The 2000 Ling-TeX Archives

1st February 2000: OT tableaux

From: John Bell 
Subject: OT tableaux

Hi,

	Apologies if this has come up before.  Has anyone found a
reasonable way of shading OT tableaux?

thanks
John Bell

2nd February 2000: Re: OT tableaux

From: Christina Thiele 
Subject: Re: OT tableaux

John Bell writes:
> Hi,
> 	Apologies if this has come up before.  Has anyone found a
> reasonable way of shading OT tableaux?
> 
> thanks
> John Bell

Forgive the ignorance ... what's an `OT tableau'?

Aha! Optimality Theory! 

If you have a copy of CJL (Cdn. Jrnl. of Ling.) handy -- issue 41,4
(Dec 96), article by Doug Pulleyblank ... (sorry, not on the web). 

If that's what you mean, then I can tell you how I did it ;-)

Use pstricks.sty in combination with colortab.sty. Print up the
colortabl.sty documentation, as well as that of pstricks.sty --
they're too handy not to have hardcopy on your shelf ;-) 

First, this is all in LaTeX 2.09 but I'll bet it's all been made to
work with 2e ... anyways, here's what I'm doing:

In my preamble:

\input pstricks.sty
\input colortab.sty

\newgray{lightgray}{.90}   <-- user-defined level of greyness

I chose 90% grey as it produces a nice even colour with my printer (HP
LP4M) and then reproduces fairly well in the issue. 

So that's the preamble stuff. The table itself is done via the usual
{tabular} and then you insert start/stop commands for the column(s) to
be greyed-in. Rather than do all the stuff here, best to leave you to
the documentation, which I found quite straightforward.

Give it a shot and see if it's `reasonable' enough for ya ;-)

Ch.

17th February 2000: Encoding hieroglyphic Luvian

From: Jed Parsons 
Subject: Encoding hieroglyphic Luvian

Greetings, all.

I should preface this message with the admission that I know absolutely
nothing about Luvian, hieroglyphic or not.  I hear they drew feet really
well.

I have recently been asked by Professor Anna M. Davies to help put
together a computer encoding scheme for hieroglyphic Luvian.  (Her purpose
is to be able to index and search text.)  Over the weekend, I composed a
minimal encoding schema that works for the two (short) documents I had
access to, and wrote a Perl script to translate encoded hieroglyphic
Luvian to HTML.  I plan to add TeX and RTF output soon.

I mention this on this list not because it's strictly a TeX-related
problem as such (yet...), but because I have the feeling that some of the
subscribers to this list might not only be among the few people in the
world who actually care or know any thing about hieroglyphic Luvian, but
also have an interest in computer encodings.  I would prefer to spend my
time on an encoding that is useful to and informed by a group of Luvian
readers rather than just one person.  So if you are intersted in
hieroglyphic Luvian, I would be grateful if you would take a look at my
first attempts at a solution.  I would welcome any suggestions or
criticism.  The relevant documents are available on my web page at
.  There is a short PDF
file describing the encoding schema, a sample encoded file, an HTML
equivalent, and the Perl script that generated it.

Many thanks to anyone who can take a look and help me improve the schema. 

Cheers,

Jed

(BTW, since you're all linguists, I'd be really delighted if you'd visit
 while you're at it...)

Jed Parsons               mailto:jed@socrates.berkeley.edu
                        http://socrates.berkeley.edu/~jed/
----------------------------------------------------------
grep(do{for(ord){(!$_&&print"$s\n")||(($O+=(($_-1)%6+1)and
grep(vec($s,$O++,1)=1,1..int(($_-6*6-1)/6))))}},(split(//,
"++,++2-27,280,481=1-7.1++2,800+++2,8310/1+4131+1++2,80\0.  What!?")));

18th February 2000: Re: Encoding hieroglyphic Luvian

From: Werner LEMBERG 
Subject: Re: Encoding hieroglyphic Luvian

> I have recently been asked by Professor Anna M. Davies to help put
> together a computer encoding scheme for hieroglyphic Luvian.  (Her
> purpose is to be able to index and search text.)  Over the weekend,
> I composed a minimal encoding schema that works for the two (short)
> documents I had access to, and wrote a Perl script to translate
> encoded hieroglyphic Luvian to HTML.  I plan to add TeX and RTF
> output soon.

I strongly suggest that you write a message to unicode@unicode.org
email list, asking whether a proposed Unicode enoding for Luvian
already exists.  This will (a) bring you probably in contact with
other Luvian experts and (b) hopefully saves you a greater amount of
time in resolving difficulties with the encoding.

    Werner

18th February 2000: Re: Encoding hieroglyphic Luvian

From: Jed Parsons 
Subject: Re: Encoding hieroglyphic Luvian

Hi, Werner, 

Thanks for your reply.

> > I have recently been asked by Professor Anna M. Davies to help put
> > together a computer encoding scheme for hieroglyphic Luvian.  
> 
> I strongly suggest that you write a message to unicode@unicode.org
> email list, asking whether a proposed Unicode enoding for Luvian
> already exists. 

Actually, unicode.org was one of the first places I went.  I didn't see
any mention of anything Luvian.  Your suggestion to contact them anyway is
a good one.  

Thanks,

Jed

Jed Parsons ...                               mailto:jed@socrates.berkeley.edu
(Just another Plautus hack)                 http://socrates.berkeley.edu/~jed/

"Try 'Wubba wubba wubba': It may not help, but it could not hurt."-Supergrover
------------------------------------------------------------------------------
use Curses;$s="Jed's another Perl hacker ";initscr;for(;;){$f=($f+10)%360;grep
(do{$g=($f-$_*10)*atan2(1,1)*4/180;for(($m,$n)=qw(cos sin)){s/.*/"11+int(12*$&
($g))"/ee}addstr($m,$n+27,(substr$s,$_,1))},0..(length $s)-1);refresh;sleep 1}

21st February 2000: Phonetic fonts: wsuipa and tipa

From: Peter Kleiweg 
Subject: phonetic fonts: wsuipa and tipa

Do you still have old LaTeX documents that use wsuipa font
commands, and a new TeX distribution without wsuipa? I wrote a
filter that translates wsuipa font commands into tipa font
commands. You can download it from CTAN:

    ftp://ftp.dante.de/tex-archive/support/wsuipa2tipa/

-- 
Peter Kleiweg       http://www.let.rug.nl/~kleiweg/

24th February 2000: UTF tools

From: Peter Kleiweg 
Subject: UTF tools

I uploaded two UTF tools to CTAN. One is for translating UTF-7
or UTF-8 into LaTeX, or any other 7- or 8-bit text format. The
other is for converting Unicode symbol numbers into UTF-7 or
UTF-8 strings.

These tools can be downloaded from:

    ftp://ftp.dante.de/tex-archive/support/utf2any/

-- 
Peter Kleiweg       http://www.let.rug.nl/~kleiweg/

3rd March 2000: Linguistic family tree

From: BP Jonsson 
Subject: linguistic family tree

I need to typeset a linguistic family tree -- moreover not the traditional
type where nodes split cleanly never to unite again; there should be dotted
lines cross-connecting some cousins and siblings.  Any suggestions how it
might be done?

/BP

3rd March 2000: Virtual fonts for linguistics

From: BP Jonsson 
Subject: virtual fonts for linguistics

I've been thinking for some time now about using the "fontinst" package to
create custom virtual fonts for linguistics/philology: subsets of TIPA
remapped for the needs of particular languages -- including prefab
letter+diacritic combos --, and similarly fonts containing diacriticized
letters as used by philologists of different languages/language groups.
There is no doubt it can be done, but I lack the TeXnical know-how and the
fontinst people are primarily concerned about remapping standard Postscript
fonts to various standard TeX encodings.  This is certainly legitimate
considering their fields of interest, but I feel that VFs have much greater
potential.  I'm also interested in the use of ligtables as a means to
simplify input of special characters -- e.g. so that typing ".d" produces
an under-dotted "d".

Any comments?

/BP

5th March 2000: Re: Linguistic family tree

From: Claire Bowern 
Subject: Re: linguistic family tree

I've typeset such trees using Tree-DVIPS; it's easy to cross-connect nodes 
that way, but I spent quite a bit of time tweaking the width of the tabular 
cells to get the spacing right. Dotted lines are done by the \dashlength 
option. I don't think it's possible to draw circles around nodes in the 
tree using this package (to represent diffusion areas), though.

Claire Bowern

At 10:16 PM 03/3/00 +0100, BP Jonsson wrote:
>I need to typeset a linguistic family tree -- moreover not the traditional
>type where nodes split cleanly never to unite again; there should be dotted
>lines cross-connecting some cousins and siblings.  Any suggestions how it
>might be done?
>
>/BP

5th March 2000: Re: Linguistic family tree

From: Emma Pease 
Subject: Re: linguistic family tree 

> At 10:16 PM 03/3/00 +0100, BP Jonsson wrote:
> >I need to typeset a linguistic family tree -- moreover not the traditional
> >type where nodes split cleanly never to unite again; there should be dotted
> >lines cross-connecting some cousins and siblings.  Any suggestions how it
> >might be done?
> >
> >/BP

Claire wrote
> I've typeset such trees using Tree-DVIPS; it's easy to cross-connect nodes 
> that way, but I spent quite a bit of time tweaking the width of the tabular 
> cells to get the spacing right. Dotted lines are done by the \dashlength 
> option. I don't think it's possible to draw circles around nodes in the 
> tree using this package (to represent diffusion areas), though.

Pstricks might do the trick perhaps in combo with other packages.

See http://www.tug.org/applications/PSTricks/

Emma

6th March 2000: Re: Virtual fonts for linguistics

From: FUKUI Rei 
Subject: Re: virtual fonts for linguistics

bpj> considering their fields of interest, but I feel that VFs have much greater
bpj> potential.  I'm also interested in the use of ligtables as a means to
bpj> simplify input of special characters -- e.g. so that typing ".d" produces
bpj> an under-dotted "d".

I agree. It is useful to create VF fonts for particular purposes.
But there has been no introductory documentation for creating VF font
files. Knuth's ``The VPtoVF processor'' (document part of vptovf.web, 
can be compiled by the weave program) contains relevant information.
Also the output of TFtoPL can be used as the basis of a new VF font.

Fukui Rei

6th March 2000: Re: Virtual fonts for linguistics

From: Alejandro Lopez-Valencia 
Subject: Re: virtual fonts for linguistics

At 10:17 PM +0100 3/3/2000, BP Jonsson wrote:
>I've been thinking for some time now about using the "fontinst" package to
>create custom virtual fonts for linguistics/philology: subsets of TIPA
>remapped for the needs of particular languages -- including prefab
>letter+diacritic combos --, and similarly fonts containing diacriticized
>letters as used by philologists of different languages/language groups.
>There is no doubt it can be done, but I lack the TeXnical know-how and the
>fontinst people are primarily concerned about remapping standard Postscript
>fonts to various standard TeX encodings.  This is certainly legitimate
>considering their fields of interest, but I feel that VFs have much greater
>potential.  I'm also interested in the use of ligtables as a means to
>simplify input of special characters -- e.g. so that typing ".d" produces
>an under-dotted "d".
>
>Any comments?
>
>/BP

Feasible. There are difficulties with the use of the \ligature command in
that it is designed for character pairs. Yet, it is not impossible to
create complex state tables with built-in intellience similar to TrueType
GX's (AAT) context sensititve morphing tables. Two excellent examples are
the klinz font, and the tibetan font CTIB. The latter has a ligature table
that eliminates the need for an input preprocessor. Right-to-left scripts
do require the use of eTeX.

- --
P. Alejandro Lpez-Valencia      Ecologist, Conservation Biologist
``Hecho de menos ...''
                                            Gabriel Garca Mrquez

7th March 2000: Line spacing

From: "Todd O'Bryan" 
Subject: Line Spacing

This isn't a particularly linguistics-related question, but I'd appreciate
any help anyone can provide.

Is there an easy way to create default line spacing of double spacing in the
body of an article, but single spacing in footnotes, examples, and tables?
I've scoured all my LaTeX resources, but can't seem to find anything. I know
to use \renewcommand{\baselinestretch}, but I'd prefer not to have to reset
the spacing every time it changes, especially since what I want is
completely predictable.

Thanks,
Todd O'Bryan
toddobryan@earthlink.net OR
obryan@ling.ucsd.edu

7th March 2000: Re: Line spacing

From: Barbara Beeton 
Subject: Re: Line Spacing

todd o'bryan asks,
    Is there an easy way to create default line spacing of double spacing
    in the body of an article, but single spacing in footnotes, examples,
    and tables?

this is what the package setspace tries to do.  you can find it
at ctan.
							-- bb

7th March 2000: Re: Line spacing

From: Alexander Holt 
Subject: Re: Line Spacing 

> Is there an easy way to create default line spacing of double spacing
> in the body of an article, but single spacing in footnotes, examples,
> and tables?

Yes, there's a package setspace.sty that's good for this.  It's in the
CTAN archive (see http://www.ctan.org/ for mirrors) at

    macros/latex/contrib/supported/setspace/setspace.sty

and there's documentation in the file itself.

Lex Holt

7th March 2000: Re: Line spacing

From: "Todd O'Bryan" 
Subject: setspace.sty

Thanks to everyone who responded so quickly with suggestions about how to
deal with single/double-spacing issues. setspace.sty worked exactly as
needed.

Todd
toddobryan@earthlink.net OR
obryan@ling.ucsd.edu

10th March 2000: Quickly, quickly! An Arabic letter for LaTeX

From: Christina Thiele 
Subject: quickly quickly! An Arabic letter for LaTeX

Does anyone know of a passable Arabic font that I can quickly access
the letter called ... Oh, damn! I dunno!

The transcription is h-underdot. Ah, here's a name: ha\cdot'

I just need it to fit into a review of a book on Arabic (natch!) and
it's just this one letter. Review's done in 9pt Times but anything'll
do! 

I've airlifted Hebrew chars in before, but this is the first time I've
needed an Arabic one. 

Anyone got a quick solution? 

Ch.

10th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: Anshuman Pandey 
Subject: Re: quickly quickly! An Arabic letter for LaTeX

Hello Christina,
There is the wonderful ArabTeX package with the accompanying Nash font. Do
you know what the character looks like? You could obviously then make the
explicit char call. (You already knew that). I'm assuming you don't know
what it looks like. In that case, I'll email you a small Postscript file
with of the character and its code within Nash.

Regards,
Anshu

On Fri, 10 Mar 2000, Christina Thiele wrote:
> Does anyone know of a passable Arabic font that I can quickly access
> the letter called ... Oh, damn! I dunno!
> 
> The transcription is h-underdot. Ah, here's a name: ha\cdot'
> 
> I just need it to fit into a review of a book on Arabic (natch!) and
> it's just this one letter. Review's done in 9pt Times but anything'll
> do! 
> 
> I've airlifted Hebrew chars in before, but this is the first time I've
> needed an Arabic one. 
> 
> Anyone got a quick solution? 
> 
> Ch.

10th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: Christina Thiele 
Subject: Re: quickly quickly! An Arabic letter for LaTeX

Anshuman Pandey writes:
> Hello Christina,
> 
> There is the wonderful ArabTeX package with the accompanying Nash font. Do
> you know what the character looks like? You could obviously then make the
> explicit char call. (You already knew that). I'm assuming you don't know
> what it looks like. In that case, I'll email you a small Postscript file
> with of the character and its code within Nash.
> 
> Regards,
> Anshu

Sorry ... I want to _avoid_ ArabTeX. I don't want a whole bunch of
overhead, just access to an Arabic font.

I do know what the char looks like -- the author has drawn it in the
margin, and I also have checked it in a book of alphabets, so no need
to trouble yourself with the.ps file. Thanks for the offer, though! 

So it's just the font. I'm tracking someone else down right now
(suggestion from another ling-tex reader) and if that fails, I'm going
to follow the suggestion made by ... the writer of the previous
message to this list, who suggested just poking through ArabTeX for a
font. 

Thanks to both of you for the ideas!

Ch.

10th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: krussll@CC.UMANITOBA.CA
Subject: Re: quickly quickly! An Arabic letter for LaTeX

On Fri, 10 Mar 2000, Christina Thiele wrote:
> I do know what the char looks like -- the author has drawn it in the
> margin, and I also have checked it in a book of alphabets, so no need
> to trouble yourself with the.ps file. Thanks for the offer, though! 

Why not just take Anshu's .ps file and plop it down (shrunk to the right
size) in the middle of the text.  I forget exactly which package does that
-- epsfig maybe.  I know I've done something like that before, but I can't
remember off the top of my head how.

-- Kevin

10th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: Christina Thiele 
Subject: Re: quickly quickly! An Arabic letter for LaTeX

krussll@CC.UMANITOBA.CA writes:
> On Fri, 10 Mar 2000, Christina Thiele wrote:
> > I do know what the char looks like -- the author has drawn it in the
> > margin, and I also have checked it in a book of alphabets, so no need
> > to trouble yourself with the.ps file. Thanks for the offer, though! 
> 
> Why not just take Anshu's .ps file and plop it down (shrunk to the right
> size) in the middle of the text.  I forget exactly which package does that
> -- epsfig maybe.  I know I've done something like that before, but I can't
> remember off the top of my head how.
> 
> -- Kevin

Hmmm ... that might be quick ... although sizing down could take some
time. So you think I should at least take Anshu's file, just in case?

Oh -- what the heck, sure, why not!

Anshuman -- if you'll forgive my hasty refusal of your PS file --
would you still be willing to send it after all? The char in question
is apparently fixed enough in its position in the alphabet that the
author's said `take the 6th letter of the Arabic alphabet and not the
eighteenth'.

And yet ...

I'm still thinking I'd be better off getting at least a whole font
... that way I get a leg up on the next paper that wants a different
Arabic character -- I'd already have it available. I know -- it's a
balancing act between getting the char right now (the pressure to
print CRC this afternoon is fast dwindling, so I seem to have a bit
more time now) and getting a solution that will apply to situations
beyond today's.

An update now -- I've had a lot of replies! I've received a .pfm and
.PFB file from someone, and also a list of all the arabtex fonts that
are on CTAN so I should now be able to take all the raw material and
get what's needed ... by whatever route ;-) I'll post a message after
I get everything sorted out ... 

Thanks again, everyone. This is a great list!

Ch.

11th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: Anshuman Pandey 
Subject: Re: quickly quickly! An Arabic letter for LaTeX

Dear Emmon,
Thank you! However, my response to Christina Thiele was in regard to
producing the actual Arabic character represented by h-underdot.

Regards,
Anshuman Pandey

On Sat, 11 Mar 2000, Emmon Bach wrote:
> To Anshuman Pandey;
> 
> If h with an underdot is what you want,m you can use \d h (pharyngeal
> fricative), as in the TeX Book, p. 52
> 
> Cheers, 
> Emmon Bach
> 
> 144 N Valley Rd
> Pelham, MA 01002 - 9766	
> home: (413) 253 7793		
> 
> Department of Linguistics
> University of Massachusetts
> Amherst, MA 01002
> office: (413) 545 0889
> email: ebach@linguist.umass.edu
> webpage: 

12th March 2000: Re: Quickly, quickly! An Arabic letter for LaTeX

From: BP Jonsson 
Subject: Re: quickly quickly! An Arabic letter for LaTeX

At 11:54 -0500 10.3.2000, Christina Thiele wrote:
>Sorry ... I want to _avoid_ ArabTeX. I don't want a whole bunch of
>overhead, just access to an Arabic font.
>
>I do know what the char looks like -- the author has drawn it in the
>margin, and I also have checked it in a book of alphabets, so no need
>to trouble yourself with the.ps file. Thanks for the offer, though!

