[Dev-luatex] Warning re comparing Word and TeX

c.a.rowley at open.ac.uk c.a.rowley at open.ac.uk
Mon Nov 2 13:09:32 CET 2009


Hi guys

Do not forget that Word etc have a different model for microtypography and how they treat font metrics and 'normal horizontal stacking' of glyph sequences.  

The big difference that has a major effect on how math layout is acheived is that they do not use the concept of nested boxes: everything is simply 'positioned glyphs'.

There are other important differences but they are not so likely to affect math typesetting.

If you get Windows 2010, I think that all the math stuff is available 'at the OS level' which I think means that it is also all in an Active-X control for developers use.  But probably with zero documentattion!

Finally, if you want to know about their algorithm, just ask them!
Although they may reply, as I would about details like that, 'we have forgotten'.

Cheers,  chris

>
> Office 2007 seems to handle sub/superscripts of big operators
> differently; superscript is placed immediately after base glyph while a
> negative kern equal to italic correction is inserted between base and
> subscript i.e. the reverse of what is described in the spec, and math
> kerning is ignored completely.

Big sigh. I'll see if I can get my hands on a copy of Word2009.

Best wishes,
Taco


-----dev-luatex-bounces at ntg.nl wrote: -----

To: dev-luatex at ntg.nl
From: dev-luatex-request at ntg.nl
Sent by: dev-luatex-bounces at ntg.nl
Date: 02/11/2009 11:00
Subject: dev-luatex Digest, Vol 46, Issue 2

Send dev-luatex mailing list submissions to
    	dev-luatex at ntg.nl

To subscribe or unsubscribe via the World Wide Web, visit
    	http://www.ntg.nl/mailman/listinfo/dev-luatex
or, via email, send a message with subject or body 'help' to
    	dev-luatex-request at ntg.nl

You can reach the person managing the list at
    	dev-luatex-owner at ntg.nl

When replying, please edit your Subject line so it is more specific
than "Re: Contents of dev-luatex digest..."


Today's Topics:

   1. Re: Luatex 0.44.0 announcement (Khaled Hosny)
   2. echoing system call argument to stderr (T T)
   3. Re: echoing system call argument to stderr
      (Manuel P?gouri?-Gonnard)
   4. Re: kpse.find_file (Reinhard Kotucha)
   5. Re: echoing system call argument to stderr (T T)
   6. Re: echoing system call argument to stderr (Akira Kakuto)
   7. Re: Luatex 0.44.0 announcement (Taco Hoekwater)


----------------------------------------------------------------------

Message: 1
Date: Sun, 1 Nov 2009 15:00:25 +0200
From: Khaled Hosny <khaledhosny at eglug.org>
To: Taco Hoekwater <taco at elvenkind.com>
Cc: luatex development list <dev-luatex at ntg.nl>
Subject: Re: [Dev-luatex] Luatex 0.44.0 announcement
Message-ID: <20091101130025.GA8814 at khaled-laptop>
Content-Type: text/plain; charset="utf-8"

On Fri, Oct 30, 2009 at 09:02:30AM +0100, Taco Hoekwater wrote:
> 
> Hi,
> 
> Khaled Hosny wrote:
> > On Tue, Oct 27, 2009 at 12:06:09PM +0100, Taco Hoekwater wrote:
> >> * Math delimiters and large operators in new MATH fonts no
> >>   longer apply the italic correction (this fixes placements
> >>   of super- and subscripts on such items for those fonts).
> > 
> > This doesn't sound the correct way to me. The described algorithm in
> > MATH specification is to set subscripts immediately after the base
> > (without italic correction) and shift superscripts horizontally by the
> > amount of italic correction, then apply math kerning if present, this
> > way fonts with no math kerning will still give acceptable output and
> > need for math kerning will be minimized.
> 
> Well, I am clearly doing something wrong, because both the old and
> new outputs look crappy. I'll rethink this.

