- dev-context - NTG Mailing Lists

i-fontfamily.xml
by Wolfgang Schuster 23 May '16

23 May '16

Hi Hans, 1. The default value for the "force" key from \definefallbackfamily is "no". <cd:command name="definefallbackfamily" file="font-sel.mkvi"> <cd:arguments> <cd:resolve name="keyword-name"/> <cd:keywords> <cd:constant type="rm" default="yes"/> <cd:constant type="ss"/> <cd:constant type="tt"/> <cd:constant type="mm"/> <cd:constant type="hw"/> <cd:constant type="cg"/> <cd:constant type="roman"/> <cd:constant type="serif"/> <cd:constant type="regular"/> <cd:constant type="sansserif"/> <cd:constant type="sans"/> <cd:constant type="support"/> <cd:constant type="teletype"/> <cd:constant type="type"/> <cd:constant type="mono"/> <cd:constant type="handwritten"/> <cd:constant type="calligraphic"/> <cd:constant type="math"/> <cd:constant type="mathematics"/> </cd:keywords> <cd:resolve name="keyword-font"/> <cd:assignments list="yes" optional="yes"> <cd:parameter name="rscale"> <cd:constant type="cd:number"/> </cd:parameter> [...] <cd:parameter name="check"> <cd:constant type="yes"/> <cd:constant type="no" default="yes"/> </cd:parameter> <cd:parameter name="force"> - <cd:constant type="yes" default="yes"/> - <cd:constant type="no"/> + <cd:constant type="yes"/> + <cd:constant type="no" default="yes"/> </cd:parameter> <cd:parameter name="features"> <cd:constant type="cd:name"/> </cd:parameter> [...] </cd:assignments> </cd:arguments> </cd:command> 2. There is a "s" missing in at the end of the assignments tag. <cd:command name="definefontfamilypreset" file="font-sel.mkvi"> <cd:arguments> <cd:resolve name="keyword-name"/> - <cd:assignment list="yes"> - <cd:inherit name="definefontfamily"/> - </cd:assignment> + <cd:assignments list="yes"> + <cd:inherit name="definefontfamily"/> + </cd:assignments> </cd:arguments> </cd:command> Wolfgang

1 0

i-math.xml
by Wolfgang Schuster 23 May '16

23 May '16

Hi Hans, I got a mail by Hans van der Meer about wrong entry for the \setupmathematics command. <cd:command name="setupmathematics" file="math-ini.mkiv"> <cd:arguments> <cd:resolve name="keyword-name-list-optional"/> <cd:assignments list="yes"> <cd:parameter name="openup"> <cd:constant type="yes"/> <cd:constant type="no" default="yes"/> </cd:parameter> [...] <cd:parameter name="textcolor"> <cd:constant type="cd:color"/> </cd:parameter> - <cd:parameter name="functionstyle"> - <cd:resolve name="value-style"/> - </cd:parameter> + <cd:parameter name="functioncolor"> + <cd:constant type="cd:color"/> + </cd:parameter> <cd:parameter name="integral"> <cd:constant type="limits"/> <cd:constant type="nolimits"/> <cd:constant type="autolimits" default="yes"/> <cd:constant type="displaylimits"/> <cd:constant type="none"/> </cd:parameter> <cd:parameter name="default"> <cd:constant type="normal"/> <cd:constant type="italic" default="yes"/> </cd:parameter> </cd:assignments> </cd:arguments> </cd:command> Wolfgang

1 0

mtx-context.lua
by Brian R. Landy 17 May '16

17 May '16

Hi, I noticed today that when running "context --jit" that return codes were not being passed back properly to the shell. For example, with a simple file like this (text.tex): \starttext test\par \mybadmacro\par \stoptext context --jit --nonstopmode test.tex ; echo ", RC=$?" The log output will print: mtx-context | fatal error: return code: 1, RC=0 But the shell return code is zero. The below patch fixes it (I'm no Lua expert so perhaps there is a better approach): --- a/tex/texmf-context/scripts/context/lua/mtx-context.lua +++ b/tex/texmf-context/scripts/context/lua/mtx-context.lua @@ -108,7 +108,7 @@ local function restart(engine_old,engine_new) local command = format("%s --luaonly %q %s --redirected",engine_new,environment.ownname,environment.reconstructcommandline()) report(format("redirect %s -> %s: %s",engine_old,engine_new,command)) local result = os.execute(command) - os.exit(result) + os.exit(result/256) end Best regards, Brian

2 1

x-ldx.mkiv
by Wolfgang Schuster 10 May '16

10 May '16

Hi, typo in line 96: \startxmlsetups ldx:code \startpacked \xmlflush{#1}\relax - \scitepacked + \stoppacked \stopxmlsetups Wolfgang