Just get the font, run it through DEK's font-table macro (what's it called
again?) to get the charcode for the letter, then use \char" to
insert it!  I do that all the time with Sanskrit, Tibetan and TIPA when
needing just an odd character.

/BP

13th March 2000: ha' (Arabic letter) success

From: Christina Thiele 
Subject: ha' (Arabic letter) success

Well, \Ha is now in place in my review! 

In response to my message from Friday, I received a great number of
replies and I'd like to pass on what I've found out ...


Like a returning boomerang, the first reply came from Steve Grathwohl
(Duke University Press), who recommended I contact Ahmed Hindawi.
Later that day, Mr. Hindawi sent me a PostScript Arabic font from a
recent Omega distribution, the .pfm and .PFB files. He cautioned:

   You should be able to use this font with LaTeX with no problems.
   Although I have to say be careful, the Omega fonts do not seem to
   be recognized by Win 2000 as proper Type 1 fonts. This font was
   actually designed to be used with Times. It should harmonize fine
   with it.

I figured I could install the font via the Adobe Type Manager (ATM) on
my laptop (where I have Y&YTeX running on NT). How I'd get the font
file(s) necessary for TeX back to the unixbox and stored in the right
places ... well ... one problem at a time, right? 

Unfortunately, when I tried to add the font via the ATM, all I got was
this error message:

   Cannot copy file omsea1.PFB to C:\PSFONTS\. 

Hmmm ... well, the laptop is networked to the unixboxes, which is
where I had first ftp'd the files (in binary) from another machine
where mail is delivered. Perhaps the laptop couldn't copy a file from
the networked `E' drive ... So I copied the two files into C:\TEMP but
got the same error message. Agh! The good news, however (although I
was unable to get this a second time), was that I was able to view the
PS font file and could see lovely Arabic shapes on the screen ;-) It
certainly looked like the file was intact and not corrupted so it had
to be my `procedures' ... I just don't know ATM and NT and PCs all
that well ... bother. 

Nevertheless, I'm sure there's a way to get these PS files to work and
I certainly thank Mr. Hindawi for having been so kind as to send them
to me. 

Now ... I did have a backup plan ...

Throughout the afternoon on Friday I had had loads of messages from
people with suggestions and information -- it was great!

Some suggested going to ArabTeX but that seemed like an awful lot of
installation overhead for a single character. Nevertheless, it was to
the arabtex/ directory on CTAN that I did end up going ... for the
fonts, that is ;-)

So, for pointing me in that direction, I'd like to thank Joost Kremers
(University of Nijmegen), Anshuman Pandey (U of Washington) who
specified the Nash fonts in ArabTeX and also offered to send me a PS
file with just the one character, Barbara Beeton (AMS) who sent a
listing of all the fonts in the arabtex/ directory on CTAN, and then
Klaus Lagally, the main person behind ArabTeX, who zeroed me in on
xnsh14 (also available in PS, he says, thanks to Taco Hoekwater).

Other suggestions came from Michael W. Daniels (Ohio State), Jacques
Andre (IRISA), Kevin Russell (U Manitoba), Alexis Dimitriadis (U of
Penn), and BP Jonsson (Sweden). 

What I ended up doing was this:

a. looking at the CTAN ArabTeX fonts, which meant staying on the
   unixbox, where I'm most comfortable at the moment -- and what 
   systems-specific stuff I don't know, my husband can help with ;-)) 

b. going to the specific CTAN directory and fetching the xnsh14.tfm
   file

c. taking a home-grown font-table generating file to get a table 
   of all the characters ... (where I do exactly as BP suggested with
   respect to identifying and accessing the specific char.)

   Well ... step c. bombed ;-( 

   It took a while (for me) to notice that while TeX will process the
   file, the previewer (xdvi) says I dunno what you want! On closer
   inspection (aka reading the screen ;-) ), I noticed that I needed
   the matching .mf file. OK. Let's go get it ... 

d. fetch xnsh14.mf from CTAN

e. process my font-table.tex file ... for more error messages

   Eventualy, I ended up fetching all the necessary .mf files
   (dvips did a good job of saying exactly which file it couldn't find
   ... ) to generate the font-table (which didn't display the font
   completely but when I ran it through the printer, the whole
   chart filled out ever so nicely ;-) ). So keep that in mind -- 
   if someone says use font X, figure you'll need not just X.tfm
   but a bunch of other .mf files as well. Of course, if you're
   used to adding fonts, the above demonstrates pathetic ignorance but
   hey! I can't know everything ... even after all these years ;-)


So, my lessons are the following:

1. post message to ling-tex to get the widest possible range of
   suggested options! 

2. try everything (including offers of PS files -- thanks, Anshu! --
   of just the specific character) that you have available to you, in
   terms of hardware/software

3. .tfm files need .mf files ... and more than just one ;-) 

4. I also found out that I need to have all these files in the same
   directory as the source file that has to access them. Around here,
   we'd rather not add things to the TeXLive texmf/ tree that will 
   be overwritten when a new CD update comes out, so I keep things
   local. Unfortunately, we don't seem to be able to set up a 
   path for fonts to my own set of macros in /texlib so if anyone
   has suggestions on how to set up fonts outside of texmf/ that
   avoids the inelegance of having .tfm and .mf files in a
   working directory, I'd be much obliged! We've tried setting
   the path in the TEXINPUTS line in my environment but that
   doesn't seem to work ... 

Additional bonus: 

4. by getting the spouse involved, I now know easier ways to
   download files via Netscape ;-))


So, to conclude, here's the code that now allows me to get exactly
what I want and need:

   \font\Arabic=xnsh14 at 9pt
   \def\Ha{{\Arabic\char'150}}


... plus a small adjustment ...

I believe the Arabic chars. are positioned correctly vis-a-vis the
baseline. However, when I used the \Ha macro into a line of English
text, its top stroke was lower than the x-height and it looked
off-balance -- in an English surround, I mean. So I raised the \Ha
char. until its top horizontal stroke was pretty much even with the
x-height of the English:

   ... traditional /\barh/ (the Arabic letter \raise.4ex\hbox{\Ha}\ts)

And here's my hand at ASCII-art of \Ha, so you can see what the
general shape is:

                      hhhhhhhhhhhhhhh
                        hhhhh
                       h
                      h
                     h
                      h
                       hh          hh
                          hh    hh
                             hh

     
====

So, I did end up getting what I wanted: for the short term, a decent
Arabic character (ha'), and for the long term, a useable Arabic font
for possible future situations (rare though they are in the texts I
see).

I'd like to thank everyone again for being so quick to answer a Friday
morning call for help, with suggestions on what fonts to look for, and
with actual font files and test files.

We may not have massive volume on this list but boy, can we get a good
answer fast!

May the week be as positive for everyone else as it looks to be for
me!

Ch.

P.S. The review is by Harold Paddock, of Alan S. Kaye (ed.), {\it
     Phonologies of Asia and Africa\/}, 2 vols. 1997. Scheduled 
     to appear in CJL 44(1). 

13th March 2000: Footnotes in tables

From: Claire Bowern 
Subject: Footnotes in Tables

Hi all,
Apologies if this is a dumb question. I can't get footnotes in tables and 
certain other environments such as tabbing to appear. The number is 
computed but the footnote is missing from the bottom of the page. For 
example, if fn 3 is in a table, fn 2 and fn 4 will appear but fn 3 
disappears somewhere. Is there any way to get around this?

Thanks
Claire
_________________________
Claire Bowern

Department of Linguistics
Harvard University
305 Boylston Hall
Cambridge, MA  02138
ph: 617-493-4230
http://www.fas.harvard.edu/~lingdept/

14th March 2000: Re: Footnotes in tables

From: Dag Langmyhr 
Subject: Re: Footnotes in Tables

-> Hi all,
-> 
-> Apologies if this is a dumb question. I can't get footnotes in tables and 
-> certain other environments such as tabbing to appear. The number is 
-> computed but the footnote is missing from the bottom of the page. For 
-> example, if fn 3 is in a table, fn 2 and fn 4 will appear but fn 3 
-> disappears somewhere. Is there any way to get around this?
-> 
-> Thanks
-> Claire

Try splitting the footnote; use \footnotemark in the table and
\footnotetext{Some text \dots} after it.

						Dag

14th March 2000: Re: F'notes in tables + useful books

From: Christina Thiele 
Subject: Re: f'notes in tables + Useful books

Claire Bowern writes:
> Hi all,
> 
> Apologies if this is a dumb question. I can't get footnotes in tables and 
> certain other environments such as tabbing to appear. The number is 
> computed but the footnote is missing from the bottom of the page. For 
> example, if fn 3 is in a table, fn 2 and fn 4 will appear but fn 3 
> disappears somewhere. Is there any way to get around this?
> 
> Thanks
> Claire

Not a dumb question ... but a very frustrating experience for sure!
I see Dag's given you a possible solution, which does seem to be the
one that's suggested in various references. 

If you have Lamport's _LaTeX_ book, see p.105 for some additional
hints. 

Also, in the Goossens, Mittelbach, and Samarin _Companion_, on
pp.132--133, there are samples and explanations of why this doesn't
work and how to get what you want ... as opposed to what LaTeX thinks
you told it ;-)

Similar information can be found in the Kopka and Daly book, _A Guide
to LaTeX2e_. Often I'll check all three books, because their approach
and presentation are slightly different -- often where one will read
like Greek to me, another will be clear as crystal. 

All of the above refs were quickly found via the index to the book
(not all items can be found so easily but it's really nice when it
works!) so I keep them within reach of my right hand here ... It
really is worth it to have these three books, at the very least, on
hand. 

And to continue the digression ... I find myself dipping into _TeX for
the Impatient_ quite often (written by Abraham, with Berry and
Hargreaves) and, of course, the good old _TeXbook_. 

Others that I have on my shelf and which I do look at on occasion
include: Jahn's _LaTeX for Everyone_, Eijkhout's _TeX by Topic_, and
Schwartz's _Introduction to TeX_. 

Anyone else have preferences or suggestions? [I'll add `useful books'
to the subject line]

Ch.

14th March 2000: Re: F'notes in tables + useful books

From: Christina Thiele 
Subject: Re: f'notes in tables + Useful books

Christina Thiele writes:
> ...
> Others that I have on my shelf and which I do look at on occasion
> include: Jahn's _LaTeX for Everyone_, Eijkhout's _TeX by Topic_, and
           ^
           Hahn 

> Schwartz's _Introduction to TeX_. 
> ...

Oops!

Ch.

14th March 2000: Re: Footnotes in tables

From: Claire Bowern 
Subject: Re: Footnotes in Tables

Thanks to everyone who suggested solutions to the footnote problem!
It's working fine now....

Claire

____________
Department of Linguistics
Harvard University
305 Boylston Hall
Cambridge  MA  02138
Ph: (617)-493-4230

14th March 2000: TIPA monospaced

From: Mike Hammond 
Subject: tipa monospaced

all

I've been using TIPA for IPA symbols (it's great), but does anybody
know of a monospaced IPA font for Latex?

mike hammond

14th March 2000: Re: TIPA monospaced

From: "Michael Covington" 
Subject: Re: tipa monospaced

It would be nice to have.  I once faked it by putting every character in an
mbox.

Michael A. Covington - Artificial Intelligence Ctr - University of Georgia
http://www.ai.uga.edu/~mc  http://www.CovingtonInnovations.com  <><

----- Original Message -----
From: "Mike Hammond" 
To: 

> all
>
> I've been using TIPA for IPA symbols (it's great), but does
> anybody know of a monospaced IPA font for Latex?
>
> mike hammond

14th March 2000: Re: TIPA monospaced

From: Peter Kleiweg 
Subject: Re: tipa monospaced

Mike Hammond skriver...
> I've been using TIPA for IPA symbols (it's great), but does
> anybody know of a monospaced IPA font for Latex? 

You can download the SIL IPA fonts from the IPA web site:

    http://www.arts.gla.ac.uk/IPA/ipa.html

Comes in three flavors: serif, sans-serif and monospaced; and
four faces: regular, bold, bold-italic, italic. The fonts
don't have as many symbols as the TIPA fonts.

These are TTF fonts.

To convert TTF font to Type1 font:

    run: ttf2pt1 -e file.ttf fontname

To use Type1 font in LaTeX:

    run: afm2tfm font -v font rfont
    the previous command results in a line of text you should
      add to your psfonts.map file. At the end of this line
      add: <font.pfb
    run: vptovf font.vpl font.vf font.tfm
    move *.vf and *.tfm files to where TeX looks for them
    run: tex testfont
       type commands:
         rfont at 16pt
         \sample
         \bye
    run: dvips testfont

In all this, change 'font' with the appropriate name.

-- 
Peter Kleiweg
http://www.let.rug.nl/~kleiweg/

21st March 2000: Examples

From: "Todd O'Bryan" 
Subject: Examples

Let me explain what I need to do, and if someone has a good idea of how to
do it, let me know...

I'm using gb4e.sty to automatically number examples throughout a paper.
However, I need to include three sub-examples as part of a single example,
as in:
(4) a. example1
    b. example2
    c. example3

No problem, you say. Use the xlist environment to get the alphabetic
numbering. Well, the problem is that each sub-example is actually four
forms: a single verb in the past, present, and subjunctive plus a gloss. Now
obviously I'd like the forms in each tense and the glosses to line up
vertically, so I'd like to use a tabular environment, or something easy to
do that gets the same result. And, thus, the problem. If I use tabular, I
can't have an xlist inside, unless it's confined to a single cell, which is
not what I want.

What I'd like is:
(4) a. exa1     exa2    exa3    gla
    b. exb1     exb2    exb3    glb
    c. exc1     exc2    exc3    glc

Any ideas?

Todd

21st March 2000: Re: Examples

From: Alexis Dimitriadis 
Subject: Re: Examples

> I'm using gb4e.sty to automatically number examples throughout a paper.
> What I'd like is:
> (4) a. exa1     exa2    exa3    gla
>     b. exb1     exb2    exb3    glb
>     c. exc1     exc2    exc3    glc

> obviously I'd like the forms in each tense and the glosses to line up
> vertically, so I'd like to use a tabular environment, or something easy to
> do that gets the same result. And, thus, the problem. If I use tabular, I

It's easy to do it right: Just define your own counter, and use it wherever
you want, independent of enclosing environments.

The attached code works with gb4e.  To use in other contexts, either modify
the definition of the {letter} counter or just say \usecounter{letter}
whenever you want it to go back to counting from a.


One more suggestion: if you never have more than 3 lines to line up, don't
use tabular-- use cgloss4e, which comes with gb4e.  The example below does one
set with tabular and one with cgloss4e.

Alexis

PS. If you want to be able to use labels (\label, \ref) with the individual
lines, you'll need a couple more lines of code; let me know and I'll send them
your way.


\begin{document}

\usepackage{gb4e,cgloss4e}

% Define counter named ``letter;'' will be zeroed whenever xnum is incremented
\newcounter{letter}[xnumi]
%
% The standard name for using the counter:
\def\theletter{\alph{letter}} 
%
% This one you can name whatever you want.  I chose \lmark
\def\lmark{\stepcounter{letter}\theletter. }


\begin{document}

\begin{exe}
\ex
\begin{tabular}[t]{lllll}
\lmark & AAA & etc. & and more \\
\lmark & BBB & .tc \\
\end{tabular}

\ex \glll
\lmark{} AAA etc.   and~more here \\
\lmark{} BBB .tc    xx       there \\
\lmark{} a   third  line here \\    
\end{exe}

\end{document}

22nd March 2000: Re: Examples

From: FUKUI Rei 
Subject: Re: Examples

Dear Todd O'Bryan

toddobryan> What I'd like is:
toddobryan> (4) a. exa1     exa2    exa3    gla
toddobryan>     b. exb1     exb2    exb3    glb
toddobryan>     c. exc1     exc2    exc3    glc

If each exmaple is not so long and its length is predictable,
it is not difficult to define a small macro like:

  \newcommand\exbox[4]{\leavevmode
    \hbox to *cm{#1\hfil}\hbox to *cm{#2\hfil}\hbox to *cm{#3\hfil}%
    \hbox{#4}}%

and then you can write something like

  \exbox{exa1}{exa2}{exa3}{gla}

within any kind of environments.

Fukui Rei

18th April 2000: Telling which column in `multicol' environment

From: "TOYOSHIMA,Masayuki" 
Subject: telling which column in multicol environment

Is there a way to tell in \begin{multicols}{n} ... \end{multicols}
environment `on WHICH COLUMN I am now' ?

I just want to make page references, not simply `p.123', 
but `p.123L' (left) or `p.123-3' (the 3rd column), hopefully by fixing
\label.

TOYOSHIMA,Masayuki / mtoyo@aa.tufs.ac.jp
Institute for the study of languages and cultures of Asia and Africa
Tokyo university of foreign studies, Kita-ku, Tokyo, 114-8580, JAPAN

23rd April 2000: Chinese boxes

From: Joseph Hilferty 
Subject: Chinese boxes

I've learned a little LaTeX in order to do construction grammar
representations. Construction grammar uses Chinese boxes (= nested
boxes) to represent constituent structure and AVMs to represent
the information encoded in each constituent.

Christopher Manning's AVM.STY works nicely for doing the AVMs,
but doing the boxes is a pain in the neck. I use \framebox to get
the boxes, which is easy enough. However, each sister box has to
be of equal height, so I try to fake it with the \raisebox command.
Unfortunately, this takes a lot of time (and patience!) to do,
because you have to calculate the extend-above and extend-below
arguments through trial and error.

Does anyone know of a macro that would do this for me automatically?

Thanks,
Joe Hilferty
__________________________________________________________
Home page: http://lingua.fil.ub.es/~hilferty/homepage.html

23rd April 2000: Re: Chinese boxes

From: Peter Kleiweg 
Subject: Re: Chinese boxes

Joseph Hilferty skriver...
> Christopher Manning's AVM.STY works nicely for doing the AVMs,
> but doing the boxes is a pain in the neck. I use \framebox to get
> the boxes, which is easy enough. However, each sister box has to
> be of equal height, so I try to fake it with the \raisebox command.
> Unfortunately, this takes a lot of time (and patience!) to do,
> because you have to calculate the extend-above and extend-below
> arguments through trial and error.

Include a rule with zero width (an invisible rule) and desired
height:

    \fbox{\rule[-2mm]{0cm}{6mm}Text}

-- 
Peter Kleiweg
http://www.let.rug.nl/~kleiweg/

25th April 2000: Re: Chinese boxes

From: Chris Manning 
Subject: Re: Chinese boxes

On 23 April 2000, Joseph Hilferty wrote:
 > Christopher Manning's AVM.STY works nicely for doing the AVMs,

Thanks!  If anyone else needs it, you can get it at:

	http://nlp.stanford.edu/~manning/tex/

 > but doing the boxes is a pain in the neck. I use \framebox to get
 > the boxes, which is easy enough. However, each sister box has to
 > be of equal height, so I try to fake it with the \raisebox command.
 > Unfortunately, this takes a lot of time (and patience!) to do,
 > because you have to calculate the extend-above and extend-below
 > arguments through trial and error.

Including a zero width rule of the same height in all boxes would be the 
easiest way to get them all the same (as Peter Kleiweg) suggested.

But it would be nice if this could be done automatically.  I've been
meaning for a while to revise avm.sty, and might see if I can add this
(but it's non-trivial, because you'll have to set all of the boxes to
work out which one is biggest before you can pad the others).

I might note that in practice a lot of construction grammar work seems
to just not bother with getting them all the same height.  E.g., in the
Kay & Fillmore What's X doing Y paper, a couple of diagrams have them
the same height but the rest don't....