Office 2007 seems to handle sub/superscripts of big operators
differently; superscript is placed immediately after base glyph while a
negative kern equal to italic correction is inserted between base and
subscript i.e. the reverse of what is described in the spec, and math
kerning is ignored completely.
So I think the current positioning of superscripts is correct, only
subscripts need to be fixed, though I'm not really sure about any of
those conclusions :)

Regards,
 Khaled

-- 
 Khaled Hosny
 Arabic localiser and member of Arabeyes.org team
 Free font developer
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
URL: <http://www.ntg.nl/mailman/private/dev-luatex/attachments/20091101/225a7e11/attachment-0001.pgp>

------------------------------

Message: 2
Date: Sun, 1 Nov 2009 20:11:11 +0000
From: T T <t34www at googlemail.com>
To: luatex development list <dev-luatex at ntg.nl>
Subject: [Dev-luatex] echoing system call argument to stderr
Message-ID:
    	<e7fba2610911011211w39d3824dud58416c1d8d95498 at mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1

Hi,

I've noticed that os.execute under texlua echos its argument to
stderr. This is not the behaviour under the standard lua interpreter.
Is it a bug or feature (and what would be the reason for it)?

Cheers,

Tomek


------------------------------

Message: 3
Date: Sun, 01 Nov 2009 22:02:45 +0100
From: Manuel P?gouri?-Gonnard <mpg at elzevir.fr>
To: T T <t34www at googlemail.com>
Cc: luatex development list <dev-luatex at ntg.nl>
Subject: Re: [Dev-luatex] echoing system call argument to stderr
Message-ID: <4AEDF775.90905 at elzevir.fr>
Content-Type: text/plain; charset=ISO-8859-1

Hi,

T T a ?crit :
> I've noticed that os.execute under texlua echos its argument to
> stderr. This is not the behaviour under the standard lua interpreter.
> Is it a bug or feature (and what would be the reason for it)?
> 
Which version of texlua? I don't observe this behaviour with texlua
0.40.6 from TL09 under x86-64-linux (nor with previous versions).

Manuel.


------------------------------

Message: 4
Date: Sun, 1 Nov 2009 22:03:37 +0100
From: Reinhard Kotucha <reinhard.kotucha at web.de>
To: Manuel P?gouri?-Gonnard <mpg at elzevir.fr>
Cc: tldev at tug.org, dev-luatex at ntg.nl
Subject: Re: [Dev-luatex] kpse.find_file
Message-ID: <19181.63401.564361.794890 at zaphod.ms25.net>
Content-Type: text/plain; charset=iso-8859-1

On 1 November 2009 Manuel P?gouri?-Gonnard wrote:

 > Hi Reinhard,
 > 
 > Reinhard Kotucha a ?crit :
 > >    kpse.set_program_name('ppower4')
 > 
 > The problem is that kpse resolves symlinks before computing SELFAUTO*
 > (which is what you want in most cases). So SELFAUTOLOC ends up being
 > texmf-dist/scripts/ppoxer4 which obviously gives the wrong result.
 >
 > A general approach to this is to use
 > 
 > kpse.set_program_name(arg[-1], 'ppower4')
 > 
 > which works on all platforms (I'm using it in texdoc and never got any
 > complaint).  As you probably know, arg[-1] is the name of the Lua
 > interpreter used for the current script, which gives good results for
 > computing SELFAUTO*. I guess that's the whole purpose of the
 > two-arguments form of kpse.set_program_name.

Hi Manuel,
thank you very much for the explanation.  Yes, it's a very good
solution, maybe the best. I fixed it this way.
 
 > (Btw, at some point there was a bug in kpse so that the first argument
 > was ignored on GNU platforms and SELFAUTO* set to "correct" values even
 > if the argument was "wrong" on platform using the GNU libc. So this kind
 > of problem was hidden on, eg Linux, when this bug was present. After the
 > bug was fixed, I looked a bit for othe occurences in TeX Live, but
 > apparently I missed some of them.)

When I checked the stuff in texmf/scripts, I obviously didn't remember
that ppower4 was moved to texmf-dist/scripts.  Maybe you overlooked it
for the same reason.

