February 2007 - ntg-pdftex - NTG Mailing Lists

[ pdftex-Bugs-736 ] font expansion leads to wrong log output of \kerns.
by noreply＠sarovar.org 22 Feb '07

22 Feb '07

Bugs item #736, was opened at 2007-02-22 03:56 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106 Category: hz Group: v1.40.1 Status: Open >Resolution: Accepted >Priority: 3 Submitted By: Robert (schlcht) Assigned to: The Thanh Han (hanthethanh) Summary: font expansion leads to wrong log output of \kerns. Initial Comment: When applying font expansion, implicit kerns will be output *with* a space in the log file, which makes them appear like explicit kerns. The following \tracingoutput1 \pdfadjustspacing2 \pdffontexpand\font 20 20 5 autoexpand bye\bye will yield ...\tenrm b ...\kern -0.27779 ...\tenrm y ...\kern -0.27779 ...etc. instead of ...\tenrm b ...\kern-0.27779 ...\tenrm y ...\kern-0.27779 ...etc. ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2007-02-22 11:46 Message: Logged In: YES user_id=710 I confirm the bug. Will try to fix it soon. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106

1 0

[ pdftex-Bugs-736 ] font expansion leads to wrong log output of \kerns.
by noreply＠sarovar.org 22 Feb '07

22 Feb '07

Bugs item #736, was opened at 2007-02-22 02:56 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106 Category: hz Group: v1.40.1 Status: Open Resolution: None Priority: 5 Submitted By: Robert (schlcht) Assigned to: The Thanh Han (hanthethanh) Summary: font expansion leads to wrong log output of \kerns. Initial Comment: When applying font expansion, implicit kerns will be output *with* a space in the log file, which makes them appear like explicit kerns. The following \tracingoutput1 \pdfadjustspacing2 \pdffontexpand\font 20 20 5 autoexpand bye\bye will yield ...\tenrm b ...\kern -0.27779 ...\tenrm y ...\kern -0.27779 ...etc. instead of ...\tenrm b ...\kern-0.27779 ...\tenrm y ...\kern-0.27779 ...etc. ---------------------------------------------------------------------- >Comment By: The Thanh Han (hanthethanh) Date: 2007-02-22 10:46 Message: Logged In: YES user_id=710 I confirm the bug. Will try to fix it soon. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106

1 0

[ pdftex-Bugs-712 ] problems with \letterspacefont
by noreply＠sarovar.org 22 Feb '07