Chris Manning

8th May 2000: Cirled letters within words

From: Christina Thiele 
Subject: cirled letters within words

Hello everyone ... this month's problem is ... ;-(

I have some words which need a letter (or two) within them circled,
and then, a curve drawn connecting to the two circles. 

If I use PSTricks, the letters before/after the circled letter(s) are
pushed away (to the left and right) of the drawn circle. I'm looking
for a circle that won't perturb the usual inter-letter spacing. 

Added wrinkle: the cirles are sized according to their contents
whereas I'd like the same circle size for all instances ... well, more
or less. At the very least, some mechanism that allows me to control
the circle diameter so I can do fine-tuning. 

If I use Emma's tree-dvips, the letter spacing in the word is left
unperturbed ... but the cirle over-writes the surrounding text! So
that's not really helpful either. 

I need a hybrid of the two! No change in inter-letter spacing, but no
over-writing of the neighbouring characters. 

I've got the _Graphics Companion_ and GUTenberg Cahier 16 (about
PSTricks) and the PSTricks manual (oooh ... v.0.93 from 1993). Does a
newer PSTricks version address this? If that's all that it'll take,
then there's my answer! 

I hope someone knows how to solve this one ;-)

Ch.

8th May 2000: Re: Cirled letters within words

From: Emma Pease 
Subject: Re: cirled letters within words 

> I have some words which need a letter (or two) within them circled,
> and then, a curve drawn connecting to the two circles. 

Pstricks is now version .97 see

http://www.tug.org/applications/PSTricks/

for latest info.

The manual I don't think has been updated and I think there is a separate
manual for pst-node. 

I'm not sure whether what you want is possible, is there space between
the letters for circles?

My first stab would be to use pstricks and to do the following

***start sample
\documentclass{article}

\usepackage{pst-node}  % this automatically loads pstricks
\begin{document}

This is a test of
retr\Rnode{a}{\psframebox[linewidth=.4pt,framesep=1pt,framearc=.5,boxsep=false]{o}}sp\Rnode{b}{\psframebox[linewidth=.4pt,framesep=1pt,framearc=.5,boxsep=false]{e}}ct


\ncbar[linewidth=.4pt,nodesep=1pt,angle=-90]{-}{a}{b}

\end{document}

***end sample


What the various bits do:

\Rnode - creates a node with reference point lined up with the baselines
\psframebox - creates a box around the item.  I think you are better
     with a rounded corner box than a pure circle if you don't want to
     overlap text
    linewidth - change the linewidth to .4pt instead of default .8pt
    framesep - change space between frame and text to 1pt from 3pt
    framearc - rounds corners, can range from 0 to 1
    boxsep - make the box irrelevant for calculating size (i.e.,
             psframebox will not affect layout)

\ncbar - to draw the line between the two nodes (you can change this
         to a curve of some sort)
    linewidth - need a thinner line
    nodesep - to compensate for the box not taking up space, should be
              the same as framesep in the psframebox command if you
              want the line touching the box. 

You could try \pscirclebox instead of \psframebox with rounded
corners, but, I don't think it is possible not to overlap text with
pscirclebox.  

Emma

8th May 2000: Re: Cirled letters within words

From: Christina Thiele 
Subject: Re: cirled letters within words

Emma Pease writes:
> > I have some words which need a letter (or two) within them circled,
> > and then, a curve drawn connecting to the two circles. 
> 
> Pstricks is now version .97 see
> 
> http://www.tug.org/applications/PSTricks/
> 
> for latest info.

Ah. Thanks. I'll get the upgrade tomorrow, if I can. 
 
> The manual I don't think has been updated and I think there is a separate
> manual for pst-node. 

OK. I'll look for the pst-node manual -- don't have that one printed
up. 

> I'm not sure whether what you want is possible, is there space between
> the letters for circles?

No. It's to show the letter(s) involved in slips of the tongue ;-)

> My first stab would be to use pstricks and to do the following
> 
> ***start sample
> \documentclass{article}
> 
> \usepackage{pst-node}  % this automatically loads pstricks
> \begin{document}
> 
> This is a test of
> retr\Rnode{a}{\psframebox[linewidth=.4pt,framesep=1pt,framearc=.5,boxsep=false]{o}}sp\Rnode{b}{\psframebox[linewidth=.4pt,framesep=1pt,framearc=.5,boxsep=false]{e}}ct
> 
> \ncbar[linewidth=.4pt,nodesep=1pt,angle=-90]{-}{a}{b}
> 
> \end{document}
> 
> ***end sample

Prett damn'd near fatal stab! I've just tested it with two sample
words -- central + choice. And the problem immediately arises: how to
get the two circles to be the same and NOT measure the size of the
letter(s) they're enclosing. That's one trick. 

But the problem with overprinted material no longer showing has been
resolved -- so you've got my attention ;-) I'm going to play with this
some more tomorrow and try to get some fine-tuning done. 

The other think I need to fiddle with is the connecting line: ms shows
a curve rather than the right-angled lines -- more like your
\nodecurve in tree-dvips. 

But this is a fabulous start, Emma. Can't thank you enough for having
problem spent a good portion of your afternoon on it. 

Best of all, though, is the description below of what the bits do. I
really really appreciate the info!
 
> What the various bits do:
> 
> \Rnode - creates a node with reference point lined up with the baselines
> \psframebox - creates a box around the item.  I think you are better
>      with a rounded corner box than a pure circle if you don't want to
>      overlap text
>     linewidth - change the linewidth to .4pt instead of default .8pt
>     framesep - change space between frame and text to 1pt from 3pt
>     framearc - rounds corners, can range from 0 to 1
>     boxsep - make the box irrelevant for calculating size (i.e.,
>              psframebox will not affect layout)
> 
> \ncbar - to draw the line between the two nodes (you can change this
>          to a curve of some sort)
>     linewidth - need a thinner line
>     nodesep - to compensate for the box not taking up space, should be
>               the same as framesep in the psframebox command if you
>               want the line touching the box. 
> 
> You could try \pscirclebox instead of \psframebox with rounded
> corners, but, I don't think it is possible not to overlap text with
> pscirclebox.  

I think your rounded rectangles are better in that they already take
up less vertical space in line. I hadn't thought of that option. 

More tomorrow -- I hope! And thanks again for the clear pointers. 
 
> Emma

Ch.

8th May 2000: Re: Cirled letters within words

From: Barbara Beeton 
Subject: Re: cirled letters within words

christina's problem:
    ... I've just tested it with two sample
    words -- central + choice. And the problem immediately arises: how to
    get the two circles to be the same and NOT measure the size of the
    letter(s) they're enclosing. That's one trick. 

suggestion:  decide on a common size.  then
 - incorporate a strut of the appropriate size into the macro
 - apply \smash to the letter being circled
code for both strut and \smash can be stolen from the tugboat macros.

happy circling.  (i'd like to see some samples when you've got
it working.)
							-- bb

9th May 2000: Re: Cirled letters within words

From: Christina Thiele 
Subject: Re: cirled letters within words

Success!!! I have just spent the last hour refining the code -- but
without the tremendous assist from Emma to start me off, I'd have
never found this. This morning, I found that Emma had posted a further
refinement, while Barbara had posted a suggestion regarding consistent
height ... and it all now looks just loverly!


So, here's how it all came together.

First, Emma posted the following sample yesterday (I've inserted
forced line breaks to make this display nicely):

   This is a test of retr\Rnode{a}{\psframebox[linewidth=.4pt,
   framesep=1pt,framearc=.5,boxsep=false]{o}}sp\Rnode{b}%
   {\psframebox[linewidth=.4pt,framesep=1pt,framearc=.5,%
   boxsep=false]{e}}ct

   \ncbar[linewidth=.4pt,nodesep=1pt,angle=-90]{-}{a}{b}


The circles were now boxes with rounded corners, which pushed less
`circling' material above/below the actual characters being
delimited. I liked that change ... 

The \ncbar, however, described a rectangular connection between the
two `circles' -- I wanted a curve ... 

I figured I'd work out the niggles this morning but I was pre-empted
by Emma, who'd spend yet more time on that curve and posted this:

   Something like
  
   \ncangle[linewidth=.4pt,nodesep=.5pt,angleA=-90,angleB=-90,%
      linearc=.15]{-}{a}{b}

   or

   \nccurve[linewidth=.4pt,angle=-90,ncurv=1.5]{a}{b}

   in place of \ncbar should do the trick for curves. linearc affects
   the curve of the corners in ncangle.

But she thought getting the circles to be more similar in size would
be tricky. 

Luckily, Barbara posted a message which addressed precisely this
issue:

   suggestion:  decide on a common size.  then
    - incorporate a strut of the appropriate size into the macro
    - apply \smash to the letter being circled
   code for both strut and \smash can be stolen from the tugboat
   macros.

So off I went to the TUGboat macros and then sort-of found the right
thing. I haven't implemented the \smash bit because I think I've got
the effect I want. I think ... 

A bit more testing and tweaking and I've now got the following, with
macros to take care of all the gory details:

   %% the strut thing from BNB:
   \def\Strut{\vrule height 6pt width 0pt}  

   %% new macro with Emma's def'n, but with some changes: 
   %%   a. framearc changed from 1.5 to just .5
   %%   b. inserted newly def'd \Strut
   \def\Circlenode#1{{\psframebox[linewidth=.4pt,framesep=1pt,
      framearc=.5,boxsep=false]{#1\Strut}}} 

   %% new macro using Emma's \ncurve macro, modified:
   %%   a. changed ncurv from 1.5 to .5, to make a shallower arc
   %%   b. added a nodesep of 1.5pt, so that the start/end points
   %%      of the curve were touching the circle, not the centre of
   %%      the node ... in the letter's shape itself
   \def\Circleconnect#1#2{{\nccurve[linewidth=.4pt,nodesep=1.5pt,%
      angle=-90,ncurv=.5]{#1}{#2}}}


And here's a sample of the material being encoded:

\begin{tabular}{lll}
\quad Onset   & \Rnode{a}{\Circlenode{c}}entral
\Rnode{b}{\Circlenode{ch}}oir
                              & \lingarrow\ kwentral sire \\ [6pt]
\quad Nucleus & f\ts\Rnode{c}{\Circlenode{ee}}\ts t
m\Rnode{d}{\Circlenode{o}}ving   
                              & \lingarrow\ fuwt meeving  \\ [6pt]
\quad Coda    & Lloy\Rnode{e}{\Circlenode{d}}
Mo\Rnode{f}{\Circlenode{s}}eby  
                              & \lingarrow\ Lloyz Modeby  \\ [10pt]
%
\quad Wo\Circlenode{rd\yy{1}/Wo}rd\yy{2}\nn{a}
              & he\Circlenode{fty/hu}sky   
                              & \lingarrow\ hesky  \\
\end{tabular}
%
\Circleconnect{a}{b}  \Circleconnect{c}{d}  \Circleconnect{e}{f}


Emma had also mentioned that perhaps a proportional font would help
deal with the variable width. I didn't do that but I have inserted
some \ts (=\thinspace) here and there for narrow letters that are
otherwise being quite hidden by the circling lines -- f, t, and i are
the main chars. at issue. 

And the other option, of using a grey line rather than black, I
haven't yet exercised. But it's possible here on my printer (an old
LJ4M) so I may yet see how that works. Of course, it's not always sure
that what I generate on hardcopy will end up being printed as cleanly
for the multiple copies of the journal but still ... a nice master is
always a good start ;-) 

===

So, many many thanks to both Emma and Barbara! It's such a small part
of this paper but it's still so nice to be able to get it all done
without hand-drawing!

If anyone finds a use for this and makes additional refinements,
please post so we can all make better `drawings' ;-) 

Thanks again. I am, as always, tremendously grateful -- and delighted
-- that this list continues to exist!

Ch.

9th May 2000: Re: Cirled letters within words

From: Christina Thiele 
Subject: Re: cirled letters within words

Christina Thiele writes:
> ...
> Emma had also mentioned that perhaps a proportional font would help
                                         ^^^^^^^^^^^^
NO NO NO!  She suggested a _fixed-width_ font. Sorry, Emma.

> ...

Ch.

10th May 2000: Chinese boxes, more thoughts

From: "David J Weber" 
Subject: Chinese boxes, more thoughts

Friends,
	Some years ago I was teaching a Quechua course in which
there were two Quechua students. I found that they grasped
"Chinese" boxes much more easily than trees or labeled
bracketings. (Well, my North American students did too!)
	The sort of diagram I wanted would include the Quehcua
text on the left and a functional/categorial label flush
right. I am attaching two attempts I did brute force. They
are ugly, especially since accented characters are
exceptional in tabbing environments, and because I had to
resort to parboxes with fixed widths...).
	I had hoped that a macro package would make entry
managable. I gave up on it. I thought input like that given
below might be managable, and give the author the
advantages of the delimiter matching and coloring that
comes with TeX or LaTeX mode in some emacs editors.
	Perhaps the current efforts along these lines might
take my thoughts into account. I'd love to be able to use
such diagrams easily.

Best, --David Weber

{chaynaw}{NX}
{awnisha}{V0}
{{{imata}{OD}{marka kumun runakuna}{SU}{munaptin}{R0}}{CND}
 {yanapa:nanpaq}{S0}
}{PRP}
{{{llapan awturida:kuna}{SU}{Pillkuchaw}{LOC}{kaqkunaqa}{S0}}{SU}
 {{{markachaw}{LOC}{tiyaq}{S0}}{MOD}{runakuna}{S0}}{SU}
  {{{kikin markachaw}{LOC}{wasita}{OD}{rurananpaq}{S0}}{PRP}
  {ristamuta}{OD}
  {ma\~{n}aptin}{R0}
 }{CND}
 {{huk inhinyeru}{SU}
  {{markaman}{META}{aywaykur}{R0}}{CND}
  {rikaptinraq}{R0}
 }{CND}
 {qonanpaq}{S0}
}{PRP}

David Weber (david_weber@sil.org)

10th May 2000: Chinese boxes, more thoughts

From: "David J Weber" 
Subject: Chinese boxes, more thoughts

Text from file 'DIAG.TEX'

\documentstyle{article}
\setlength{\topmargin}{-1.0in}
\setlength{\textheight}{10.3in}
\setlength{\textwidth}{7.0in}
\setlength{\oddsidemargin}{-.5in}
\setlength{\evensidemargin}{-.5in}
% \setlength{\headheight}{0.0in}
% \setlength{\headsep}{0.0in}
% \setlength{\footheight}{0.0in}
% \setlength{\footskip}{0.0in}
% \renewcommand{\baselinestretch}{.7}
% \newcommand{\as}{}%{\hspace{.1in}\vspace{.1in}}
% \pagestyle{empty}
% \raggedbottom
% \sloppy

\begin{document}
\fbox{\parbox{3.5in}{
      \begin{tabbing}
      chaynaw \`NX\\
      awnisha \`*V0\\
      \fbox{\parbox{3in}{
            \begin{tabbing}
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  imata\`OD\\
                  marka kumun runakuna\`SU\\
                  munaptin\`*R0
                  \end{tabbing}
                 }}\`CND\\
            yanapa:nanpaq\`*S0
            \end{tabbing}
           }}\`PRP\\
      \fbox{\parbox{3in}{
            \begin{tabbing}
            \fbox{\parbox{2.5in}{
                 \begin{tabbing}
                 llapan awturida:kuna\`SU\\
                 Pillkuchaw\`LOC\\
                 kaqkunaqa\`*S0
                 \end{tabbing}
                }}\`SU\\
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        markachaw tiyaq runakuna\`*S0
%                        \fbox{\parbox{1.5in}{
%                              \begin{tabbing}
%                              markachaw\`LOC\\
%                              tiyaq\`*S0
%                              \end{tabbing}
%                             }}\`MOD\\
%                        runakuna\`*S0
                        \end{tabbing}
                       }}\`SU\\
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        kikin markachaw\`LOC\\
                        wasita\`OD\\
                        rurananpaq\`*S0
                        \end{tabbing}
                       }}\`PRP\\
                  ristamuta\`OD\\
                  ma\~{n}aptin\`*R0
                  \end{tabbing}
                 }}\`CND\\
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  huk inhinyeru\`SU\\
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        markaman\`META\\
                        aywaykur\`*R0
                        \end{tabbing}
                       }}\`CND\\
                  rikaptinraq\`*R0
                  \end{tabbing}
                 }}\`CND\\
            qonanpaq\`*S0
            \end{tabbing}
            }}\`PRP
      \end{tabbing}
      }}
\end{document}

10th May 2000: Chinese boxes, more thoughts

From: "David J Weber" 
Subject: Chinese boxes, more thoughts

Text from file 'DIAG.GUD'

\documentstyle{article}
\setlength{\topmargin}{-1.0in}
\setlength{\textheight}{10.3in}
\setlength{\textwidth}{7.0in}
\setlength{\oddsidemargin}{-.5in}
\setlength{\evensidemargin}{-.5in}
% \setlength{\headheight}{0.0in}
% \setlength{\headsep}{0.0in}
% \setlength{\footheight}{0.0in}
% \setlength{\footskip}{0.0in}
% \renewcommand{\baselinestretch}{.7}
% \newcommand{\as}{}%{\hspace{.1in}\vspace{.1in}}
% \pagestyle{empty}
% \raggedbottom
% \sloppy

\begin{document}
\fbox{\parbox{3.5in}{
      \begin{tabbing}
      chaynaw \`NX\\
      awnisha \`*V0\\
      \fbox{\parbox{3in}{
            \begin{tabbing}
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  imata\`OD\\
                  marka kumun runakuna\`SU\\
                  munaptin\`*R0
                  \end{tabbing}
                 }}\`CND\\
            yanapa:nanpaq\`*S0
            \end{tabbing}
           }}\`PRP\\
       \fbox{\parbox{3in}{
            \begin{tabbing}
            llapan awturida:kuna\`SU\\
            Pillkuchaw\`LOC\\
            kaqkunaqa\`*S0
            \end{tabbing}
           }}\`SU\\
      \fbox{\parbox{3in}{
            \begin{tabbing}
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        markachaw tiyaq runakuna\`*S0
%                        \fbox{\parbox{1.5in}{
%                              \begin{tabbing}
%                              markachaw\`LOC\\
%                              tiyaq\`*S0
%                              \end{tabbing}
%                             }}\`MOD\\
%                        runakuna\`*S0
                        \end{tabbing}
                       }}\`SU\\
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        kikin markachaw\`LOC\\
                        wasita\`OD\\
                        rurananpaq\`*S0
                        \end{tabbing}
                       }}\`PRP\\
                  ristamuta\`OD\\
                  ma\~{n}aptin\`*R0
                  \end{tabbing}
                 }}\`CND\\
            \fbox{\parbox{2.5in}{
                  \begin{tabbing}
                  huk inhinyeru\`SU\\
                  \fbox{\parbox{2in}{
                        \begin{tabbing}
                        markaman\`META\\
                        aywaykur\`*R0
                        \end{tabbing}
                       }}\`CND\\
                  rikaptinraq\`*R0
                  \end{tabbing}
                 }}\`CND\\
            qonanpaq\`*S0
            \end{tabbing}
            }}\`PRP
      \end{tabbing}
      }}
\end{document}

10th May 2000: Line numbers in plain TeX

From: BP Jonsson 
Subject: line numbers in plain TeX

Well the subject line says it all: I need numbered lines in a document made 
in plain TeX.  Can anybody recommend anything?