Regards, and thanks again,
  Reinhard

-- 
----------------------------------------------------------------------------
Reinhard Kotucha    	    	    	              Phone: +49-511-3373112
Marschnerstr. 25
D-30167 Hannover    	                      mailto:reinhard.kotucha at web.de
----------------------------------------------------------------------------
Microsoft isn't the answer. Microsoft is the question, and the answer is NO.
----------------------------------------------------------------------------


------------------------------

Message: 5
Date: Sun, 1 Nov 2009 22:11:58 +0000
From: T T <t34www at googlemail.com>
To: Manuel P?gouri?-Gonnard <mpg at elzevir.fr>
Cc: luatex development list <dev-luatex at ntg.nl>,    	Akira Kakuto
    	<kakuto at fuk.kindai.ac.jp>
Subject: Re: [Dev-luatex] echoing system call argument to stderr
Message-ID:
    	<e7fba2610911011411g5f0fb864jd0a08bf2c25c8cb1 at mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1

2009/11/1 Manuel P?gouri?-Gonnard <mpg at elzevir.fr>:
> Hi,
>
> T T a ?crit :
>> I've noticed that os.execute under texlua echos its argument to
>> stderr. This is not the behaviour under the standard lua interpreter.
>> Is it a bug or feature (and what would be the reason for it)?
>>
> Which version of texlua? I don't observe this behaviour with texlua
> 0.40.6 from TL09 under x86-64-linux (nor with previous versions).

Version 0.40.6-2009103121 but maybe older ones as well. Perhaps it's
only win32. In that case CC Akira.

Cheers,

Tomek


------------------------------

Message: 6
Date: Mon, 02 Nov 2009 11:27:25 +0900
From: "Akira Kakuto" <kakuto at fuk.kindai.ac.jp>
To: T T <t34www at googlemail.com>
Cc: Manuel_P?ouri?Gonnard     	<mpg at elzevir.fr>, luatex development list
    	<dev-luatex at ntg.nl>
Subject: Re: [Dev-luatex] echoing system call argument to stderr
Message-ID: <125712884643853100006a7b at sniffer>
Content-Type: text/plain; charset=iso-2022-jp

Hi Tomek,

> >> I've noticed that os.execute under texlua echos its argument to
> >> stderr. This is not the behaviour under the standard lua interpreter.
> >> Is it a bug or feature (and what would be the reason for it)?
> >>
> > Which version of texlua? I don't observe this behaviour with texlua
> > 0.40.6 from TL09 under x86-64-linux (nor with previous versions).
> 
> Version 0.40.6-2009103121 but maybe older ones as well. Perhaps it's
> only win32. In that case CC Akira.

Please update the kpathsea dll to the latest version.
In some version, I fogot to erase a line
for debugging purpose.



------------------------------

Message: 7
Date: Mon, 02 Nov 2009 08:36:21 +0100
From: Taco Hoekwater <taco at elvenkind.com>
To: Khaled Hosny <khaledhosny at eglug.org>
Cc: luatex development list <dev-luatex at ntg.nl>
Subject: Re: [Dev-luatex] Luatex 0.44.0 announcement
Message-ID: <4AEE8BF5.2040903 at elvenkind.com>
Content-Type: text/plain; charset=UTF-8



Khaled Hosny wrote:
> 
> Office 2007 seems to handle sub/superscripts of big operators
> differently; superscript is placed immediately after base glyph while a
> negative kern equal to italic correction is inserted between base and
> subscript i.e. the reverse of what is described in the spec, and math
> kerning is ignored completely.

Big sigh. I'll see if I can get my hands on a copy of Word2009.

Best wishes,
Taco


------------------------------

_______________________________________________
dev-luatex mailing list
dev-luatex at ntg.nl
http://www.ntg.nl/mailman/listinfo/dev-luatex


End of dev-luatex Digest, Vol 46, Issue 2
*****************************************

---------------------------------------------------------------------------
The Open University is incorporated by Royal Charter (RC 000391), an exempt charity in England & Wales and a charity registered in Scotland (SC 038302)


More information about the dev-luatex mailing list