- dev-context - NTG Mailing Lists

MetaPost 1.803 announcement
by Taco Hoekwater 16 Jun '13

16 Jun '13

I am happy to announce a new release of MetaPost: ---------------------------------------------------------------------- MetaPost 1.803 ---------------------------------------------------------------------- This is a bugfix release for 1.802. The win32 binary as well as a source archive can be downloaded immediately from the files section at https://foundry.supelec.fr/projects/metapost/ If you want to build MetaPost yourself, you can also do an anonymous subversion checkout: $ svn checkout --username anonsvn https://foundry.supelec.fr/svn/metapost/tags/1.803 The password is anonsvn ---------------------------------------------------------------------- What is new in MetaPost version 1.803: ---------------------------------------------------------------------- * Fixed a bug in the path tension handling that sometimes caused distorted output. * A number of small alterations to the build system making it easier to integrate metapost with W32TEX. * Fix the internal rounding functions in --numbersystem=double so that they round exactly like in --numbersystem=scaled mode. * From Elie Roux: a fix to graph.mp to prevent log-scaled graphs with large ranges from causing the "Value is too large" error. Have fun! Taco Hoekwater, for the MetaPost team, April 5, 2013.

2 1

type interference with \placefigure
by Mojca Miklavec 19 Apr '13

19 Apr '13

Hello, maybe this has never worked, but \placefigure[force]{Second page from \type{fullpage-example}, included with \type+\externalfigure+}{\externalfigure[fullpage-example][page=2]} fails in MKII with figures : dimensions of fullpage-example loaded from figurefile itself (./gnuplot-context-doc.tuo) ! Missing } inserted. <inserted text> } <to be read again> \endgroup \putcompletecaption ...trut #2\endstrut \endgraf } \fi \dostopattributes \doi... \docheckcaptioncontent ...completecaption {#1}{#2} }\ifdim \wd \tempcaptionbo... \dosetpagfloat ... \docheckcaptioncontent {#2}{#3} \ifcase \floatparameter \c... \dosetfloatbox ... \dosetpagfloat \fi {#1}{#2}{#3} \setlocalfloatdimensions {... ... l.195 ...externalfigure[fullpage-example][page=2]} ? I'm not asking to fix it, but I'm unable to compile the document with MKIV. Mojca

2 2

MkIV "isolated expression" with \(start|stop|use)staticMPfigure
by Sanjoy Mahajan 14 Apr '13

14 Apr '13

With the attached test file, which uses MPinclusions for a staticMPfigure, I get an "Isolated expression". I think it's a parser bug somewhere, because deleting almost any word in the file removes the error. The file itself looks a bit crazy. It is a minimal example generated by a program to help with debugging. The program randomly deletes a word in the file, then tests whether the error still happens. If yes, it deletes another word. If no, it restores the word, and tries again by choosing another word to delete. That process reduced the file from 4KB to 2KB. Here's the relevant piece of the transcript. The same error happens at live.contextgarden.net (so I don't think it's a Debian-specific issue). ! terminal: >> r_ladder ! Isolated expression. <to be read again> ; <*> r_ladder; ; . system > tex > error on line 73 in file test-mpfigures.tex: terminal: >> r_ladder ! Isolated expression. <to be read again> ; <*> r_ladder; ; ... 63 64 \startMPinclusions 65 def r_ladder := draw unitsquare; enddef; 66 \stopMPinclusions 67 68 \startstaticMPfigure{resistiveladder} 69 r_ladder; 70 \stopstaticMPfigure 71 72 \usestaticMPfigure[resistiveladder] 73 >> 74 \stoptext 75 <inserted text> ...0;;]===], [===[;;]===], false)} \meta_process_graphic ...!!es , \MPaskedfigure )}} \egroup \placeMPgraphic \m... \meta_handle_reusable_graphic ...cess_graphic {#3} }\setxvalue {\??mpgraphic ... \meta_use_graphic ...aphic \currentMPgraphicname } \empty \meta_end_graphic_g... <to be read again> \fi \dodoubleemptyonespaced #1#2->#1[{#2}][] ... l.73 And the start of the output, showing the version information: mtx-context | run 1: luatex --fmt="/var/lib/texmf/luatex-cache/context/f919750633122078d706dd10f1cfd39b/formats/cont-en" --lua="/var/lib/texmf/luatex-cache/context/f919750633122078d706dd10f1cfd39b/formats/cont-en.lui" --backend="pdf" "./test-mpfigures.tex" \stoptext This is LuaTeX, Version beta-0.70.2-2012070301 (TeX Live 2012/Debian) \write18 enabled. (test-mpfigures.tex ConTeXt ver: 2012.05.30 11:26 MKIV fmt: 2013.4.11 int: english/english system > cont-new.mkiv loaded If you can think of further tests to run, I'll be happy to try them. -- -Sanjoy