TIA,

/BP

16th May 2000: Tables in landscape format

From: Malcolm Ross 
Subject: Tables in landscape format

I need to produce quite large tables in association with my work in 
comparative linguistics, often in landscape format, i.e. rotated 
ninety degrees. I can do this using the package _lscape_ and the 
associated environment _landscape_. But this destroys the ability of 
the table to float, leaving white space on the previous page at the 
point in the file where the table occurs.

Does anyone know of a package that will allow floating objects in 
landscape format? Or a work-around that will achieve this? My 
attempts to date have failed.

Malcolm Ross

________________________________________________________________________
(Dr) Malcolm Ross
Senior Fellow
Department of Linguistics
Research School of Pacific
      and Asian Studies
Australian National University
CANBERRA A.C.T.
Australia 0200

Phone: +61 2 6249 3147
Fax:   +61 2 6249 4896

http://linguistics.anu.edu.au/biomdr.html

16th May 2000: Re: Tables in landscape format

From: Peter Kleiweg 
Subject: Re: Tables in landscape format

Malcolm Ross skriver...

> Does anyone know of a package that will allow floating objects in 
> landscape format? Or a work-around that will achieve this? My 
> attempts to date have failed.

To rotate all, figure and caption, use:

  \usepackage{rotating}

  \begin{sidewaysfigure}
  ...
  \end{sidewaysfigure}

  \begin{sidewaystable}
  ...
  \end{sidewaystable}

If you just want to rotate the table, but not the caption, use:

  \usepackage{graphicx}

  \begin{table}
  \centerline{\rotatebox{90}{\begin{tabular}...
  ...
  \end{tabular}}}
  \caption{...}
  \end{table}

-- 
Peter Kleiweg
http://www.let.rug.nl/~kleiweg/

18th May 2000: Issues of Russian transliteration

From: BP Jonsson 
Subject: Issues of Russian transliteration

Issues of Russian transliteration:

A client wants me to typeset russian words in latin transliteration with 
accent markings, and she has a couple of very firm preferences for what it 
should look like.  The problem relates to the letter "e oborotnoe", which 
she wants typeset as schwa -- i.e. upside down latin e --, but also as 
upside-down E, rather than enlarged schwa, when capitalized.  Fore some 
reason ;) she finds \'\.e awful, while \'\"e is OK for accented /jo/!  I 
suggested using \`e for unaccented e oborotnoe, doubled \` (however that 
may be symbolized in generic TeX! :-) for accented e oborotnoe and \H for 
accented \"e, but she will have none of that.  Neither, for some reason, 
does she want cyrillic, in spite of my pointing out that it is easy to 
get.  So now I wonder (1) how to get that upper-case rotated E, and (2) if 
there is any precedence for using schwa for e oborotnoe?  I can see the 
logic of it, but the capitalization business makes it hard.

Regards,
/BP

18th May 2000: Re: Issues of Russian transliteration

From: doug@essex.ac.uk (Arnold D J)
Subject: Re: Issues of Russian transliteration

A probably useless suggestion is the observation that upside down "E"
is the logical symbol \exists, so something like $\mathrm{\exists}$
might be a starting point....

Best
Doug

18th May 2000: Re: Issues of Russian transliteration

From: Barbara Beeton 
Subject: Re: Issues of Russian transliteration

from philip jonsson:

    Issues of Russian transliteration:

    A client wants me to typeset russian words in latin transliteration with 
    accent markings, and she has a couple of very firm preferences for what it 
    should look like.  The problem relates to the letter "e oborotnoe", which 
    she wants typeset as schwa -- i.e. upside down latin e --, but also as 
    upside-down E, rather than enlarged schwa, when capitalized.  ...

shudder!

    ...  Fore some 
    reason ;) she finds \'\.e awful, while \'\"e is OK for accented /jo/!  I 
    suggested using \`e for unaccented e oborotnoe, doubled \` (however that 
    may be symbolized in generic TeX! :-) for accented e oborotnoe and \H for 
    accented \"e, but she will have none of that.  Neither, for some reason, 
    does she want cyrillic, in spite of my pointing out that it is easy to 
    get.  So now I wonder (1) how to get that upper-case rotated E, and (2) if 
    there is any precedence for using schwa for e oborotnoe?  I can see the 
    logic of it, but the capitalization business makes it hard.

*if*
 - you're using latex
 - your output processing includes postscript
there is a package rotating.sty that will allow you to rotate even
single letters; if you can't find a real schwa, it could be concocted
that way.  in the graphics/x package there is a command \reflectbox;
that would produce a reflected cap E, which seems to me slightly
more attractive than a full rotation (although maybe not acceptable
to your client; on the other hand, if it *is* acceptable, maybe
she'd also accept a reflected lowercase e instead of a schwa).

i know of no precedent for using schwa for e oborotnoe -- phonetically
it's pure nonsense.

i think your client has rocks in her head, but "the client is always
right" (ha!), and the request *is* possible of fulfilment.

good luck.
							-- bb

18th May 2000: Re: Issues of Russian transliteration

From: Joost Kremers 
Subject: Re: Issues of Russian transliteration

At 12:12 18-5-00 +0000, you wrote:
>A probably useless suggestion is the observation that upside down "E"
>is the logical symbol \exists, so something like $\mathrm{\exists}$
>might be a starting point....

a suggestion i had already done, but accidentally just sent to the original
poster:

******

a very simplistic way of getting the upper-case rotated E is to use \exists
in mathmode (so $\exists$): this is the logical existential quantifier.
it's not very pretty, though, because the E you get is in (sort of) sans
serif font, whereas your text is presumably in CM or Times.

i guess if you know your way around fonts it should be possible to define a
command that produces an inverted E, but i have no idea how that should be
done.

******

note that AFAIK you don't need the \mathrm command: \exists is already
upright, and there is just no Times Roman version available...

joost kremers

--------------------------------------------------------
Joost Kremers (Mr.)
University of Nijmegen - The Netherlands
Department of Languages and Cultures of the Middle-East
PO Box 9103
6500 HD Nijmegen - The Netherlands
phone: +31 24 3612996
fax: +31 24 3611972

19th May 2000: Re: Issues of Russian transliteration

From: BP Jonsson 
Subject: Re: Issues of Russian transliteration

At 08:39 18.5.2000 -0400, Barbara Beeton wrote:
>i think your client has rocks in her head, .

I just spoke to one of the guys she shares a supervisor with, and he would 
agree with you 111%! :-)

>but "the client is always
>right" (ha!), and the request *is* possible of fulfilment

Well, I asked that colleague of hers to "bribe" the supervisor to tell her
that if it gets out that she used TeX, and didn't use cyrillic, everybody's
gonna think she's got rocks in her head! ;-)  I'm afraid it'll mean I'll 
have to fix *his* thesis at a reduced rate, but WTH, he gave me too a sip 
off that bottle of hotelny vodka! :-)

>good luck.

Thanks, I'll need it, it seems...

/BP

  B.Philip Jonsson  mailto:melroch@mail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~__
             A h-ammen pennuid i phith!                \ \
    __  ____ ____    _____________ ____ __   __ __     / /
    \ \/___ \\__ \  /___  _____/\ \\__ \\ \  \ \\ \   / /
    / /   / /  /  \    / /Melroch\ \_/ // /  / // /  / /
   / /___/ /_ / /\ \  / /Melarocco\_  // /__/ // /__/ /
  /_________//_/  \_\/ /Eowine__   / / \___/\_\\___/\_\
Gwaedhvenn Angelmiel\ \_____/ / a/ /_adar Merthol naun
~~~~~~~~~Cuinondil~~~\_______/~~~\__/~~~Noolendur~~~~~~
|| Lenda lenda pellalenda pellatellenda cuivie aiya! ||
"A coincidence, as we say in Middle-Earth" (JRR Tolkien)

19th May 2000: Re: Issues of Russian transliteration

From: Alejandro Lopez-Valencia 
Subject: Re: Issues of Russian transliteration

At 12:39 PM +0200 19/5/0, BP Jonsson wrote:
>gonna think she's got rocks in her head! ;-)  I'm afraid it'll mean I'll
>have to fix *his* thesis at a reduced rate, but WTH, he gave me too a sip
>off that bottle of hotelny vodka! :-)
>

Ahhh! But nothing like Finish vodka ;) And what about Bombay Saphire with
English tonic? Rats! If I only could find a bottle of Legoine. ;)

Anyway, I just looked and found I have some aboobie cyr fonts (Times Ten
and Baskerville). Interested?

--
P. Alejandro Lpez-Valencia      Ecologist, Conservation Biologist

23rd May 2000: Multilingual scanning

From: Claire Bowern 
Subject: Multilingual Scanning

Hi all, 

This isn't about TeX but it is about text processing and scanning and
since in many dialects of English final t is unreleased I hope you don't
mind a question off the subject...

I've been asked to scan and do OCR on a whole load of Zuni/English texts.
The originals are typed and are very clear, but the Zuni has a lot of
diacritics (umlauts, glottalisation and so on) and some non-standard
characters. I've been mucking aroudn with the OCR program but I can't get
anything sensible out of the Zuni (I've tried all the language settings to
try to get it to "see" diacritics (by pretending it's Finnish or German
and so on) but nothing's come up well)

So, my question is a) whether it's worth trying to muck around any more or
whether it would be quicker to type in the 150 pages :-) and b) if there's
anything I can do to improve the character recognition (eg, setting to
look for individual characters rather than whole words)?

Thanks!
Claire

____________
Claire Bowern
Department of Linguistics
Harvard University
305 Boylston Hall
Cambridge  MA  02138
Ph: (617)-493-4230

24th May 2000: Re: Multilingual scanning

From: Emmon Bach 
Subject: Re: Multilingual Scanning

About scanning nonstandard characters:

I have only used a couple of scanners for this problem, Textbridge and
Omnipage (8.0 was the last one I used).  Of these Omnipage is the only one
that allows you to specify arbitrary TrueType fonts for recognition.  I
have used this with some success on texts from Awik'ala (an Upper Northern
Wakashan language from Rivers Inlet BC).  The orthography for this language
(from John Rath) has extensive use of diacritics: haceks over x, g, a,
grave accents over a bunch of vowels including schwa. My fonts were made by
me with Fontographer. The scans required a lot of editing, but I think on
balance it was worth it.  A drawback is that you can only see the
characters in some of the editing modes, as I recall.  Textbridge I have
used with success on Western Abenaki, which has only a couple of special
glyphs that can be found in your usual ASCII fonts (o's with circumflex).
You can't tell it to use any old font for a base, and the insertion of
special characters is a pain, requiring pulling down a list of special
characters, with no encoding known to me. 

I would be delighted to hear about more possibilities.

Incidentally, does anybody know about scanning possibilities with Linux?

Emmon Bach

# 14
Kitimat Motel
656 Dadook Crescent
Box 477
Kitimat BC V8C 2R7
CANADA                    PHONE: 250 632 6677 #14

24th May 2000: `eng' in TIPA

From: Malcolm Ross 
Subject: eng in TIPA

I recently had cause to re-read the discussion started by Martin 
Haase in April 1999 about the seeming absence of the character eng 
(velar nasal) from TIPA.

It may be relevant to no one, but I recently encountered the same 
problem (eng printed as a black square). Before trying the 
suggestions from that discussion, I thought I would try the 
alternative inputs, i.e. \ng and \textipa{N}. To my surprise, only 
\ng produces the black square, whilst \textipa{N} produces a 
perfectly good eng.

I don't know enough about TeX and LaTeX to know what this might mean 
(and I don't need to know!), but I thought I would offer the 
information anyway.

I am using OzTeX 4 on a Mac.

Malcolm Ross

________________________________________________________________________
(Dr) Malcolm Ross
Senior Fellow
Department of Linguistics
Research School of Pacific
      and Asian Studies
Australian National University
CANBERRA A.C.T.
Australia 0200

Phone: +61 2 6249 3147
Fax:   +61 2 6249 4896

http://linguistics.anu.edu.au/biomdr.html

24th May 2000: Re: `eng' in TIPA

From: Avery Andrews 
Subject: Re: eng in TIPA

I get \ng to work fine if I use the T1 encoding
(\usepackage[T1]{tipa}

or T3:  \usepackage[t3]{fontenc}

This is empirical, I don't really know what's going on with encodings.

24th May 2000: Re: `eng' in TIPA

From: Dag Langmyhr 
Subject: Re: eng in TIPA

-> I recently had cause to re-read the discussion started by Martin 
-> Haase in April 1999 about the seeming absence of the character eng 
-> (velar nasal) from TIPA.
-> 
-> It may be relevant to no one, but I recently encountered the same 
-> problem (eng printed as a black square). Before trying the 
-> suggestions from that discussion, I thought I would try the 
-> alternative inputs, i.e. \ng and \textipa{N}. To my surprise, only 
-> \ng produces the black square, whilst \textipa{N} produces a 
-> perfectly good eng.
-> 
-> I don't know enough about TeX and LaTeX to know what this might mean 
-> (and I don't need to know!), but I thought I would offer the 
-> information anyway.
-> 
-> Malcolm Ross

The \ng and \NG glyphs in LaTeX produce the Sami character `eng' which
can be found in the T1 encoding but not in the older T0; consequently
you need

	\usepackage[T1]{fontenc}

However, very few fonts contain these glyphs; in fact, I don't know of
any apart from EC (Extended Computer Modern). The black square
indicates that you are using a font without these glyphs.

						Dag L

24th May 2000: Re: `eng' in TIPA

From: BP Jonsson 
Subject: Re: eng in TIPA

At 15:20 24.5.2000 +1000, Malcolm Ross wrote:
>It may be relevant to no one, but I recently encountered the same problem 
>(eng printed as a black square). Before trying the suggestions from that 
>discussion, I thought I would try the alternative inputs, i.e. \ng and 
>\textipa{N}. To my surprise, only \ng produces the black square, whilst 
>\textipa{N} produces a perfectly good eng.

Try this:
         \font\mypa=tipa12\def\ng{{\mypa\char78}}
or if you like \def\ng{\textipa{N}} probably works dandy in LaTeX -- I'm a 
purist myself :-)

/BP

  B.Philip Jonsson  mailto:melroch@mail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~__
             A h-ammen pennuid i phith!                \ \
    __  ____ ____    _____________ ____ __   __ __     / /
    \ \/___ \\__ \  /___  _____/\ \\__ \\ \  \ \\ \   / /
    / /   / /  /  \    / /Melroch\ \_/ // /  / // /  / /
   / /___/ /_ / /\ \  / /Melarocco\_  // /__/ // /__/ /
  /_________//_/  \_\/ /Eowine__   / / \___/\_\\___/\_\
Gwaedhvenn Angelmiel\ \_____/ / a/ /_adar Merthol naun
~~~~~~~~~Cuinondil~~~\_______/~~~\__/~~~Noolendur~~~~~~
|| Lenda lenda pellalenda pellatellenda cuivie aiya! ||
"A coincidence, as we say in Middle-Earth" (JRR Tolkien)

25th May 2000: Re: Multilingual scanning

From: "Kolb, Hap" 
Subject: RE: Multilingual Scanning

Hi,

why not try a trainable OCR-package like FineReader
(http://www.abbyyusa.com/)? With a little bit of luck (and some manual
effort) it'll work for pages 11 through 150...

Greetings, ...hap

--
hans-peter kolb                          email: 	kolb@tpd.tno.nl
TNO-TPD/TU Delft, DIS/MMT                  tel: 	+31-15-269 2456
Stieltjesweg 1, P.O.Box 155                fax: 	+31-15-269 2111
NL-2600 AD Delft			 	    mobile: 	+31-6-2280
1132

25th May 2000: Re: Multilingual scanning

From: Claire Bowern 
Subject: RE: Multilingual Scanning

THanks to everyone who sent me replies about multilingual scanning.
Omnipage and FineReader were suggested as trainable OCRs. Another
suggestion was an older version of Textbridge (my version has very limited
training capabilities).

I'll let the list know the results of what works...

Claire
____________
Claire Bowern
Department of Linguistics
Harvard University
305 Boylston Hall
Cambridge  MA  02138
Ph: (617)-493-4230

1st June 2000: Endnotes

From: "Harold F. Schiffman" 
Subject: endnotes

I am hoping someone can tell me how to handle this problem:  when I use
endnotes.sty, LaTeX begins the first endnote right up smack-dab underneath
the word NOTES which is in about 16 pt font.  There's absolutely no space
between them, and it looks really stupid.  I have tried to force spaces, I
have tried inserting vspace commands, \bigskip, what-not, it ignores this
and continues to print this in what is clearly a violation of general TeX
and LaTeX principles. 

Can somebody tell me how to edit the endnotes.sty file to make it put some
space in there?  Thanks,

Hal Schiffman

1st June 2000: Re: Endnotes

From: Christina Thiele 
Subject: Re: endnotes

Harold F. Schiffman writes:
> I am hoping someone can tell me how to handle this problem:  when I use
> endnotes.sty, LaTeX begins the first endnote right up smack-dab underneath
> the word NOTES which is in about 16 pt font.  There's absolutely no space
> between them, and it looks really stupid.  I have tried to force spaces, I
> have tried inserting vspace commands, \bigskip, what-not, it ignores this
> and continues to print this in what is clearly a violation of general TeX
> and LaTeX principles. 
> 
> Can somebody tell me how to edit the endnotes.sty file to make it put some
> space in there?  Thanks,
> 
> Hal Schiffman

Hello, Harold.

I use endnotes.sty for a few of my journals. Here's how I make \NOTES
give more space. 

I actually define another macro that's used in the source file -
either \Note or \Notes, which then allows me to add space around
\NOTES and changes in font size to the whole endnotes area. 

While it's not the most elegant solution it certainly does work for
me. 

Ch.

=====

\newcommand{\Note}{{\vspace{6truept}
                     \begin{center}
                        \footnotesize NOTE
                     \end{center}
                     \NOTES                %% see Endnotes section
                     \vspace{-3truept}
                     \enotesize}}          %% see Endnotes section

\newcommand{\Notes}{{\vspace{6truept}
                     \begin{center}
                        \footnotesize NOTES
                     \end{center}
                     \NOTES                %% see Endnotes section
                     \vspace{-3truept}
                     \enotesize}}          %% see Endnotes section
... 

and then further down, in the `Endnotes section', I have these two
def's:

%% Added explicit baselineskip command, which apparently has to have
%%   that blank space after it, in order to be properly applied.
\def\enotesize{\small\baselineskip=13pt }

%% Modified \enoteformat to suit ESC style:
\def\enoteformat{\rightskip\z@ \leftskip=11.5pt%
                 \noindent%
                 \leavevmode\llap{\nn{\ttr\@theenmark\en}}}

\newcommand{\NOTES}{\begingroup 
                    \theendnotes
                    \endgroup\par
                   \par}

I've got a few other bits and pieces in here ... :

%% Where an endnote has an indented paragraph, use this:
\newcommand{\enoteindent}{\mbox{\hspace{1em}}}

%% An un-numbered endnote for thanks, acknowledgements, etc.:
\def\acknotetext{\@ifnextchar [{\@xendnotenext}%
   {\begingroup\let\protect\noexpand%
      \xdef\@theenmark{\null}\endgroup%
    \@endnotetext}}
 
\def\acknote#1{{\acknotetext{\baselineskip=11.5pt \leftskip=0pt%
                                 {\footnotesize\noindent #1}%
                              \vskip3truept}%
          }}

That \acknote ends up printing as the first `note' under the `Notes'
section at the end of the text (or wherever you'd got the endnotes
printing). I haven't used it in so long that I can't swear to that,
though ;-)

And one more -- big -- proviso: I'm using the above code under
2.09. And I know that 2e is sometimes less forgiving than 2.09 for
sloppy def's so ... well ... you know ... take all the above with a
grain or two of salt ... 