1 0

typo-tal.lua line 281 (alignment character isssue)
by Brian R. Landy 04 May '16

04 May '16

Hi, I was hoping you would accept this patch to typo-tal.lua: index a67cd0a..100d95d 100644 --- a/tex/texmf-context/tex/context/base/mkiv/typo-tal.lua +++ b/tex/texmf-context/tex/context/base/mkiv/typo-tal.lua @@ -278,7 +278,7 @@ function characteralign.handler(originalhead,where) if char then local font = getfont(current) -- local unicode = unicodes[font][char] - local unicode = fontcharacters[font][char].unicode or char -- ignore tables + local unicode = (fontcharacters[font][char] and fontcharacters[font][char].unicode or char) -- ignore tables if not unicode then -- no unicode so forget about it elseif unicode == separator then -- This permits using certain special characters like the Unicode word joiner as the alignment character in a natural table. Without the patch ConTeXt aborts with an error: ... typo-tal.lua:281: attempt to index field '?' (a nil value) Thank you! Best regards, Brian Here is a simple example that aborts without the patch: \edef\mywordjoiner{\normalUchar"2060} \starttext \bTABLE \setupTABLE[c][1][aligncharacter=yes,alignmentcharacter={\mywordjoiner},align=middle] \bTABLEbody \bTR \bTD 1\mywordjoiner.234 \eTD \bTD 1\mywordjoiner.234 \eTD \eTR \bTR \bTD 123\mywordjoiner.4 \eTD \bTD 123\mywordjoiner.4 \eTD \eTR \eTABLEbody \eTABLE \stoptext

1 0

page-mak.mkvi
by Wolfgang Schuster 19 Apr '16

19 Apr '16

Hi, is the empty second argument on purpose? \definemakeup [\v!page] [] Wolfgang

2 1

font-ini.mkvi
by Wolfgang Schuster 18 Apr '16

18 Apr '16

Hi, use of "b" instead of "#". \unexpanded\def\font_basics_define_body_font_c[#whatever][#dummya][#dummyb]% body|identifier - {\font_basics_define_body_font_a[bwhatever][\s!rm][]} + {\font_basics_define_body_font_a[#whatever][\s!rm][]} Wolfgang

1 0

Re: [dev-context] No format file with last beta
by Wolfgang Schuster 11 Apr '16

11 Apr '16

> Hans Hagen <mailto:pragma@wxs.nl> > 10. April 2016 um 19:16 > i wonder at what point it crashes (where in what file) It stops at the following line in node-ini.lua local boundarycodes = mark(getsubtypes("boundary")) Wolfgang

3 4

No format file with last beta
by Wolfgang Schuster 10 Apr '16

10 Apr '16

Hi Hans, I’m unable to generate a format file for MkIV with the last beta, the last line in the log is: open source > 1 > 34 > /Users/wolf/context/beta/tex/texmf-context/tex/context/base/mkiv/node-ini.mkiv loading : ConTeXt Node Support / Initialization Could this be related to the LuaTeX version for the Mac: This is LuaTeX, Version beta-0.90.0 (TeX Live 2016) Wolfgang

2 1

On LuaTeX beta-0.95.0
by luigi scarso 07 Apr '16

07 Apr '16

Hi all, we have put beta-0.95.0 on the luatex svn already from while, but the tag is not frozen because: 1) Hans is working on the manual, 2) I keep source/ in sync with TeXLive (kpathsea files etc ), 3) we fix bugs, of course --- the building process not always goes well on every the platforms that we support. The initial plan was to put 0.90.0 for TeXLive, but now we are pretty confident that 0.95.0 is ok, and we have no plan to touch the codebase in source/ (bug fixing apart). (@Karl: I can keep Master/texmf-dist/doc/luatex/base in sync, if it's ok for you) For those that build from source: A. Kakuto keeps the TeXLive repo aligned with the luatex repo, assuring that the windows binaries are ok. It's clear that, having two repos with the same program, is a bit hard to decide which program is the latest one, so when I import from TeXLive in the commit message I specify the revision (e.g. "sync with TL rev. 40299." ); in the luatex repo, trunk and tags/beta-0.95.0 are always aligned, always first trunk and immediately after beta-0.95.0. I found more simple to have both the repos (texlive and luatex) on my machine, but it's my choice. Unfortunately currently the contextgarden server has some problems, and perhaps not all context users have the latest luatex --- not a big problen, we are working on this. The formal announcement of Luatex 0.95.0 very likely will be when the TeXLive is frozen. around 4may. -- luigi

1 0