2 3

PATCH: fix meta-pdf.lua choking on grayscale metapost output
by Sanjoy Mahajan 11 Apr '13

11 Apr '13

I am converting myself to MkIV, and ran into the following error: ! LuaTeX error /usr/share/texmf/tex/context/base/meta-pdf.lua:355: attempt to call global 'pdfgrayliteral' (a nil value) stack traceback: /usr/share/texmf/tex/context/base/meta-pdf.lua:355: in function </usr/share/texmf/tex/context/base/meta-pdf.lua:353> [C]: in function 'lpegmatch' /usr/share/texmf/tex/context/base/meta-pdf.lua:519: in function 'parse' /usr/share/texmf/tex/context/base/meta-pdf.lua:539: in function 'convertmpstopdf' <main ctx instance>:1: in main chunk. It was during converting metapost to pdf. The metapost source has: defaultcolormodel := 3; % grayscale which may be why this problem wasn't noticed before. But the patch below fixes it for me (after regenerating the format, which took me a while to realize): --- /usr/share/texmf/tex/context/base/meta-pdf.lua 2012-05-14 13:38:55.000000000 -0400 +++ /usr/share/texmf/tex/context/base/meta-pdf.lua 2013-04-11 14:54:45.000000000 -0400 @@ -352,7 +352,7 @@ function mps.setgray(s) pdfcode(pdffinishtransparencycode()) - pdfcode(pdfgrayliteral(mps.colormodel,s)) + pdfcode(pdfgraycode(mps.colormodel,s)) end function mps.specials(version,signal,factor) -- 2.0 123 1000

1 0

math-ali.mkiv
by Wolfgang Schuster 10 Apr '13

10 Apr '13

Hi, the redefinition of \startalignment in math-ali.mkiv doesn’t cover \stopalignment. \ifdefined \startalignment \let\align_math_normal_start\startalign \let\align_math_normal_stop \stopalign \let\align_text_normal_start\startalignment \let\align_text_normal_stop \stopalignment \unexpanded\def\startalign {\ifmmode - \let\stopalign\align_math_normal_stop % cannot be an unexpanded def ... lookahead in align + \let\stopalign \align_math_normal_stop % cannot be an unexpanded def ... lookahead in align + \let\stopalignment\align_math_normal_stop % cannot be an unexpanded def ... lookahead in align \expandafter\align_math_normal_start \else - \let\stopalign\align_text_normal_stop + \let\stopalign \align_text_normal_stop + \let\stopalignment\align_text_normal_stop \expandafter\align_text_normal_start \fi} \let\stopalign \relax + \let\stopalignment\relax \let\startalignment\startalign - \let\stopalignment \stopalign \fi Wolfgang

1 0

Merging font support modules
by Aditya Mahajan 07 Apr '13

07 Apr '13

Hi everyone, I was browsing through the third party modules, and noticed that there are a few packages that provide typescripts for fonts. I think that it will be worthwhile to clean up and merge these modules, so as to reduce the amount of clutter. I am listing all the font related modules below, and asking if these are still needed; or should some of these be moved to an obsolete/ branch or merged with other modules. - Chinese font support (by Wolfgang Schuster; provides fonts, perhaps merge with cwTeX font module, also by Wolfgang) - GFSDidot Opentype fonts (I cannot browse through that! Should this be moved to obsolete branch?) - Greek (by Thomas Schmitz; provides fonts as well as typescripts, so this module cannot be merged) - Storm font support (Again, I cannot browse through that. Should this be moved to obsolete branch?) - Lucida bright font support (by Vit Zyka: do we still need these support files, or can this be moved to obsolete?) - Support for cwTeX fonts (by Wolfgang Schuster provides fonts and typescripts. Can this be merged with Chinese font support?) - Typescript (by Wolfgang Schuster, provides typescripts for a variety of fonts. I think that all modules that only distribute font typescripts should be merged with this one.) - URW Garamond support (by Peter Münster; is this still needed? If so, this should be merged with the typescript module) - URW Gothic support (by Taco Hoekwater; is this still needed? If so, this should be merged with the typescript module). What do others feel about such pruning and merging of packages? Aditya