Ch.

2nd June 2000: Re: Endnotes

From: Klaus Lagally 
Subject: Re: endnotes

Just a hint, I did not look into the problem itself:

if you write your own macros containing dimensions, and also numbers, 
leave one explicit space after the number and/os the dimension, e.g.:
24pt or 3.24! This will cause TeX to stop the scanning
of the external representation of the number/dimension, and will eat
all following spaces without harm. Otherwise TeX might be going on with
expanding the input looking for the end of the quantity, and even though
it will back out later, once it has resolved the situation, the spurious
expansions might have led to side effects which will not be undone.
I fell into this trap several times.

Another safe solution is putting \relax after the quantity to stop TeX
reading on.

Cheers 

Klaus
-- 
Prof. Dr. Klaus Lagally  | lagally@informatik.uni-stuttgart.de
Institut fuer Informatik | Tel.  +49-711-7816392 |  Zeige mir deine Uhr,
Breitwiesenstrasse 20-22 | FAX   +49-711-7816370 |  und ich sage dir,
70565 Stuttgart, GERMANY |             (changed) |  wie spaet es ist.

5th June 2000: Re: Endnotes

From: Anshuman Pandey 
Subject: Re: endnotes

On Thu, 1 Jun 2000, Christina Thiele wrote:
> Harold F. Schiffman writes:

[snip]

> I use endnotes.sty for a few of my journals. Here's how I make \NOTES
> give more space. 
> 
> I actually define another macro that's used in the source file -
> either \Note or \Notes, which then allows me to add space around
> \NOTES and changes in font size to the whole endnotes area. 
> 
> While it's not the most elegant solution it certainly does work for
> me. 

It surprised me that this is still not part of any LaTeX class file. I
would assume that enough people use endnotes.sty -- or local variants of
it -- which would justify its inclusion in the "tools" package or its
addition to the core classes. Wouldn't this be ideal? 

The implementation would roughly follow the current practice: \footnote
contains the note and its placement is dictated through the use of a macro
\endnotes, or \endnotes=1. If the \endnotes macro is absent LaTeX typesets
the footnotes in the standard manner, if \endnotes exists or is true, then
they are placed at the end of the document. Perhaps something like
\makeendnotes would indicate where you'd want to place them. Part of this
automation would include attributing the standard LaTeX sectioning
parameters to set the appropriate vspace between "Notes" and the actual
enumeration of the notes.

Regards,
Anshuman Pandey

5th June 2000: Re: Endnotes

From: Mike Hammond 
Subject: endnotes

all

i hope this helps. i had a similar problem with endnotes and fixed it. i 
can't remember what i did, but i'm attaching the edited enotes.sty file, 
which i renamed as mhenotes.sty.

i invoke it as follows.

\newpage
\section*{Notes}

\indent\theendnotes

\newpage

mike hammond

5th June 2000: Re: Endnotes

At 05:31 AM 6/5/00, Anshuman Pandey wrote:
>It surprised me that this is still not part of any LaTeX class file. I
>would assume that enough people use endnotes.sty -- or local variants of
>it -- which would justify its inclusion in the "tools" package or its
>addition to the core classes. Wouldn't this be ideal?

It would be more ideal if people simply stopped producing endnotes. As far 
as I can tell, endnotes were invented due to the difficulty of making 
footnotes; surely anyone who's ever had to produce a paper with a 
typewriter would prefer endnotes to footnotes. But these days (and 
especially when programs like TeX are available), there's no excuse for 
using endnotes.  Perhaps this has changed from the days before 
parenthetical in-text citation, but I find that the notes in the papers I 
read are generally worth reading.  It's a pain to keep the notes at the end 
of the chapter (or even worse, the book) bookmarked so you can read them 
simultaneously with the main text.

I realize that many people are writing for draconian publishing houses or 
journals or to thesis guidelines that haven't been updated since 1970, and 
for these cases, the endnotes package gets a lot of use. But the people who 
control the LaTeX distribution have a prescriptivist bent to them, and I'd 
bet that they're hoping endnotes will just vanish completely.

--
Michael W. Daniels          | "But the evil is that they hold for
daniels@ling.ohio-state.edu |    certain that they are in the light."
Dept. of Linguistics        |

5th June 2000: Re: Endnotes

From: Emma Pease 
Subject: Re: endnotes 

> It would be more ideal if people simply stopped producing endnotes. As far 
> as I can tell, endnotes were invented due to the difficulty of making 
> footnotes; surely anyone who's ever had to produce a paper with a 
> typewriter would prefer endnotes to footnotes. But these days (and 
> especially when programs like TeX are available), there's no excuse for 
> using endnotes.  Perhaps this has changed from the days before 
> parenthetical in-text citation, but I find that the notes in the papers I 
> read are generally worth reading. 

I tend to agree except for two reasons 

1. The latex file is not going to be used directly and the output is
not meant to be camera ready so the notes and the text are going to be
re-typeset

2. Long notes.  If the notes are often a significant part of the page,
then endnotes make sense.  I've worked on one book[1] where the notes
often took up two-thirds or more of the page (looking back, I wish we
had used endnotes); at that point getting good pagebreaks is a
problem.

Emma

[1] A philosophy book not a linguistics book.

5th June 2000: Re: Endnotes

From: Michael Doob 
Subject: Re: endnotes

Michael W. Daniels writes:
> It would be more ideal if people simply stopped producing endnotes. As far 
> as I can tell, endnotes were invented due to the difficulty of making 
> footnotes; surely anyone who's ever had to produce a paper with a 
> typewriter would prefer endnotes to footnotes. But these days (and 
> especially when programs like TeX are available), there's no excuse for 
> using endnotes.  Perhaps this has changed from the days before 
> parenthetical in-text citation, but I find that the notes in the papers I 
> read are generally worth reading.  It's a pain to keep the notes at the end 
> of the chapter (or even worse, the book) bookmarked so you can read them 
> simultaneously with the main text.

I think that anyone who has had to typeset books (for draconian
publishers or otherwise) that have lots of footnotes and displayed
mathematics whould disagree with you. Once you have mulitple-page
footnotes, they conflict with floats and multiline equations resulting
in very underfull vboxes being produced by the output routine.

Cheers,
Michael

5th June 2000: Re: Endnotes

From: "Michael W. Daniels" 
Subject: Re: endnotes

At 01:43 PM 6/5/00, you wrote:
>mathematics whould disagree with you. Once you have mulitple-page
>footnotes, they conflict with floats and multiline equations resulting
>in very underfull vboxes being produced by the output routine.

I have to admit that I've never seen a book with multiple-page footnotes; I 
can imagine that such a thing would cause problems. I'd personally turn it 
into an appendix, or re-work it into the main text; I can appreciate that 
the freedom to do so isn't always there.

--
Michael W. Daniels          | "But the evil is that they hold for
daniels@ling.ohio-state.edu |    certain that they are in the light."
Dept. of Linguistics        |

6th June 2000: LaTeX for linguists --- a web site

From: doug@essex.ac.uk (Arnold D J)
Subject: LaTeX for linguists --- a web site

                     LaTeX for linguists

Perhaps some of you already know this site, it has been around,
incomplete and unpublicised for years, but it has now reached a
certain level of `maturity' (i.e. its been around for a while) ...
Anyway, its high time we `officially announced it' 

	  http://clwww.essex.ac.uk/latex4ling/

I quote:

     These pages provide information on how to use LaTeX for writing
     Linguistics papers (articles, books, etc.). In particular, they
     provide instructions and advice on creating the things Linguists
     standardly need, like trees, numbered examples, and so on, as
     well as advice on some things that most people need (like
     bibliographies), but with an eye to standard Linguistic practice.

     Our idea is that you can have your browser pointing at these
     pages all the time you are writing; whenever you need one of the
     LaTeX constructs described, you can just point, click, drag and
     drop (etc.) one of the examples we give into your text and then
     change it to suit.

     The pages are: 
     + not an introduction to LaTeX; 
     + not complete or finished -- there are lots of gaps, 
     + not comprehensive -- even when finished, there will probably be topics that do
       not get covered (but see How you can contribute) 
     + not original (most of the packages we describe existed before these pages
       were created) 
     + not closed or self-contained (see `How you can contribute', below). 

Still, you may find it useful as a point of reference.

Doug Arnold
------------------------------------------------------------------
Doug Arnold,  		     	   doug@essex.ac.uk
Dept. of Language & Linguistics,   http://clwww.essex.ac.uk/~doug
University of Essex,	     	   
Wivenhoe Park,		     	   Tel: +44 1206 872084 (direct)
Colchester, CO4 3SQ, UK.     	   Fax: +44 1206 872198
------------------------------------------------------------------

18th June 2000: Tables

From: "Benjamin T. Bruening" 
Subject: tables

Dear ling-tex,

I'd be very happy to never use Microsoft Word again, but I find 
working with tables on my computer screen to be very useful for 
organizing and analyzing data.  LaTeX prints tables fine, but it's 
very hard to work with them on your computer screen.  So I've been 
doing tables in MS Word, which is far from optimal (tables don't even 
display properly in Word 98 and sometimes crash the program). I 
haven't tried to insert Word tables into a LaTeX document; I imagine 
it could be done but only after multiple conversions.

Can anyone recommend a way of working with tables that doesn't 
require a bloated program like MS Word?  Ideally any tables created 
could be put into a LaTeX document with minimal hassle.  (I'm using 
OzTeX on a G3 Mac.)

Thanks,
Benjamin
-- 

-------------------------------------------
Benjamin Bruening
Dept. of Linguistics
MIT E39-245
77 Massachusetts Ave.
Cambridge, MA 02139

18th June 2000: Re: Tables

From: Christina Thiele 
Subject: Re: tables

Benjamin T. Bruening writes:
> Dear ling-tex,
> 
> I'd be very happy to never use Microsoft Word again, but I find 
> working with tables on my computer screen to be very useful for 
> organizing and analyzing data.  LaTeX prints tables fine, but it's 
> very hard to work with them on your computer screen.  So I've been 

Clarify `hard' ;-) 

That is, is it that LaTeX code is hard (perhaps getting in the way?)
or that the text editor is getting in the way.

Remember: inputting text has far more to do with how useful your
editor is -- or rather, how much you know about what your editor can
really do AND whether you have the time to customise it to do the
grunt work for you ;-)

Perhaps if you explain what's hard that will also make it easier to
recommend alternatives, be they a better editor for TeX or just a
different program entirely ... 

I've done masses of tables and I don't find inputting them
particularly hard ... but maybe inputting is not where the problem
lies?

> ...
> Can anyone recommend a way of working with tables that doesn't 
> require a bloated program like MS Word?  Ideally any tables created 
> could be put into a LaTeX document with minimal hassle.  (I'm using 
> OzTeX on a G3 Mac.)
> 
> Thanks,
> Benjamin
> ...

Ch.

18th June 2000: Re: Tables

From: "Benjamin T. Bruening" 
Subject: Re: tables

>Clarify `hard' ;-)
>
>That is, is it that LaTeX code is hard (perhaps getting in the way?)
>or that the text editor is getting in the way.

>Perhaps if you explain what's hard that will also make it easier to
>recommend alternatives, be they a better editor for TeX or just a
>different program entirely ...
>
>I've done masses of tables and I don't find inputting them
>particularly hard ... but maybe inputting is not where the problem
>lies?

No, inputting is fairly easy.  It's manipulating tables afterwards. 
Say I want to insert an entire column in the middle of a table, or 
cut out a column and insert it in the middle of another table.  These 
aren't one-click operations in the text editors I'm familiar with, 
since you can't select all and only a column.  The code also does get 
in the way if one is attempting to analyze data in tabular format, 
especially if one is using IPA fonts or other symbols and wants to 
see at a glance how certain things compare with each other.  One 
might also not know the best way to lay out a table until one has 
tried several alternatives, but having to run LaTeX for each 
alteration can be a bit cumbersome.  In other words, I'd like some 
kind of WYSIWYG table program whose output could be inputted into 
LaTeX easily, like the ones that exist for pictures and figures.  My 
point in asking is the hope that someone else already did the grunt 
work.

Benjamin
-- 
-------------------------------------------
Benjamin Bruening
Dept. of Linguistics
MIT E39-245
77 Massachusetts Ave.
Cambridge, MA 02139

19th June 2000: Re: Tables

From: Joseph Hilferty 
Subject: Re: tables

"Benjamin T. Bruening" wrote:
> So I've been
> doing tables in MS Word, which is far from optimal (tables don't even
> display properly in Word 98 and sometimes crash the program). I
> haven't tried to insert Word tables into a LaTeX document; I imagine
> it could be done but only after multiple conversions.

Programs such as RTF2LaTeX should be able to do this. There's a free
program called Excel2LaTeX that might be able to help you out too (I
guess--I use WordPerfect myself).

http://www.jam-software.com/software.html

N.B.: These are Windows programs.

> Can anyone recommend a way of working with tables that doesn't
> require a bloated program like MS Word?  Ideally any tables created
> could be put into a LaTeX document with minimal hassle.  (I'm using
> OzTeX on a G3 Mac.)

Though you'd have to run it under your Windows emulator, you might
try Pedro Luis Luque's Tablas program:

http://www.fie.us.es/~calvo/latex.html

Joe
__________________________________________________________
Home page: http://lingua.fil.ub.es/~hilferty/homepage.html

19th June 2000: Re: Tables

From: Alexis Dimitriadis 
Subject: Re: tables

> In other words, I'd like some kind of WYSIWYG table program whose output
> could be inputted into LaTeX easily, like the ones that exist for pictures
> and figures.

Two suggestions:

a. There are some near-WYSIWYG wrappers for LaTeX, perhaps you could just use
one of those? The only one I have used ran on unix (LyX) and was not quite
ready for prime time, but something like that might make it easier to fiddle
with tables.

b. Alternately: have you tried using Excel to build your tables?  They can be
saved in Text Only mode with the fields separated by tabs; you could then
substitute &'s and \\s manually or via some editor macro, but it might be
simpler to just put the &s and \\s right in the Excel fields before you save.
If you do the latter it should be possible (unless my recollection is VERY
mistaken) to \input the saved file directly into your latex document without
further processing.  It's not ideal, since you cannot see what the IPA
characters look like, for example, but it might do if no specialized
application exists; and you can't beat Excel for row and column manipulations!

Alexis Dimitriadis

19th June 2000: Re: Tables

From: Martin Steer 
Subject: Re: tables

As usual, Chistina is right on the ball.

Christina wrote:
>>That is, is it that LaTeX code is hard (perhaps getting in the way?)
>>or that the text editor is getting in the way.
>>
>>Perhaps if you explain what's hard that will also make it easier to
>>recommend alternatives, be they a better editor for TeX or just a
>>different program entirely ...
>>
>>I've done masses of tables and I don't find inputting them
>>particularly hard ... but maybe inputting is not where the problem
>>lies?

I use Alpha as a text editor with OzTeX. Both of these are Mac programs.
Alpha has a large number of ready-made LateX macros and is also easy to
customize. It works well with OzTeX. I'd like to know what other linguists
use. I think that once you get used to LaTeX, there are few problems with
using tables, but some of the sorting of data and so on, that you can do in
a Word table, is not really a job for LaTeX. It's a job for... ?

OxTeX is shareware and is available at:
http://www.trevorrow.com/

Alpha is shareware and is available at:
http://alpha.olm.net/

Benjamin wrote:
>No, inputting is fairly easy.  It's manipulating tables afterwards.
>Say I want to insert an entire column in the middle of a table, or
>cut out a column and insert it in the middle of another table.  These
>aren't one-click operations in the text editors I'm familiar with,
>since you can't select all and only a column.

Alpha allows 'rectangular editing', so you can indeed select and move a column.

> The code also does get
>in the way if one is attempting to analyze data in tabular format,
>especially if one is using IPA fonts or other symbols and wants to
>see at a glance how certain things compare with each other.  One
>might also not know the best way to lay out a table until one has
>tried several alternatives, but having to run LaTeX for each
>alteration can be a bit cumbersome.

One of the advantages of Alpha with OzTeX is that it allows you to select a
portion of your text to typeset, so in a long document you can select just
your table, typeset it, and have a look at it.

> In other words, I'd like some
>kind of WYSIWYG table program whose output could be inputted into
>LaTeX easily, like the ones that exist for pictures and figures.  My
>point in asking is the hope that someone else already did the grunt
>work.

This sounds messy, but good luck!

There is a WYSIWYG interface (LyX) available for TeX, but as far as I know
it doesn't run on a Mac. LyX is freeware. If you have money there is also a
TeX system called Textures, which features 'live rendering' of your LaTeX
code. If I may quote from another discussion list (the OzTeX list):

On Mon, May 08, 2000 at 10:29:34AM -0700, C.M. Connelly wrote:
> Textures is, by far, the slickest TeX system I've seen.  Being
> able to type your source code and see your typeset text appear is
> pretty neat.  But Textures requires a *very* fast machine,
> especially if you're going to be doing any work of any significant
> length.

If anyone uses these I'd be interested to hear about it.

+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
Martin Steer
Department of Linguistics
Research School of Pacific and Asian Studies
Australian National University

19th June 2000: Re: Tables

From: Fredrik Karlsson 
Subject: Re:tables

I'd suggest you check out gnumeric at

http://www.gnome.org/projects/gnumeric/

It lets you work with the data much like the way you would in Excel,
and also lets you save the result in LaTeX-format.

The result is table-code that should be easy enough to adapt for your
specific needs.

/Fredrik

-----------------------------------------------------------------
Fredrik Karlsson, Research Assistant

Department of Philosophy and Linguistics
Ume University
S-901 87 UME
SWEDEN
Tel: +46 90 786 69 02 	Web: http://www.ling.umu.se/~fredkarl/
Fax: +46 90 786 63 77	Email: fredkarl@ling.umu.se
-----------------------------------------------------------------

19th June 2000: Re: Tables

From: doug@essex.ac.uk (Arnold D J)
Subject: Re: tables

Try using Lyx -- it is an almost wsywyg interface to LaTeX.

The web site is:

	http://www.lyx.org/

It is really excellent, especially in respect of tables. (contrary to
another correspondent, I think it really is ready for prime time).

Best wishes
Doug Arnold

19th June 2000: Re: Tables

From: Frederik Fouvry 
Subject: Re: tables

,-- On Mon, 19 Jun, Martin Steer wrote:
| 
| Alpha allows 'rectangular editing', so you can indeed select and move a column.

[...]

| One of the advantages of Alpha with OzTeX is that it allows you to select a
| portion of your text to typeset, so in a long document you can select just
| your table, typeset it, and have a look at it.

FYI: Emacs can do rectangular editing as well (including sorting on a
table on the data in any given column), and in the latex-mode, you can
also select a piece of a document to run LaTeX on.

-- 
Frederik Fouvry

SfB 340					fouvry@sfs.nphil.uni-tuebingen.de
Eberhard-Karls-Universitt		http://www.sfs.nphil.uni-tuebingen.de/~fouvry/
Kleine Wilhelmstrasse 113		Tel. +49 7071 29 7 73 13 (direct)
D-72074 Tbingen			Fax  +49 7071 550 520
Deutschland

19th June 2000: Diagrams for Bora derivations

From: "David J Weber" 
Subject: diagrams for Bora derivations

Friends,

I have been using LaTeX as I coauthor a grammar of Bora (a Witotoan
language of Peru and Colombia). Bora has a complicated tone
system which we describe in terms of the cyclic addition of
suffixes, each suffix bringing along tones imposed on its host, and
some floating tones.

We use diagrams like those given below (also attached), embedding
them in verbatim contexts. (If they look like gibberish, try looking
at the attached file with a fixed-width font.)