22 Feb '07

Bugs item #712, was opened at 2006-12-12 02:53 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=712&group_id=106 Category: hz Group: v1.40.0 Status: Closed Resolution: Fixed Priority: 5 Submitted By: Robert (schlcht) Assigned to: The Thanh Han (hanthethanh) Summary: problems with \letterspacefont Initial Comment: % Testing the new \letterspacefont, I found: % % (1) it's certainly not correct to have the spacing % relative to character widths (look at `minimal'!) % (2) everything's shifted to the right % (3) when trying to expand the letter-spaced font, I % get the warning `font should be expanded before % its first use' (even before its first use), and % then pdftex segfaults. \pdfadjustspacing2 \hsize13cm This is normal text This is normal text This is normal text This is normal text This is normal text This is normal text {\letterspacefont\lspfont=\font 1000 %\pdffontexpand\lspfont 20 20 5 autoexpand % segfault \lspfont This isn't normal text This isn't normal text This isn't normal text This is minimal text} This is normal text This is normal text This is normal text \bye ---------------------------------------------------------------------- >Comment By: The Thanh Han (hanthethanh) Date: 2007-02-22 09:31 Message: Logged In: YES user_id=710 yes it sounds reasonable. Or we can extend the syntax of \letterspacefont so that if an optional keyword like 'nolig' is found then ligatures will be disabled. ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2007-02-22 02:54 Message: Logged In: YES user_id=2217 The current situation with regards to the lack of kerning of letterspaced fonts is not really satisfying. After some consideration, I therefore think it is actually better *not* to disable ligatures, and hence kerning, for letterspaced fonts. So I would propose to remove the line: set_no_ligatures(k); {disable ligatures for letter-spaced fonts} It is easy enough for the high-level interfaces (microtype, context) to (automatically) disable selected ligatures only, so that kerning for the rest of the characters is still available. What do you think? \letterspacefont\lsfont\font 100 \lsfont \tagcode\lsfont`f -1 VALID off -- \bye ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2007-02-02 13:54 Message: Logged In: YES user_id=710 the warning message can be removed. I left it there since I don't know what happens if a virtual font is loaded, used and then expanded and used again with font expansion. It should be alright to do so, but I am not very sure. Anyway, I will change the warning to be issued for non-virtual fonts only. It turned to be a design mistake that the \pdffontexpand primitive takes a base font and turns it to an expanded font. It would be better it was more like \letterspacefont, ie it should create a new font and leave the original font intact. ---------------------------------------------------------------------- Comment By: Taco Hoekwater (taco) Date: 2007-02-02 07:30 Message: Logged In: YES user_id=1608 > I hope luatex will abolish with this nonsensical connexion of ligatures and kerns? Yes. ligatures and kerns are totally separate in luatex. ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2007-02-01 19:02 Message: Logged In: YES user_id=2217 If the warning is misleading, should it not be removed? On a different matter: Since \letterspacefont issues set_no_ligatures, there won't be any kerning, either. There's no way around that, I guess? (I hope luatex will abolish with this nonsensical connexion of ligatures and kerns?) ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2007-01-04 15:59 Message: Logged In: YES user_id=710 The warning is misleading but harmless. The reason of the warning is that pdftex checks whether a font has not been used before expansion (its type has not been examined and is new/unknown yet). \letterspacefont sets the type of \y as a virtual font, thus pdftex thinks the font might have been used before. ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2007-01-04 00:40 Message: Logged In: YES user_id=2217 I'm a bit confused about the status of (3): letterspacing and expansion: When expanding a letterspaced font, pdftex issues the warning: ``font should be expanded before its first use'', but apart from that, I cannot see any problems -- the font is both letterspaced and expanded. Or is there a reason for the warning? \pdfadjustspacing2 \font\x=cmr10 \x \letterspacefont\y=\x 100 \y \pdffontexpand\y 20 20 5 autoexpand ... \bye ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2006-12-21 16:09 Message: Logged In: YES user_id=421 A patch for this will be included in the next version (1.40). ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2006-12-12 17:47 Message: Logged In: YES user_id=710 oops, I disabled ligature for letter-spaced font, but this should not affect the base font. Will have to rethink. Most likely the syntax will change like follows: \letterspacefont\f=cmr10 at 11pt 100 \f text ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2006-12-12 15:50 Message: Logged In: YES user_id=2217 % and here's % (4) ligatures will be broken up, not only in the % letter-spaced font (which is probably acceptable), % but in all subsequent uses of the base font, too fi\par {\letterspacefont\lspfont=\font 100 \lspfont fi}\par fi \bye ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2006-12-12 13:31 Message: Logged In: YES user_id=710 I fixed (1) and (2); (3) is disabled now, ie you cannot expand a letter-spaced font. Please wait for another release from Martin. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=712&group_id=106

1 0

[ pdftex-Bugs-736 ] font expansion leads to wrong log output of \kerns.
by noreply＠sarovar.org 22 Feb '07

22 Feb '07

Bugs item #736, was opened at 2007-02-22 02:56 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106 Category: hz Group: v1.40.1 Status: Open Resolution: None Priority: 5 Submitted By: Robert (schlcht) Assigned to: The Thanh Han (hanthethanh) Summary: font expansion leads to wrong log output of \kerns. Initial Comment: When applying font expansion, implicit kerns will be output *with* a space in the log file, which makes them appear like explicit kerns. The following \tracingoutput1 \pdfadjustspacing2 \pdffontexpand\font 20 20 5 autoexpand bye\bye will yield ...\tenrm b ...\kern -0.27779 ...\tenrm y ...\kern -0.27779 ...etc. instead of ...\tenrm b ...\kern-0.27779 ...\tenrm y ...\kern-0.27779 ...etc. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=736&group_id=106