4 5

Luajittex 0.76.0 announcement
by luigi scarso 06 Apr '13

06 Apr '13

New beta release of luajittex. Site: https://foundry.supelec.fr/projects/luajittex/ svn repo.: svn checkout --username anonsvn https://foundry.supelec.fr/svn/luajittex/tags/beta-0.76.0 -- luigi

1 0

Luatex 0.76.0 announcement
by Taco Hoekwater 05 Apr '13

05 Apr '13

Hi, I have just uploaded the archives for a new luatex release, 0.76.0, which has a few fixes for beta 0.75.0, and synchronizes the code with the development version of TeXLive 2013. Changes: * Metapost 1.801. * Small patches from TeXLive 2013, as well as an update to the latest libraries (libpng, poppler, etc.) * Speed up printing (of TeX strings) to the terminal and log. * New lua function node.end_of_math(). * Fixed a memory leak in the lua function pdfscanner.scan(). The archives can be downloaded from supelec as usual: https://foundry.supelec.fr/projects/luatex/ You could also check out the sources via anonymous svn: svn co --username anonsvn --password anonsvn \ https://foundry.supelec.fr/svn/luatex/tags/beta-0.76.0 Bugs and feature requests can be added to the issue tracker at http://tracker.luatex.org Have fun, Taco

1 0

MetaPost 1.801 announcement
by Taco Hoekwater 05 Apr '13

05 Apr '13

I am happy to announce a new release of MetaPost: ---------------------------------------------------------------------- MetaPost 1.801 ---------------------------------------------------------------------- This is a bugfix release for 1.800. The win32 binary as well as a source archive can be downloaded immediately from the files section at https://foundry.supelec.fr/projects/metapost/ If you want to build MetaPost yourself, you can also do an anonymous subversion checkout: $ svn checkout --username anonsvn https://foundry.supelec.fr/svn/metapost/tags/1.801 The password is anonsvn ---------------------------------------------------------------------- What is new in MetaPost version 1.801: ---------------------------------------------------------------------- * fixed an off-by-one error in the SVG output that prevented the last font's glyphs from actually being output to the svg file under prologues:=3; * fixed a bug in bbox calculations in that caused the max y to be set based on the maximum x coordinate internal value instead of the maximum y coordinate. * fixed a bug where metapost incorrectly produced a 'degenerate spec' error. * build fixes for various platforms imported from TeXLive 2013, as well as the latest kpathsea development etc. Have fun! Taco Hoekwater, for the MetaPost team, April 5, 2013.

1 0

Nested \setupXXX commands
by Wolfgang Schuster 26 Mar '13

26 Mar '13

Hi Hans, there is a problem with the command handler and nested \setup… commands. In the following example the setting for the “two” environment are ignored. <example> \defineenumeration[one][text=One] \defineenumeration[two][text=Two] \setupenumeration[one,two][alternative=serried,color=red] \starttext \startone First enumeration. \stopone \starttwo Second enumeration. \stoptwo \stoptext </example> The reason for this behavior is the \setupcounter call in \everysetupenumeration. When you look at the internal code for \setupenumeration (mult-aux.mkiv) you see that the first element in \setupernumeration (i.e. “one”) is processed with \mult_interfaces_with_comma_list_element, at the end of this macro \everysetupenumeration. At this point \setupcounter is used which redefines \mult_interfaces_with_comma_list_element. This redefined version is used for the next element (i.e. “two”) in \setupenumeration and the values are no passed (“alternative” and “color”) to \setupcounter. \unexpanded\def\mult_interfaces_install_setup_handler#1#2#3#4#5#6#7#8% {\ifx#3\relax\let#3\empty\fi \unexpanded\def#2{\dodoubleempty#4}% \unexpanded\def#6{\mult_interfaces_get_parameters{#1#3:}}% no every ! don't change it \newtoks#5% \newtoks#8% \unexpanded\def#4[##1][##2]% maybe helper {\let#7#3% \ifsecondargument \def\mult_interfaces_with_comma_list_element####1% we will have a simple one as well {\edef#3{####1}% \mult_interfaces_get_parameters{#1#3:}[##2]% \the#5}% \processcommalist[##1]\mult_interfaces_with_comma_list_element \else \let#3\empty \mult_interfaces_get_parameters{#1:}[##1]% \the#5% \fi \let#3#7% \the#8}} Wolfgang

2 1