There are two disadvantages to using the verbatim context: First, the
result is ugly; the diagrams stick out like sore thumbs.  Second, we
must use the practical orthography rather than (T)IPA. Consequently,
we would like a macro package for typesetting such diagrams, one that
would allow us to escape the limitation to fixed-width fonts, and
allow us to use IPA for the Bora.

Is there something we could use? ...adapt? ...anyone who would like to
help develop such a package?

Help! --David

Examples follow:

 hvete   finish
 |  L |   lexical
 |  : +        blocked
 |  : L_tso    causative
 |  : |   |
 |  : |   L_te    go to do
 |  : |   :  +        blocked
 |  : |   :  L_ro     CE
 |  : |   :  |  |
 |  : |   :  |  L_:be    
 |  : |   :  |  :   |
 |  : |   :  |  :   L   FDLT
 H  : H   :  H  :   :   DHT
 :  : :   :  :  : H :   PLTS
 :  : :   :  :  : : :
 hvet-tso-t-ro-be `In vain did he go to make it stop.'

majcho   to eat                                  hvete   to finish
 |   |                                           |  L |   lexical
 |   L_tso    cause                              |  : +        blocked
 |   $   |       delinked                        |  : L_tso    cause
 |   |   L_te    go to do                        |  : |   |
 |   |   |  |                                    |  : |   L_je   return from doing
 |   |   L__._juco:   already                    |  : |   |  |
 |   |   :  |  L |    lexical                    |  : |   L__._juco:   already
 |   |   :  |  $ |        delinked               |  : |   :  :  L |    lexical
 |   |   :  |  | L__ro    CE                     |  : |   :  :  $ |       delinked
 |   |   :  |  | $   |        delinked           |  : |   :  :  | L__ra   CE
 |   |   :  |  | |   L_:be                  |  : |   :  :  | |   |
 |   |   :  |  | |   :   :                       |  : |   :  :  | L___._hi   
 |   |   :  |  | |   :   L    FDLT               |  : |   :  :  | :   |  |
 H   H   :  H  H H   :   :    DHT                |  : |   :  :  | :   |  L   FDLT
 :   :   :  :  : :   : H :    PLTS               H  : H   :  :  H :   H  :   DHT
 :   :   :  :  : :   : : :                       :  : :   :  :  : :   :  :
mjch-tso-t-jc:-ro-be                       hvet-tso-j-jco:-r-hi
`He already went to make them eat                `(Someone) already 
returned from trying to make
 (but without succeeding).'                      him/her/it finish (but without 
succeeding).'

In the following, G represents a floating tone:

di  ahbe      brother
 |\  L  |\     lexical
 | + |  | |    blocked by ahbe's L
 | G |  | |
 |   |  | +           blocked by ahbe's L
 |   H  | G ta:ba     wife (PPHE, GEN)
 |   :  |    |  |     
 |   :  |    |  L_ke     objAn
 |   :  |    |  :  |
 |   :  |    |  :  L     FDLT
 H   :  H    H  :  :     DHT
 :   :  :    :  :  :
d  hb   t:ba-ke `to my brother's wife'

David Weber (david_weber@sil.org)

20th June 2000: Re: Tables

From: Michael Doob 
Subject: Re: tables

> Can anyone recommend a way of working with tables that doesn't 
> require a bloated program like MS Word?  Ideally any tables created 
> could be put into a LaTeX document with minimal hassle.  (I'm using 
> OzTeX on a G3 Mac.)

There have been several replies generated by the above message. 
Perhaps the following TeX code is the type of thing you want:

\documentclass{article}
\begin{document}
 
% We change the  to be a TeX alignment character and
% the  to be \cr
{\catcode`\^^I=4 \catcode`\^^M=\active \let^^M=\cr%
\halign{#&&\quad#
abc	def	ghi
jkl	mno	pqr
stu	vwx	yz$\alpha$
}}%
 
\end{document}

The table data are input by separating columns by a  and putting
each line of the table on a separate line in the TeX file.

Cheers,
Michael

20th June 2000: Yet another tree macro

From: "Kirk Lowery" 
Subject: YATM! :-)

In an odd corner of the web I just found Yet Another Tree Macro -- the 
QobiTree package:

http://www.neci.nj.nec.com/homepages/qobi/software.html

Notes:

1. Written by Jeff Siskind, available for free usage.
2. It only takes one file, and using it seems pretty straight forward.
3. It was written for LaTeX 2.09, so be cautious!
4. The example file that comes with it compiled fine in LaTeX 2.09 
compatibility mode, and the output looks pretty good.

I don't remember having seen mention of this package before, so I hope 
this will be of some interest. Maintainers of tree packages might find 
some ideas in his code...

Kirk

- ----
Kirk Lowery

20th June 2000: Tree macros (was: YATM! :-)

From: Joseph Hilferty 
Subject: Tree macros (was: YATM! :-)

I'm supposed to be teaching an English Syntax course next semester,
and I was wondering if anybody had an opinion on the various tree
macros that exist here and there.

Specifically, what I'm looking for is a tree-macro package that
would compile nicely for both PS and PDF documents (without having
to resort to Distiller in the latter case). Unfortunately, this
rules out most of the tree macros out there.

I've tried Doug Arnold's parsetree package, which is very
straightforward and is the only tree macro I've seen that works
with both with LaTeX and PdfLaTeX. In this sense it's a dream
come true. Unfortunately, parsetree only allows three branches
per node, and the branches don't render smoothly (at least on
my PC).

I've also tried Jeffrey Siskind and Alexis Dimitriadis's qtree. It
doesn't seem to work with PdfLaTeX, but it does work with DviPdfm.
The branches render smoothly in PS, and a little less beautifully
in PDF.

If anyone has any recommendations, I'd be very grateful.

Thanks in advance,
Joe Hilferty

20th June 2000: Re: Tree macros

From: Christina Thiele 
Subject: Re: Tree macros (was: YATM! :-)

I've no idea how it works when trying to change into PDF files but I
find Emma Pease' tree-dvips really good. It's pretty much the first
`tree program' I tried and I've stuck with it ever since.

Avery Andrews has a preprocessor for it, called pstrees. 

Info on both of these and other (oldish but still stylish ;-) )
packages can be found in the file ling-mac.tex on CTAN in

   tex-archive/info/

And every time I mention it, I also go through a little ritual -- a
mea culpa song and dance, actually -- that I promise I'll update the
file but you know how those good intentions seem to wander off to some
other part of the room a few days later ... 

Perhaps someone who's already using tree-dvips has gone the PDF route
and can tell you/us if its results are any better/worse than what
you've already found out. 

Ch.

21st June 2000: Re: Tree macros

From: Emma Pease 
Subject: Re: Tree macros (was: YATM! :-) 

> I've no idea how it works when trying to change into PDF files but I
> find Emma Pease' tree-dvips really good. It's pretty much the first
> `tree program' I tried and I've stuck with it ever since.

I doubt it would work if converting straight from tex to pdf or dvi to
pdf (without going via postscript).  It is dependent on postscript
specials. 

Emma

ps. who one day will update the tree-dvips macros.

21st June 2000: Re: Tree macros

From: Christina Thiele 
Subject: Re: Tree macros (was: YATM! :-)

Ah shucks, Emma ... you're now joining me in the ritual `Some day I'll
update this thing. Soon, maybe. I think.' ;-)) 

Nice to have some company ;-)) 

Ch.

22nd June 2000: Web site information

Hi there, everyone! 

Last week, Doug Arnold announced a pretty nifty website for 
linguistcs using TeX:

     http://clwww.essex.ac.uk/latex4ling/

We've now got this link in the ling-tex homepage under the
heading `Other interesting websites'. 

Now I'd like to invite anyone else who has a relevant 
website they'd like to have linked from/to ling-tex
to post the info here and Dag Langmyhr, our webmaster, will  
see that it's added. 

And if you haven't gone to the ling-tex website, here's
where to go:

    http://www.ifi.uio.no/~dag/ling-tex/

Notice that mail to ling-tex is archived and available
in chunks of a year -- it's a nice way to read related 
messages on a given subject. 

And of course, if you have suggestions for other additions
to the site, please let us know!

Thanks to everyone -- our list has been going since 1994,
and is archived in Norway since 1995. So thanks also go to
Dag for keeping it up!

Ch.

22nd June 2000: Re: Web site information

From: BP Jonsson 
Subject: Re: Web site information

>Hi there, everyone!
>
>Last week, Doug Arnold announced a pretty nifty website for
>linguistcs using TeX:
>
>      http://clwww.essex.ac.uk/latex4ling/
                                 ^^^^^
So we plain-users are out of luck again? ;-)>

Well, seriously, on the site there is mentioned that

>Covington's. which are a LaTeX adaptation of M. deGroot's Midnight gloss 
>TeX-macros.

Where can I find these presumably plain macros?

And BTW, has anyone succeeded typesetting the edmac documentation?

/BP

 B.Philip Jonsson  bpj@netg.se 
melroch@mail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"Doubt grows with knowledge" -Goethe

22nd June 2000: Re: Web site information

From: Barbara Beeton 
Subject: Re: Web site information

bp jonsson asks

    Well, seriously, on the site there is mentioned that

    >Covington's. which are a LaTeX adaptation of M. deGroot's Midnight gloss 
    >TeX-macros.

    Where can I find these presumably plain macros?

ctan: macros/generic/midnight/

    And BTW, has anyone succeeded typesetting the edmac documentation?

this was at one time intended as a tug texniques edition, and i have
the files from that.  i'll have to check with the authors to find out
what is their intention for the file; i can't think that they would
want to limit its availability, but i do have the obligation to check.
i've never tried to generate the documentation from the files in the
zip bundle on ctan, but if the authors say "okay", i'll examine those
to see what the difference is.  perhaps a .ps or .pdf file in the
ctan info area would be a suitable resolution?

of course, if someone else comes forward first and says "here it is",
that would save a lot of work ...
							-- bb

22nd June 2000: Re: Web site information

From: BP Jonsson 
Subject: Re: Web site information

>perhaps a .ps or .pdf file in the
>ctan info area would be a suitable resolution?

OK with me.