1 0

[ pdftex-Bugs-712 ] problems with \letterspacefont
by noreply＠sarovar.org 22 Feb '07

22 Feb '07

Bugs item #712, was opened at 2006-12-12 02:53 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=712&group_id=106 Category: hz Group: v1.40.0 Status: Closed Resolution: Fixed Priority: 5 Submitted By: Robert (schlcht) Assigned to: The Thanh Han (hanthethanh) Summary: problems with \letterspacefont Initial Comment: % Testing the new \letterspacefont, I found: % % (1) it's certainly not correct to have the spacing % relative to character widths (look at `minimal'!) % (2) everything's shifted to the right % (3) when trying to expand the letter-spaced font, I % get the warning `font should be expanded before % its first use' (even before its first use), and % then pdftex segfaults. \pdfadjustspacing2 \hsize13cm This is normal text This is normal text This is normal text This is normal text This is normal text This is normal text {\letterspacefont\lspfont=\font 1000 %\pdffontexpand\lspfont 20 20 5 autoexpand % segfault \lspfont This isn't normal text This isn't normal text This isn't normal text This is minimal text} This is normal text This is normal text This is normal text \bye ---------------------------------------------------------------------- >Comment By: Robert (schlcht) Date: 2007-02-22 02:54 Message: Logged In: YES user_id=2217 The current situation with regards to the lack of kerning of letterspaced fonts is not really satisfying. After some consideration, I therefore think it is actually better *not* to disable ligatures, and hence kerning, for letterspaced fonts. So I would propose to remove the line: set_no_ligatures(k); {disable ligatures for letter-spaced fonts} It is easy enough for the high-level interfaces (microtype, context) to (automatically) disable selected ligatures only, so that kerning for the rest of the characters is still available. What do you think? \letterspacefont\lsfont\font 100 \lsfont \tagcode\lsfont`f -1 VALID off -- \bye ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2007-02-02 13:54 Message: Logged In: YES user_id=710 the warning message can be removed. I left it there since I don't know what happens if a virtual font is loaded, used and then expanded and used again with font expansion. It should be alright to do so, but I am not very sure. Anyway, I will change the warning to be issued for non-virtual fonts only. It turned to be a design mistake that the \pdffontexpand primitive takes a base font and turns it to an expanded font. It would be better it was more like \letterspacefont, ie it should create a new font and leave the original font intact. ---------------------------------------------------------------------- Comment By: Taco Hoekwater (taco) Date: 2007-02-02 07:30 Message: Logged In: YES user_id=1608 > I hope luatex will abolish with this nonsensical connexion of ligatures and kerns? Yes. ligatures and kerns are totally separate in luatex. ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2007-02-01 19:02 Message: Logged In: YES user_id=2217 If the warning is misleading, should it not be removed? On a different matter: Since \letterspacefont issues set_no_ligatures, there won't be any kerning, either. There's no way around that, I guess? (I hope luatex will abolish with this nonsensical connexion of ligatures and kerns?) ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2007-01-04 15:59 Message: Logged In: YES user_id=710 The warning is misleading but harmless. The reason of the warning is that pdftex checks whether a font has not been used before expansion (its type has not been examined and is new/unknown yet). \letterspacefont sets the type of \y as a virtual font, thus pdftex thinks the font might have been used before. ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2007-01-04 00:40 Message: Logged In: YES user_id=2217 I'm a bit confused about the status of (3): letterspacing and expansion: When expanding a letterspaced font, pdftex issues the warning: ``font should be expanded before its first use'', but apart from that, I cannot see any problems -- the font is both letterspaced and expanded. Or is there a reason for the warning? \pdfadjustspacing2 \font\x=cmr10 \x \letterspacefont\y=\x 100 \y \pdffontexpand\y 20 20 5 autoexpand ... \bye ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2006-12-21 16:09 Message: Logged In: YES user_id=421 A patch for this will be included in the next version (1.40). ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2006-12-12 17:47 Message: Logged In: YES user_id=710 oops, I disabled ligature for letter-spaced font, but this should not affect the base font. Will have to rethink. Most likely the syntax will change like follows: \letterspacefont\f=cmr10 at 11pt 100 \f text ---------------------------------------------------------------------- Comment By: Robert (schlcht) Date: 2006-12-12 15:50 Message: Logged In: YES user_id=2217 % and here's % (4) ligatures will be broken up, not only in the % letter-spaced font (which is probably acceptable), % but in all subsequent uses of the base font, too fi\par {\letterspacefont\lspfont=\font 100 \lspfont fi}\par fi \bye ---------------------------------------------------------------------- Comment By: The Thanh Han (hanthethanh) Date: 2006-12-12 13:31 Message: Logged In: YES user_id=710 I fixed (1) and (2); (3) is disabled now, ie you cannot expand a letter-spaced font. Please wait for another release from Martin. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=712&group_id=106