/BP

  B.Philip Jonsson  mailto:melroch@mail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~__
             A h-ammen pennuid i phith!                \ \
    __  ____ ____    _____________ ____ __   __ __     / /
    \ \/___ \\__ \  /___  _____/\ \\__ \\ \  \ \\ \   / /
    / /   / /  /  \    / /Melroch\ \_/ // /  / // /  / /
   / /___/ /_ / /\ \  / /Melarocco\_  // /__/ // /__/ /
  /_________//_/  \_\/ /Eowine__   / / \___/\_\\___/\_\
Gwaedhvenn Angelmiel\ \_____/ / a/ /_adar Merthol naun
~~~~~~~~~Cuinondil~~~\_______/~~~\__/~~~Noolendur~~~~~~
|| Lenda lenda pellalenda pellatellenda cuivie aiya! ||
"A coincidence, as we say in Middle-Earth" (JRR Tolkien)

22nd June 2000: Re: Using LaTex styles with plain TeX (was: Web site information)

From: doug@essex.ac.uk (Arnold D J)
Subject: Re: Web site information

Just a short response to BP (B.Philip Jonsson)   wrote 
bemoaning the lack of support for plain \TeX (e.g. at the latex4ling
site).

Actually, of course, a lot of the stuff there actually does work with
plain \TeX -- but of course it is not guaranteed.

However, I think the gb4e macros for example numbering and glosses do
work for plain \TeX.

     http://clwww.essex.ac.uk/latex4ling/

------------------------------------------------------------------
Doug Arnold,  		     	   doug@essex.ac.uk
Dept. of Language & Linguistics,   http://clwww.essex.ac.uk/~doug
University of Essex,	     	   
Wivenhoe Park,		     	   Tel: +44 1206 872084 (direct)
Colchester, CO4 3SQ, UK.     	   Fax: +44 1206 872198
------------------------------------------------------------------

22nd June 2000: Re: Using LaTex styles with plain TeX

From: Alexis Dimitriadis 
Subject: Re: Web site information

> However, I think the gb4e macros for example numbering and glosses do
> work for plain \TeX.

Sorry to contradict, but don't bother to try.  They don't.  Nothing that
uses LaTeX environments (\begin{X} ... \end{...}) will.

Alexis

22nd June 2000: Re: Tree macros

From: BP Jonsson 
Subject: Re: Tree macros (was: YATM! :-)

>Specifically, what I'm looking for is a tree-macro package that
>would compile nicely for both PS and PDF documents (without having
>to resort to Distiller in the latter case). Unfortunately, this
>rules out most of the tree macros out there.

Have you tried dvipdfm?

FWIW I do my trees in a Prolog program which outputs PS.  Unfortunately 
I've promised not to samizdat it...

/BP

 B.Philip Jonsson  bpj@netg.se 
melroch@mail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

"Doubt grows with knowledge" -Goethe

22nd June 2000: Re: Tree macros

From: Michel Eytan 
Subject: Re: Tree macros (was: YATM! :-)

> FWIW I do my trees in a Prolog program which outputs PS.  Unfortunately
> I've promised not to samizdat it...
>
> /BP

I remember indeed a Prolog program for Mac done with LPA MacProlog. I seem
to have lost or mislaid it...

BTW I deplore the steep price of LPA versus the now (alas) defunct
AAIS_Prolog that was a pure marvel -- bought in 1994 it still runs under
MacOS 9.0.4!!!

Cheers
 ~=michel
--
Michel Eytan                                       eytan@dpt-info.u-strasbg.fr
                                    In a World without Fences who needs Gates?

23rd June 2000: Re: Tree macros

From: "Malouf R." 
Subject: Re: Tree macros (was: YATM! :-)

Hi,

While we're at it, let me mention that pstricks is the best solution I've
found for drawing trees.  The pstricks syntax is a little cumbersome, so I've
written a set of interface macros to make using them easier, and also to
integrate them with Chris Manning's AVM macros.  The big advantange these
macros have over tree-dvips is that they handle node placement as well as edge
placement. If anyone's interested, my macros are available at:

	http://www.let.rug.nl/~malouf/rtrees.tar.gz

In the long run, what we probably need is a set of tree-drawing macros for
metapost.  That way we'd be able to produce trees in either eps or pdf
directly, without having to wrestle with postscript \special's and such.
- --
Rob Malouf
malouf@let.rug.nl

23rd June 2000: Re: Tree macros

From: Joseph Hilferty 
Subject: Re: Tree macros (was: YATM! :-)

First, thanks for all the replies. They are very much appreciated.

BP Jonsson wrote:
> >Specifically, what I'm looking for is a tree-macro package that
> >would compile nicely for both PS and PDF documents (without having
> >to resort to Distiller in the latter case). Unfortunately, this
> >rules out most of the tree macros out there.
> 
> Have you tried dvipdfm?

Yes, it works with qtree, for example, though the output to PDF
really isn't quite as nice as the PS version from DviPs. DviPdfm
cannot handle anything by tree-dvips (for example, the arrows!),
which is a shame.

BTW, Alexis Dimitriadis pointed out to me that there is a way to
make qtree run under PdfLaTeX: all you have to do is comment out
the line containing "\@input{eepic.sty}." It seems that there's an
incompatability between eepic specials and PdfLaTeX. Unfortunately,
the branches don't look as well on screen as they would if eepic
could be used. In any event, the trees seem to look better printed
out.

Joe Hilferty
__________________________________________________________
Home page: http://lingua.fil.ub.es/~hilferty/homepage.html

23rd June 2000: Re: Using LaTex styles with plain TeX

From: "Michael W. Daniels" 
Subject: Re: Web site information

>Sorry to contradict, but don't bother to try.  They don't.  Nothing that
>uses LaTeX environments (\begin{X} ... \end{...}) will.

Not always the case -- one can sometimes replace \begin{X} with \X and 
\end{X} with \endX (and perhaps some \begingroup and \endgroup commands) to 
use the package.

For example, the avm.sty package works with \avm and \endavm just as well 
as with \begin{avm} and \end{avm}.

Now if the definitions of \X and \endX use latex-internal commands 
themselves, then you have more work to do. It's not too difficult to 
convert certain types of latex macros to plain tex macros.

For example, I have a version of tipa hacked up to work under plain tex. To 
do this, I just ran the file through plain tex; every time I found an 
unrecognized command, I searched through latex.ltx for its definition and 
made the appropriate replacement. After you do this enough times, you'll 
have something that only uses tex primitives.

--
Michael W. Daniels          | "But the evil is that they hold for
daniels@ling.ohio-state.edu |    certain that they are in the light."
Dept. of Linguistics        |

23rd June 2000: Re: Using LaTex styles with plain TeX

From: Alexis Dimitriadis 
Subject: Re: Using LaTex styles with plain TeX

> >Sorry to contradict, but don't bother to try.  They don't.  Nothing that
> >uses LaTeX environments (\begin{X} ... \end{...}) will.
> 
> Not always the case -- one can sometimes replace \begin{X} with \X and 
> \end{X} with \endX (and perhaps some \begingroup and \endgroup commands) to 
> use the package.

Indeed; if, as you say, the package does not use latex-specific features.

The styles in question (gb4e and cgloss4e) use the \begin{}/\end{} syntax
inside the style files, so they cannot be used without modifications.

Using plain TeX syntax to invoke a LaTeX-style command is one thing, but
converting a style to TeX syntax is a much bigger undertaking.  The
example macros in gb4e use the LaTeX {list} environments, which are big and
messy, depend on all sorts of other latex services like latex-style counters,
and are simply a pain to puzzle out (I've tried).  I bet it would be easier to
just write a TeX implementation from scratch.

There's no reason cgloss4e.sty should not work with plain TeX--after all, it
is a TeX package with cosmetic changes (but why not just use the TeX
original?).

It's a shame, really, that more packages aren't written so they work with
either TeX or LaTeX.  Sometimes this would be easy but the authors just don't
think about plain TeX, I guess.  I'm guilty of this myself.  Font styles are a
good example: as far as I know, there's no reason they could not all be
written to work with plain TeX, with the LaTeX-specific parts conditional on
some \if switch.  (Perhaps the authors of tipa could incorporate your
revisions to make it TeX compatible?)

Alexis "Used to use plain TeX till I saw the writing on the wall" Dimitriadis

> For example, the avm.sty package works with \avm and \endavm just as well 
> as with \begin{avm} and \end{avm}.
> 
> Now if the definitions of \X and \endX use latex-internal commands 
> themselves, then you have more work to do. It's not too difficult to 
> convert certain types of latex macros to plain tex macros.
> 
> For example, I have a version of tipa hacked up to work under plain tex. To 
> do this, I just ran the file through plain tex; every time I found an 
> unrecognized command, I searched through latex.ltx for its definition and 
> made the appropriate replacement. After you do this enough times, you'll 
> have something that only uses tex primitives.
> 
> --
> Michael W. Daniels          | "But the evil is that they hold for
> daniels@ling.ohio-state.edu |    certain that they are in the light."
> Dept. of Linguistics        |

23rd June 2000: Re: Using LaTex styles with plain TeX

From: "Michael W. Daniels" 
Subject: Re: Using LaTex styles with plain TeX

>think about plain TeX, I guess.  I'm guilty of this myself.  Font styles are a
>good example: as far as I know, there's no reason they could not all be
>written to work with plain TeX, with the LaTeX-specific parts conditional on
>some \if switch.  (Perhaps the authors of tipa could incorporate your
>revisions to make it TeX compatible?)

The problem is the NFSS -- porting anything (like tipa) that depends on it 
would be quite complex. I had to modify the tipa code to assume certain 
things about my documents that the latex version didn't have to (i.e. like 
what font I was using for the base text and such). One would have to port 
the NFSS itself first to get something that could be distributed...

--
Michael W. Daniels          | "But the evil is that they hold for
daniels@ling.ohio-state.edu |    certain that they are in the light."
Dept. of Linguistics        |

23rd June 2000: Re: Web site information

From: Joseph Hilferty 
Subject: Re: Web site information

BP Jonsson wrote:
> Well, seriously, on the site there is mentioned that
> 
> >Covington's. which are a LaTeX adaptation of M. deGroot's Midnight gloss
> >TeX-macros.
> 
> Where can I find these presumably plain macros?

I think that you can find covington style at CTAN. Or you can
try here, I think:

ftp://ftp.ai.uga.edu/pub/tex/

I don't know if it will work with plain TeX.

Joe Hilferty
__________________________________________________________
Home page: http://lingua.fil.ub.es/~hilferty/homepage.html

23rd June 2000: Re: Using LaTex styles with plain TeX

From: Werner LEMBERG 
Subject: Re: Using LaTex styles with plain TeX

> The problem is the NFSS -- porting anything (like tipa) that depends
> on it would be quite complex. I had to modify the tipa code to
> assume certain things about my documents that the latex version
> didn't have to (i.e. like what font I was using for the base text
> and such). One would have to port the NFSS itself first to get
> something that could be distributed...

This has changed meanwhile.  Han The Thanh has ported NFSS to Plain
TeX; he calls this `plnfss'.  It is part of his vntex package
(Vietnamese for TeX and LaTeX) -- not on CTAN yet, AFAIK.  I don't
know exactly how it works (since I've never used it), but it seems
that high-level LaTeX 2e font encoding stuff works unmodified.  In
combination with Vladimir Volovich's `plainenc' package (also in vntex
-- originally developed for the Cyrillic T2* encodings) which provides
full Plain TeX support for LaTeX's inputenc mechanism, it should be
simpler now to adapt many LaTeX 2e packages to Plain TeX.

Another useful thing in vntex is Frank Mittelbach's dblaccnt package
which defines a nice interface to access characters with up to two
accents (as needed for Vietnamese).  I can imagine that some of you
can need such things for other purposes also.

vntex can be found at

  http://www.fi.muni.cz/~thanh/download/vntex.zip (~600KB)

    Werner

23rd June 2000: Re: Web site information

From: Alexis Dimitriadis 
Subject: Re: Web site information

> Now I'd like to invite anyone else who has a relevant 
> website they'd like to have linked from/to ling-tex
> to post the info here and Dag Langmyhr, our webmaster, will  
> see that it's added. 

http://www.ling.upenn.edu/advice/latex.html
    TeX/LaTeX information for linguists.

    This page is intended to be selective, not encyclopaedic; I tried to
    provide access to basic latex materials as well as a basic suite of
    linguist tools (numbered examples, references, trees, glosses, IPA,
    graphics) and links to more resources.

Some other suggestions for links:

http://www.sil.org/computing/fonts/
    SIL's Fonts in Cyberspace

    A huge database of language fonts that includes some TeX fonts (not
    sure how thorough the coverage is, though; no IPA fonts for TeX come
    up)

Not linguist-specific:
    
http://www.giss.nasa.gov/latex/ltx-2.html
    A hypertext index of LaTeX commands

    Wonderfully organized by command or by topic, courtesy of NASA's Goddard
    Institute for Space Studies. (Even rocket scientists need to read the
    manual!)

It would also be good to have links to CTAN and TUG (ctan.tug.org,
www.tug.org) along with the linguist-specifics links.

Glad you're putting together a page of LaTeX pages!  I'll be looking
forward to the completed product.

Alexis

23rd June 2000: Re: Web site information

From: Chris Manning 
Subject: Web site information

Although it appears not to have been updated for a bit, Geert-Jan
Kruijff has a quite good listing of relevant stuff at:

	http://kwetal.ms.mff.cuni.cz/~gj/latex.html

and my own contributions are now canonically at:

	http://nlp.stanford.edu/~manning/tex/
Chris.

23rd June 2000: Re: Web site information

From: Geert-Jan Kruijff 
Subject: Re: Web site information

> Although it appears not to have been updated for a bit, Geert-Jan
> Kruijff has a quite good listing of relevant stuff at:

> 	http://kwetal.ms.mff.cuni.cz/~gj/latex.html

Thanks Chris ;-) The site is also available via
, which is the official way to reach
it. An update will hopefully follow sooner or later, once I'm done
writing my dissertation (so, let's hope that it will be that
"sooner"-thing :-) ...).

Totally unrelated question:

Does anyone have some lines of code for creating a "list of symbols"?
I've been trying today to set up something like that, using the
following bit:


%%\makeatletter
%%\setlength{columnseprule}{0pt}
%%\newcommand{\listofsymbols}{\@starttoc{smb}}
%%\newcommand{\symtt}[1]{\addcontentsline{smb}{symbols}{#1}#1}  
%%\newcommand{\l@symbols}{\@dottedtocline{1}{1.5em}{2.3em}}
%%\newcommand{\l@symbols}[2]{\par\noindent\ensuremath{#1} ... {\it #2}}
%%\makeatother

later followed by 

%%\chapter*{Index symbolicae}
%%\begin{multicols}{2}
%%\listofsymbols
%%\end{multicols}

Now my problem is that LaTeX stumbles on the math-stuff. Does anyone
know of a way around that problem? 

Cheers, Geert-Jan

24th June 2000: Re: Using LaTex styles with plain TeX

From: Anshuman Pandey 
Subject: Re: Using LaTex styles with plain TeX

> Another useful thing in vntex is Frank Mittelbach's dblaccnt package
> which defines a nice interface to access characters with up to two
> accents (as needed for Vietnamese).  I can imagine that some of you
> can need such things for other purposes also.

[snip]

How does this dblaccnt package differ in function from diatop?

Regards,
Anshuman Pandey

24th June 2000: Re: Using LaTex styles with plain TeX

From: Werner LEMBERG 
Subject: Re: Using LaTex styles with plain TeX

> > Another useful thing in vntex is Frank Mittelbach's dblaccnt
> > package which defines a nice interface to access characters with
> > up to two accents (as needed for Vietnamese).  I can imagine that
> > some of you can need such things for other purposes also.
> 
> How does this dblaccnt package differ in function from diatop?

No idea :-)  I've never seen diatop (and currently no access to my
CTAN CDs).

dblaccnt isn't very sophisticated; it `simply' defines one additional
command for font encoding packages.  Example (taken from t5enc.def for
Vietnamese):

  \DeclareTextDoubleComposite{\~}{T5}{\^}{o}{233}

Now you can say \~\^o and you will get glyph 233 in T5 encoding.  Such
double accents can be also used in input encodings.  Example (taken
from viscii.def):

  \DeclareInputText{178}{\~\^o}

This maps input character 178 to \~\^o.  Note that everything works
with \Make{Upper,Lower}case.

Frank's file could be extended to support triple accents which would
be useful for ancient Greek.  After some necessary minor clean-ups in
the LaTeX kernel it probably will become an `official' extension to
LaTeX 2e.

    Werner

28th June 2000: List of symbols

From: Klaus Lagally 
Subject: list of symbols

> From: Geert-Jan Kruijff 

[ ... ]
> 
> Totally unrelated question:
> 
> Does anyone have some lines of code for creating a "list of symbols"?
> I've been trying today to set up something like that, using the
> following bit:
> 
> 
> %%\makeatletter
> %%\setlength{columnseprule}{0pt}
> %%\newcommand{\listofsymbols}{\@starttoc{smb}}
> %%\newcommand{\symtt}[1]{\addcontentsline{smb}{symbols}{#1}#1}  
> %%\newcommand{\l@symbols}{\@dottedtocline{1}{1.5em}{2.3em}}
> %%\newcommand{\l@symbols}[2]{\par\noindent\ensuremath{#1} ... {\it #2}}
> %%\makeatother

You have two defining instances of \l@symbols; 
individually they work with me :-)

Klaus

> 
> later followed by 
> 
> %%\chapter*{Index symbolicae}
> %%\begin{multicols}{2}
> %%\listofsymbols
> %%\end{multicols}
> 
> Now my problem is that LaTeX stumbles on the math-stuff. Does anyone
> know of a way around that problem? 

My LaTeX does not stumble there. What exactly happens?

> Cheers, Geert-Jan
 
Klaus
- -- 
Prof. Dr. Klaus Lagally  | lagally@informatik.uni-stuttgart.de
Institut fuer Informatik | Tel.  +49-711-7816392 |  Zeige mir deine Uhr,
Breitwiesenstrasse 20-22 | FAX   +49-711-7816370 |  und ich sage dir,
70565 Stuttgart, GERMANY |             (changed) |  wie spaet es ist.

28th June 2000: Re: List of symbols

From: Geert-Jan Kruijff 
Subject: Re: list of symbols

>> Does anyone have some lines of code for creating a "list of
>> symbols"?  I've been trying today to set up something like that,
>> using the following bit:
>> 
>> 
>> %%\makeatletter %%\setlength{columnseprule}{0pt}
>> %%\newcommand{\listofsymbols}{\@starttoc{smb}}
>> %%\newcommand{\symtt}[1]{\addcontentsline{smb}{symbols}{#1}#1}
>> %%\newcommand{\l@symbols}{\@dottedtocline{1}{1.5em}{2.3em}}
>> %%\newcommand{\l@symbols}[2]{\par\noindent\ensuremath{#1} ... {\it
>> #2}} %%\makeatother

> You have two defining instances of \l@symbols; individually they
> work with me :-)

I use either one or the other - but the "commenting out" doesn't show
here, I agree ;-) 

>>  later followed by
>> 
>> %%\chapter*{Index symbolicae} %%\begin{multicols}{2}
>> %%\listofsymbols %%\end{multicols}
>> 
>> Now my problem is that LaTeX stumbles on the math-stuff. Does
>> anyone know of a way around that problem?

> My LaTeX does not stumble there. What exactly happens?

The problem I get is the following:

---------------------------BEGIN LATEX RUN-----------------------------------

 (phd-main.smb
! Improper alphabetic constant.
 
                   \spacefactor 
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Improper \spacefactor.
 \spacefactor 
                             
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Improper alphabetic constant.
 
                   \spacefactor 
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Improper \spacefactor.
 \spacefactor 
                             
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Missing $ inserted.
 
                $
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Extra }, or forgotten $.
 }
                 
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
! Extra }, or forgotten $.
\l@symbols ... \protect {\mbox {\ensuremath {#1}}}
                                                   ... {\it #2}
l.1 ...`\spacefactor \@m 12\relax ${}\sb {i}$}{59}
                                                  
)
! Missing $ inserted.
 
                $
l.331 \listofsymbols
                    
! Missing } inserted.
 
                }
l.331 \listofsymbols
                    
! Missing } inserted.
 
                }
l.331 \listofsymbols
                    

---------------------------END LATEX RUN-----------------------------------

I am loading the following packages, in the order as stated:

\usepackage{amsthm,amsmath}
\usepackage{latexsym,amssymb}
\usepackage{multind}
\usepackage{proof}
\usepackage{prooftree}
\usepackage{stdendnotes} % Mark's endnotes
\usepackage{names} 
\usepackage{terms}
\usepackage{gb4e,cgloss4e}
\usepackage{aclbib}
\usepackage{multicol}
\usepackage{epic,ecltree}
\usepackage{diagrams}
\usepackage{avm}

I am trying to index:

\symtt{\atsign_i}

whereby \atsign is defined as follows:

\newcommand{\atsign}{\makeatletter{}@\makeatother}

Maybe that's where things go wrong ... 

Cheers, Geert-Jan

28th June 2000: Re: Using LaTex styles with plain TeX

From: BP Jonsson 
Subject: Re: Using LaTex styles with plain TeX

At 22:43 24.6.2000 +0200, Werner LEMBERG wrote:
> > How does this dblaccnt package differ in function from diatop?
>
>No idea :-)  I've never seen diatop (and currently no access to my
>CTAN CDs).
[snip]
>   \DeclareTextDoubleComposite{\~}{T5}{\^}{o}{233}

I've done something similar like this:

      \def\?#1{{\diatop:\u,\=#1;}}

As you can see I changed the diatop syntax a bit! (Don't even remember the 
original one! :-)

/BP
  B.Philip Jonsson  mailto:melroch@mail.com

28th June 2000: Re: Using LaTex styles with plain TeX

From: Christina Thiele 
Subject: Re: Using LaTex styles with plain TeX

BP Jonsson writes:
> At 22:43 24.6.2000 +0200, Werner LEMBERG wrote:
> > > How does this dblaccnt package differ in function from diatop?
> >
> >No idea :-)  I've never seen diatop (and currently no access to my
> >CTAN CDs).
> [snip]
> >   \DeclareTextDoubleComposite{\~}{T5}{\^}{o}{233}
> 
> I've done something similar like this:
> 
>       \def\?#1{{\diatop:\u,\=#1;}}
> 
> As you can see I changed the diatop syntax a bit! (Don't even remember the 
> original one! :-)
> 
> /BP

The original \diatop definition, straight from the source file (mine
;-) ). I believe we also put it into wsuipa macros from the late
80s. At least, I've got it stashed in my copy of ipamac.tex ...

\def\diatop[#1|#2]{\leavevmode{\setbox1=\hbox{{#1{}}}%
                     \setbox2=\hbox{{#2{}}}%
                     \dimen0=\ifdim\wd1>\wd2\wd1\else\wd2\fi%
                     \dimen1=\ht2\advance\dimen1by-1ex%
                     \setbox1=\hbox to1\dimen0{\hss#1\hss}%
                     \rlap{\raise1\dimen1\box1}%
                     \hbox to1\dimen0{\hss#2\hss}}}%
 
I don't know that it's been changed at all. Oh, one change from the
very first publication in my 1987 paper: there was no \leavevmode,
which I'd added a while before that time. I think that's the only
change that's ever been made to it. Yeah, I just checked. 

Now, I didn't write the macro -- as anyone who knows me would know
;-))). I think it was either Jean-Pierre Paillet or Michael Dunleavy,
both at Carleton University in the mid-80s when this was designed.

But that's the original and it's been pretty handy for a lot of
people. If anyone teaches TeX classes, might be fun to set this as a
test: how many different accents, above and below, can you get \diatop
to successfully place ;-)

Oh well ... back to work.

Ch.

28th June 2000: Re: Using LaTex styles with plain TeX

From: BP Jonsson 
Subject: Re: Using LaTex styles with plain TeX

At 09:31 28.6.2000 -0400, Christina Thiele wrote:
>The original \diatop definition, straight from the source file (mine
>;-) ).

I changed it for a very pedestrian reason: to have the necessary keys 
easily available on my Swedish keyboard!

/BP
  B.Philip Jonsson  mailto:melroch@mail.com

30th July 2000: Mapping lines

From: "Yoshinari Fujino" 
Subject: mapping lines

Dear Subscribers.

I would appreciate it if anyone could give me some advice on
reproducing the mapping (lines) which appeared in
Lass, Roger (1984) _Phonology_ Cambridge: CUP, p296. (12.3)
The lines themselves look similar to association lines of autosegmental
phonology.
However, the numbers of branching from the single node are maximally four.
I have no idea what package would work for my case.

Thank you very much.

Yoshinari Fujino
Email: yfujino@geocities.co.jp
Web: http://www.geocities.co.jp/CollegeLife-Labo/9551/

30th July 2000: Re: Mapping lines

From: "Michael W. Daniels" 
Subject: Re: mapping lines

>I would appreciate it if anyone could give me some advice on
>reproducing the mapping (lines) which appeared in
>Lass, Roger (1984) _Phonology_ Cambridge: CUP, p296. (12.3)
>The lines themselves look similar to association lines of autosegmental
>phonology.

I haven't seen this kind of diagram, but I have had adequate results doing 
autosegmental diagrams using the xy package:

\xymatrix@-1pc{
\ar@{--}[drr]\txt{C} & {\txt{+}} & \ar@{-}[d]\txt{C}                       \\
                      &           & \ar@{-}[dr]\txt{root}                   \\
                      &           &                       & {\txt{-approx}} \\
}

The basic idea is that you have a normal TeX-style table layout, and the 
\ar commands let you connect the cells.

So I have three rows, each of which has four columns. In the first row, I 
put a C and draw a dashed line {--} to the cell one down and two to the 
right (labelled 'root'). I put a '+' in the next cell, and a C in the third 
cell, drawing a solid {-} line down to the same root node.  A command like 
\ar@{-}|=[d] would 'interrupt' the line with an equal sign (to represent 
deleted associations). You can get arrowheads and -tails by using line 
styles like {>->}, squiggly lines with {~}, and so on.

The package is much more complex than I've described; a lot of people tend 
to avoid it because of its complexity. By default, I believe it works 
entirely with tex commands and metafonts, so there shouldn't be problems 
with pdf conversion down the line if that's where you're headed.

As for availibility, I believe it's included with modern versions of teTeX, 
and it's definitely on CTAN.

--
Michael W. Daniels          | "But the evil is that they hold for
daniels@ling.ohio-state.edu |    certain that they are in the light."
Dept. of Linguistics        |

30th July 2000: Re: Mapping lines

From: "Yoshinari Fujino" 
Subject: Re: mapping lines

I will try xy package and see whether it is far beyond of my ability or not.
I think I have to be equipped with lots of techniques to be a texnichian.

Yoshinari Fujino
Email: yfujino@geocities.co.jp
Web: http://www.geocities.co.jp/CollegeLife-Labo/9551/

30th July 2000: Re: Mapping lines

From: Christina Thiele 
Subject: Re: mapping lines

I didn't hop in on this query when it first arose because ... well
... I don't think I know what `autosegmental phonology' is ;-)) That
is, I probably would recognise the structures if I saw them, but the
term itself ... 

I was hoping someone else to would ... euh ... draw a picture, so to
speak, and I would learn-by-lurking ... ;-) 

If anyone has copies of CJL: (Cdn Jrnl. of Ling.) ... are there any
articles in there that use the assocation lines that Yoshinari is
looking for? If yes, tell me and I'll check the files. I've just
looked at article titles for the past several years and no
`Autosegmental' turns up. 

Just wondering if there's an easier solution than xy, which so far
seems to be his only option ... 

Ch.

30th July 2000: Re: Mapping lines

From: krussll@CC.UMANITOBA.CA
Subject: Re: mapping lines

Here are a couple of articles from 1995.  I'm pretty sure that's not the
last year I remembered to renew my CLA membership, but it's all I have
lying around the office :-).

Mohamed Elmedlaoui    p 39  ff
Emmanuel Nikiema      p 319 ff

> Just wondering if there's an easier solution than xy, which so far
> seems to be his only option ... 

I commonly use the various ps... packages for autosegmental
diagrams.  They work just fine.  If PDF output is a concern, I've never
had any problems converting the resulting PS files to PDF using
Ghostscript.

-- Kevin Russell

30th July 2000: Re: Mapping lines

From: Christina Thiele 
Subject: Re: mapping lines

Wonderful! Thanks!!

OK. I've got Nikiema's file here ... Yup, I was using Emma's
tree-dvips package. Same would be true of Elmedlaoui's article.

It's possible that this might be an easier solution. Here's what I
have in my cjl style file for the macros -- you can find her stuff on
CTAN. 

====

%% TREE DIAGRAMS:

%% The following .sty file by Emma Pease allows for fantastic tree
%%   diagrams to be drawn within LaTeX, using dvips. Avoids a lot of
%%   cut-and-paste work!

\input tree-dvips.sty


% Shorthand macros for trees (taken from zanuttini-l.sty):

%% \dots\ alone in a tree diagram don't have enough height.
%%   Add an \rlap'd \phantom F to get the xheight:
\newcommand{\ndots}{{\dots\rlap{\p{F}}}}  %% = `nodedots'

%% for a phantom node:
\newcommand{\phnode}[1]{{\p{#1}}}

====

And here's a sample of the input from Nikiema -- ex. 15 on p.331:

%% Ex 15:
\Example{Effets du ton flottant issu de l'\'elision vocalique

{\tabcolsep=4pt
\begin{tabular}[t]{lccccccccccccl}
%% Row 1:
a. & B & H & & 
      & B & \node{a}{H} & & 
         & B & \node{b}{H} & \\
   & $\vert$ &   & & 
      & $\vert$ & $\vert$ & & 
         & $\vert$ &   & \\
   & a\rlap{\ts k} & v\rlap{\nn{0}} & \ts \llap{w}\ts i 
            & $\longrightarrow$
      & a\rlap{\ts k} & v\rlap{\nn{0}} & \ts \llap{w}\ts \node{c}{i} 
                  & $\longrightarrow$
         & a\rlap{\ts k} & \eset & \llap{w}\node{d}{\ts i} 
                        & [\`akw\'\i] \\ [6pt]
%% Row 2:
b. & B & B & H & 
      & B & B & H & 
         & B & \node{e}{B} & H \\
   & $\vert$ &  &  & 
      & $\vert$ & $\vert$ & $\vert$ & 
         & $\vert$ &  & $\vert$ \\
   & e\rlap{\ts k} & v\rlap{\nn{0}} & \ts \llap{w}\ts e 
            & $\longrightarrow$
      & e\rlap{\ts k} & v\rlap{\nn{0}} & \ts \llap{w}\ts e 
                   & $\longrightarrow$
         & e\rlap{\ts k} & \eset & \llap{w}\node{f}{\ts e} 
                          & [\`ekw\v e] \\ [6pt]
%% Row 3:
c. & B & B & H & 
      & B & B & H & 
         & B & \node{g}{B} & H \\
   & $\vert$ &  &  & 
      & $\vert$ & $\vert$ & $\vert$ & 
         & $\vert$ &  & $\vert$ \\
   & a\rlap{\ts b} & v\rlap{\nn{0}} & \ts \llap{w}\ts i 
        & $\longrightarrow$
      & a\rlap{\ts b} & v\rlap{\nn{0}} & \ts \llap{w}\ts i 
               & $\longrightarrow$
         & a\rlap{\ts b} & \eset & \llap{w}\ts \node{h}{i} 
                     & [\`abw\v\i] \\
\end{tabular}
%
{\dashlength=3pt
\nodeconnect{a}{c} \nodeconnect{b}{d} 
\nodeconnect{e}{f} \nodeconnect{g}{h}
}}}


====

Notes:

1. I use \llap and \rlap a lot, to fudge the measurable width of a
   given column to get things to line up nicely

2. \ts = \thinspace. You'll notice \rlap's with a \ts + a space
    + a letter, all being done in an effort to get better visual
    alignments. 

3. \nn is a shorthand of mind, def'd as: 

   \newcommand{\nn}[1]{{\leavevmode%
                  \setbox0=\hbox{\raise.7ex\hbox{\footnotesize #1}}%
                  \ht0=0pt\box0{}}}

3. I'm not averse to using $\vert$ from math mode for a short
   vertical rule rather than coding up two nodes and drawing
   a line between then. Also, it's sometimes tricky to get a 
   true vertical rule between 2 nodes supposedly one above the
   other -- \vert is guaranteed to be vertical as it doesn't
   deal with node connects at all. 

4. \dashlength is what you use to make dashed connecting lines;
   if you set the value to 1pt, you get a dotted line.

And thanks to Emma -- it's a great package!

Ch.

10th July 2000: TIPA (and other files) update

From: "Todd O'Bryan" 
Subject: TIPA (and other files)

My computer recently crashed and I'm having to reconstruct how I got TIPA
and a few other packages (like natbib) to work. The OzTeX manual recommends
making inputs, tfm, and vf folders inside a mytex folder so that you can
just copy the mytex folder when you upgrade. Unfortunately, I'm not sure
what to put where. Can anybody point me to a resource that explains where to
put things?

Todd

11th July 2000: Re: TIPA (and other files) update

From: Joost Kremers 
Subject: Re: TIPA (and other files)

hi todd,

well, my strategy (i use miktex on W95 and tetex on linux) is simply to
imitate the directory structure of the tex directory (whatever that is in
oztex) in the local tex directory (`mytex' in your case).

so for example on linux, i have (among others) the following directories in
the tex tree:

/usr/share/texmf/fonts/tmf
/usr/share/texmf/fonts/vf

i just created the following corresponding directories in the local tex tree:

/usr/local/share/texmf/fonts/tmf
/usr/local/share/texmf/fonts/vf

etc.

i believe there is a document somewhere that describes this, but i couldn't
find it. perhaps someone else knows? 

BTW, i thought macs didn't crash... :-)

HTH
joost

>My computer recently crashed and I'm having to reconstruct how I got TIPA
>and a few other packages (like natbib) to work. The OzTeX manual recommends
>making inputs, tfm, and vf folders inside a mytex folder so that you can
>just copy the mytex folder when you upgrade. Unfortunately, I'm not sure
>what to put where. Can anybody point me to a resource that explains where to
>put things?
>
>Todd

----------------------------------------
Joost Kremers
Katholieke Universiteit Nijmegen
Talen en Culturen van het Midden-Oosten
Postbus 9103
6500 HD Nijmegen
tel: 024-3612996
fax: 024-3611972

12th July 2000: Re: TIPA (and other files) update

From: Martin Steer 
Subject: Re: TIPA (and other files)

Hi Todd,

My understanding is that as far as OzTeX is concerned it doesn't matter
where you put things as long as they can be found! You set OzTeX up to find
them by modifying your *local* file, which is in the main OzTeX folder.
Instructions as to how to do this are in the *local* file itself.

I think, like Joost, that a good way to organise things is just to imitate
the main directory structure within your own mytex folder, which should be
within the OzTeX folder. So, something you add -- such as all the natbib
stuff -- that would otherwise go in the tools subfolder in the main
directory, goes in the mytex:tools:natbib subfolder. You let OzTeX know
where it is by adding <input_folders  = $c :mytex:tools:* :mytex:tools:> to
your local file, as shown below. Similarly for other subfolders and
applications. When you upgrade, take your local file and your mytex folder
and throw the rest away!

%%%%%%%%%%%%% put OzTeX-only keywords after the next line %%%%%%%%%%%%%%
current_app = OzTeX

input_folders  = $c :mytex:tools:* :mytex:tools:
%line added 20.2.00


+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+
Martin Steer
Department of Linguistics
Research School of Pacific and Asian Studies
Australian National University

13th July 2000: Re: TIPA (and other files) update

From: "Todd O'Bryan" 
Subject: TIPA (and other files) update

Thanks to Martin Steer and Joost Kremers for their help with reconstructing
my TeX system. The problem turned out to be much more insidious and
confusing than I could have imagined.

Somehow, my FTP program decided that the .tfm files I had downloaded to work
with TIPA were text files, not data files, so it converted them in the
process of downloading. What I thought were messages telling me that TeX
couldn't find the .tfm file were, in fact, messages telling me that the .tfm
files were not good. (I guess I should pay more attention to error messages,
huh? :-)

Anyway, thanks again.

Todd

17th July 2000: LaTeX under Windows

From: "Yoshinari Fujino" 
Subject: LaTeX under Windows

Dear Subscribers

I have a general curiosity that what free distributions of LaTeX2e for
MS-Windows users would be recommended.

TUG website lists two ports for Windows users but I am not certain which
would be easier to install for newcomers. I myself use pLaTeX2e run on
Japanese system but I have no idea if it is a better choice.

I currently wrote install instruction of pLaTeX and TIPA for Windows users
in
Japanese.(http://www.geocities.co.jp/CollegeLife-Labo/9551/pLaTeXintro-jp.ht
ml)
I am preparing to do the same thing in English.

Yoshinari Fujino
Email: yfujino@geocities.co.jp
Web: http://www.geocities.co.jp/CollegeLife-Labo/9551/

17th July 2000: Re: LaTeX under Windows

From: Zsuzsanna Nagy 
Subject: Re: LaTeX under Windows

I've been using emtexgi for more then two years. It is a Win application
that runs tex and latex in their emtex version (developed for DOS).
It is very stable and easy to work with, here is the web-page where you
can find the information:
http://www.wfu.edu/Academic-departments/Economics/ftp/emtexgi.html

Emtexgi was not "upgraded" for Win98, but I had no problem with
upgraded my OS from Win95 to Win98, I didn't have to touch anything
related to emtexgi/latex/etc. I assume you can just install emtexgi
under Win98 right away.

I know, running a Dos compiler under Win is not really what Donald Knuth
might have dreamt about, but emtexgi works fine, makes it easy to run
bibtex, makeindex, gsview... on top of LaTeX.

I also installed TIPA fonts on my computer and through emtexgi I can just
load the TIPA stuff as any regular font.

I hope it was helpful!

Good luck,
Zsuzsa Nagy

17th July 2000: Headings

From: "Todd O'Bryan" 
Subject: Headings

Dear Ling-TeXers,

A friend asked me to help type-set a document for a conference proceedings,
and we're having trouble with the headers and footers.

Using the myheadings pagestyle, we were able to get the running heads that
go at the top of each page (different left and right ones, of course), but
now the headers run into the text. I'm guessing that judicious playing with
\topmargin, \headsep, and such will fix that, but if anybody has suggestions
or warnings I'd be most appreciative.

The big problem is that the first page does not have a header, but a footer.
We tried doing what it suggests to do in Lamport's LaTeX-A Document
Preparation System, namely define our own pagestyle (by using
\newcommand{\ps@myfoot} and redefining the four commands that produce the
header and footer text) and then using \thispagestyle{myfoot} on the first
page, but we can't get it to TeX correctly.

Either we get a "Missing \begin{document}" warning (if we try to put the
definitions above the \begin{document} with \thispagestyle in the text of
the first page) or we get strange text in the *header* including all the
commands we were trying to redefine.

The document uses a few packages: times, tipa, and one defined by the
author, so I suppose they could be interfering.

Anyway, if someone has come up against this problem and knows what we're
doing wrong, we'd greatly appreciate some words of advice.

Todd

17th July 2000: Re: Headings

From: krussll@CC.UMANITOBA.CA
Subject: Re: Headings

A quick, dirty, but wasteful work-around is to generate a blank page at
the beginning that gets the footer, and start the real text on page two.
So if the numbering for the conference paper should start on page 173,
start the document with:

\setcounter{page}{172}
\clearpage

-- Kevin Russell

18th July 2000: Re: Headings

From: alexis@babel.ling.upenn.edu (Alexis Dimitriadis)
Subject: Re: Headings

> Using the myheadings pagestyle, we were able to get the running heads that
> go at the top of each page (different left and right ones, of course), but
> now the headers run into the text. I'm guessing that judicious playing with
> \topmargin, \headsep, and such will fix that, but if anybody has suggestions
> or warnings I'd be most appreciative.
>
> The big problem is that the first page does not have a header, but a footer.
> [...]

Use fancyhdr.sty, it will make your life much easier.  To do footers on the
first page and headers on the rest, you can use the "fancyplain" feature: this
lets you redefine the "plain" style to include a footer, and you activate it
with \thispagestyle.

I will email you separately a document class I wrote that produces headings in
the same layout; if you have trouble adapting them, let me know.

Alexis

18th July 2000: Re: Headings

From: alexis@babel.ling.upenn.edu (Alexis Dimitriadis)
Subject: Re: Headings

On second thought, here's just the commands needed to define the headings.
They use fancyheadings.sty, actually an earlier version of fancyhdr.sty,
so there are simpler ways to do the same; but this works anyway.

This gets you a footer on the first page, and headers with the page number,
author and title on the rest.  (You get to provide the commands that define
author and title...)

\fancyplain's first argument is active in plain style; its second argument
is active in fancy style.

- -----------------
\input{fancyheadings.sty}

\headrulewidth=0pt

%      Even (right) pages                       Odd (left) pages

\lhead [\fancyplain{}{\thepage}]                 {}
\rhead []                                        {\fancyplain{}{\thepage}}
\chead [\fancyplain{}{\def\and{\& }\@pgauthor}]  {\fancyplain{}{\@pgtitle}}

\lfoot{}
\cfoot{}
% \rfoot is only defined for odd pages, since we always begin on an odd page:
\rfoot[]{\fancyplain{\it \FirstFooter}{}}

\def\FirstFooter{U. Penn Working Papers in Linguistics,
        Volume 7.1, 2000}

\pagestyle{fancyplain}
% You'll also need to specify \thispagestyle{plain} on the first page, unless
% you use \maketitle (which issues this automatically).
- ----------------

Just in case you still have trouble getting header and footer margins to work
right:  Latex computes these things when it loads the document class; if you
want the header to be higher, you must adjust both \topmargin and \headsep,
otherwise your text will not begin where it is supposed to!  (There is no
similar problem with footers).

Here's a command you can use to raise the header:

% change \topmargin, \headsep in tandem, or the top margin will move!
\def\raiseheader#1{\@tempdima=#1\relax
  \advance\headsep by \@tempdima \advance\topmargin by -\@tempdima \relax}

% Example:
% \raiseheader{-0.08in}
% \advance \footskip by -0.08in
 
Alexis

18th July 2000: Re: Headings

From: "Todd O'Bryan" 
Subject: Headings

Thanks to all who gave me advice and help on the headings problem.

We moved the definition of our own pagestyle into a .sty file and everything
seems to be working. I don't know if commands with @ in them are trouble
outside of .sty files or what, but that seems to have solved the problem.

Todd

19th July 2000: Bibliography indenting

From: "Todd O'Bryan" 
Subject: Bibliography indenting

Hey y'all,

How does one set the left indent of all lines except the first in a
bibliography entry to 0.5in? The author used BibTeX, so I assume I just have
to change a value or two, but don't know where to change them. Currently,
the indent is more like 0.25in.

Thanks much,

Todd

27th July 2000: TIPA and PK files

From: "Sandra P. Kirkham" 
Subject: TIPA and Pk files

Hello All!
Has anyone successfully loaded tipa on Scientific Workplace? I've managed
to load everything except the Pk files, which, incidently, are not
available on the net in a precompiled format for my printer.
Questions:
1) Are there precompiled pk files available anywhere for the Brother HL-760
series(600dpi) printer and where are they located?
2) Failing the availability of the precompiled files, has anyone run
Metafont using Scientific Workplace and how do you do it? 
3) Failing running Metafont on Scientific Workplace, has anyone run
Metafont to create PK files using the DOS shell in Windows-NT and how do
you do it?
4) If the latter two options are possible, which version of Metafont is
used and what is the site location for downloading it?  

Thanks much,

Sandra Patricia Kirkham
Department of Linguistics,
University of Victoria

30th July 2000: Graphics

From: Mike Hammond 
Subject: graphics

all

Hi. I've been trying to embed graphics in a latex document and things are 
behaving rather strangely. What i want is to have the embedded graphic 
display in a dvi file, but also display appropriately when converted to ps 
(dvips) or pdf (dvipdfm).

If I use graphicx, i can embed an eps file and get it to display in a dvi, 
but not convert to ps or pdf. If I convert the graphic to a jpg, then I can 
get it to display in pdf (with dvipdfm again), but not in the dvi or ps 
file.

the eps code:

	\usepackage{graphicx}
	...
	\includegraphics[width=4in]{jl-edit.eps}

the jpg code (using ebb):

	\usepackage[dvipdfm]{graphicx}
	...
	\fig{\includegraphics{jl-edit2.jpg}}

the jpg code (doing ebb offline):

	\usepackage[dvipdfm]{graphicx}
	...
	\includegraphics[bb = 0 0 288 190]{jl-edit2.jpg}

help!

(and incidently, i'm doing this all with miktex under windows; i can't seem 
to find an equivalent to dvipdfm in tetex....)

mike h

31st July 2000: TeX for Arabic

From: Christina Thiele 
Subject: TeX for Arabic

Brigitte, I hope this yields some answers for you ...  If you want
more info on the ling-tex list, you can find it at:

      http://www.ifi.uio.no/~dag/ling-tex.html

====

Fellow ling-tex'ers ... 

I have received a query from Brigitte Dorange-Naulevade 

      dorange@cybercable.fr

asking the following:

> Hello,
> I am looking for informations about using TeX in Arabic.
> ...

I think the quickest response will come if I post this to ling-tex,
cc'd to Brigitte. Please note -- SHE IS NOT ON ling-tex.

If you have information on the subject, please post directly to her,
or make sure that your reply to ling-tex is cc'd to her. 

Ch.

31st July 2000: Re: TeX for Arabic

From: Anshuman Pandey 
Subject: Re: TeX for Arabic

Hello Brigitte,

In my opinion the best TeX package for Arabic is ArabTeX by Klaus 
Lagally. It's available from CTAN/nonfree/language/arabtex/ or directly 
from the author's ftp site at ftp.informatik.uni-stuttgart.de/pub/arabtex.
It functions with both plain TeX and LaTeX. The transliteration scheme
closely follows that of the ZDMG (Zeitschrift der Morgenlaendischen
Geselschaft). The user manual is also very useful. ArabTeX supports Hebrew
as well.

Regards,
Anshuman Pandey

31st July 2000: Re: TeX for Arabic

From: Michel Goossens 
Subject: Re: TeX for Arabic

The best package is arabtex of Prof. lagally.
On CTAN in tex-archive/nonfree/language/arabtex
Can use many encodings, in partucular UTF8 (Unicode). m

31st July 2000: pdfLaTeX

From: Joost Kremers 
Subject: pdflatex

hi all,

i wonder if anyone can help me with the following. until now, i only used
latex and dvips to create my documents. the other day, i tried to create a
.pdf file from a latex source, but i encountered a problem. i use the
tree-building package qtree, and i found that pdflatex does not handle the
lines connecting the nodes correctly. all diagonal lines are absent, and
the vertical lines vary in thickness.

on windows i can use dvipdfm to create a .pdf file from a .dvi file, but i
do most of my work on linux, (teTeX), and that distribution seems to lack
something like dvipdfm. so does anyone know how to make pdflatex work for
qtree? (qtree uses epic.sty and eepic.sty to draw the lines, so i guess
that's where the problem is.)

thanks in advance,

joost kremers

--------------------------------------------------------
Joost Kremers (Mr.)
University of Nijmegen - The Netherlands
Department of Languages and Cultures of the Middle-East
PO Box 9103
6500 HD Nijmegen - The Netherlands
phone: +31 24 3612996
fax: +31 24 3611972

31st July 2000: Re: pdfLaTeX

From: Jean-Pierre A Koenig 
Subject: Re: pdflatex

You can use ps2pdf which is present on most Linux distributions and is
also available on Solaris. As it names indicate it creates a .pdf file
from a .ps file and I have found to be reliable with all kinds of
graphics.

Best,
Jp
______________________________________________________________________
______________________________________________________________________

Jean-Pierre Koenig
Assistant Professor
Linguistics Dept.
609 Baldy Hall
SUNY at Buffalo
Buffalo, NY, 14260-1030
Phone: (716) 645-2177 (ext. 717)
Fax: (716) 645-3825
jpkoenig@acsu.buffalo.edu
http://wings.buffalo.edu/linguistics/koenig/index.html

Last updated 2000/12/22 by Dag Langmyhr.