1 0

[ pdftex-Patches-735 ] patch for simplified horizontal positional tracking
by noreply＠sarovar.org 19 Feb '07

19 Feb '07

Patches item #735, was opened at 2007-02-19 22:00 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=495&aid=735&group_id=106 Category: Positioning Group: v1.40.0 Status: Open Resolution: None Priority: 5 Submitted By: Hartmut Henkel (hhenkel) Assigned to: Nobody (None) Summary: patch for simplified horizontal positional tracking Initial Comment: 1. simpler calculations for horizontal position update when inside []TJ arrays, now by pure additions within a fontsize/10000 raster; no rounding errors can accumulate. Gives about 7 % pdftex speed gain when typesetting fiction-type stuff. 2. writet3.c now also with standard /FontMatrix [0.001 0 0 0.001 0 0], tidied up a little (.pgc part untested). No need anymore (hope it's right) to handle PK fonts specially by adv_char_width() in pdftex.web. 3. wrong use of printf() at a few places corrected by Martin. Work in progess, not safe for production. Needs more testing, particularly with PK fonts. Regards, Hartmut ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=495&aid=735&group_id=106

1 0

[ pdftex-Bugs-733 ] CVE-2007-0104
by noreply＠sarovar.org 16 Feb '07

16 Feb '07

Bugs item #733, was opened at 2007-02-16 15:37 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=733&group_id=106 Category: xpdf Group: v1.40.0 >Status: Closed >Resolution: Fixed Priority: 3 Submitted By: Martin Schröder (oneiros) Assigned to: Martin Schröder (oneiros) Summary: CVE-2007-0104 Initial Comment: Another bug in xpdf was found: http://nvd.nist.gov/nvd.cfm?cvename=CVE-2007-0104 http://projects.info-pull.com/moab/MOAB-06-01-2007.html As pdftex uses xpdf, this can be used to crash pdftex. A patch for this is attached and will be included in the next release. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=733&group_id=106

1 0

[ pdftex-Bugs-733 ] CVE-2007-0104
by noreply＠sarovar.org 16 Feb '07

16 Feb '07

Bugs item #733, was opened at 2007-02-16 15:37 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=733&group_id=106 Category: xpdf Group: v1.40.0 Status: Open Resolution: None Priority: 3 Submitted By: Martin Schröder (oneiros) Assigned to: Martin Schröder (oneiros) Summary: CVE-2007-0104 Initial Comment: Another bug in xpdf was found: http://nvd.nist.gov/nvd.cfm?cvename=CVE-2007-0104 http://projects.info-pull.com/moab/MOAB-06-01-2007.html As pdftex uses xpdf, this can be used to crash pdftex. A patch for this is attached and will be included in the next release. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=733&group_id=106

1 0

runptex API
by Thanh Han The 13 Feb '07

13 Feb '07

Hi, I have put the documented API of runptex, a wrapper to run pdftex, to http://runptex.sarovar.org/ Comments welcome very much. Source codes, installation and usage guide will be added soon. Thanh

1 0

Re: [NTG-pdftex] another problem -- of pdftex or gs?
by Thanh Han The 11 Feb '07

11 Feb '07

On Sat, Feb 10, 2007 at 10:52:09PM +0100, Thanh Han The wrote: > yes I also got very varying results from gs, depending on > version, configuration and also the output device. > However, I know of another way to check whether a ps file is > valid. sorry, typo. I *don't* know of another way to check ps files. Thanh

1 0