Hi, I uploaded a new beta. The most important changes are: - Environments and modules are no longer loaded under protection so this has to be done explicitly in the files themselves. So far we haven't encountered problems with that change. - Some more cont-*.tex files have been split in mkii and mkiv versions. In mkii a tex file is loaded when no mkii file is found, but in mkiv only files with suffix mkiv are loaded. That way we stay compatible. This also means that cont-sys.tex is no longer loaded in mkiv which is okay as normally nothing in there concerns mkiv. If you use that file for setting up for instance another default paperformat, you'd better create a cont-sys.mkiv file. (Maybe that file should go and we should use the configuration file instead.) - The multilingual interface and messages are now loaded at the \LUA\ end. If you observe problems, please let me know. - If you are adventurous you can add the following to cont-new.mkiv \enablemode[*nofonts] This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while. - There have been some fixes to linebreaks in verbatim. Hopefully I didn't mess things up too much. - An upcoming version will also have \nonknuthmode as default. You can test this by putting this macro in cont-new.mkiv. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
Hi Hans, On Linux with debian minimals: l.42 \loadmarkfile{catc-ini} Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini} Cheers, Renaud Le 18/01/2011 19:55, Hans Hagen a écrit :
Hi,
I uploaded a new beta. The most important changes are:
- Environments and modules are no longer loaded under protection so this has to be done explicitly in the files themselves. So far we haven't encountered problems with that change.
- Some more cont-*.tex files have been split in mkii and mkiv versions. In mkii a tex file is loaded when no mkii file is found, but in mkiv only files with suffix mkiv are loaded. That way we stay compatible. This also means that cont-sys.tex is no longer loaded in mkiv which is okay as normally nothing in there concerns mkiv. If you use that file for setting up for instance another default paperformat, you'd better create a cont-sys.mkiv file. (Maybe that file should go and we should use the configuration file instead.)
- The multilingual interface and messages are now loaded at the \LUA\ end. If you observe problems, please let me know.
- If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
- There have been some fixes to linebreaks in verbatim. Hopefully I didn't mess things up too much.
- An upcoming version will also have \nonknuthmode as default. You can test this by putting this macro in cont-new.mkiv.
Hans
----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl ----------------------------------------------------------------- ___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________
On 18-1-2011 8:13, Renaud AUBIN wrote:
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
afaik nothing changed there ... installs ok here on windows and xubuntu Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
Ok, using /usr/local/ConTeXt_version_2010.11.26 will not please the resolver or something else because of the dot… mv do the job It would be nice to think of those like me who use dot into their naming scheme… ;) Renaud Le 18/01/2011 20:13, Renaud AUBIN a écrit :
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
Cheers,
Renaud
Renaud AUBIN
It would be nice to think of those like me who use dot into their naming scheme… ;)
See also http://www.ntg.nl/pipermail/ntg-context/2010/052339.html ;) -- Peter Münster Contact information: http://pmrb.free.fr/contact/
On 18-1-2011 10:04, Peter Münster wrote:
Renaud AUBIN
writes: It would be nice to think of those like me who use dot into their naming scheme… ;)
See also http://www.ntg.nl/pipermail/ntg-context/2010/052339.html ;)
well, having a * in a filename is a bad idea anyway Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
Except that I used . And not *
"Hans Hagen"
On 18-1-2011 10:04, Peter Münster wrote:
Renaud AUBIN
writes: It would be nice to think of those like me who use dot into their naming scheme… ;)
See also http://www.ntg.nl/pipermail/ntg-context/2010/052339.html ;)
well, having a * in a filename is a bad idea anyway
Hans
----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl ----------------------------------------------------------------- ___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________
-- Envoyé de mon téléphone Android avec K-9 Mail. Excusez la bièveté.
On 18-1-2011 9:32, Renaud AUBIN wrote:
Ok, using /usr/local/ConTeXt_version_2010.11.26 will not please the resolver or something else because of the dot… mv do the job
It would be nice to think of those like me who use dot into their naming scheme… ;)
hm, i wonder what goes wrong, as catc-ini is not the first file that is loaded
Renaud
Le 18/01/2011 20:13, Renaud AUBIN a écrit :
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
Cheers,
Renaud
___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________
-- ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On Tuesday 18 January 2011 20:13:36 Renaud AUBIN wrote:
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
There appears to be a problem (for quite some time, not just for the current beta) of the robustness of \loadmarkfile{}. Indeed, quite often, the conditions under which this occurs is not clear to me, but almost systematically, updating the minimals gives such errors. The update succeeds by entering the full path, such as /usr/local/context/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv and repeating the whole path many times, for all instances of unfound markfiles (under X11, right-click is your friend)! It is strange, as the update process finds the other files perfectly well, and is really annoying. Alan
On 26-1-2011 8:50, Alan BRASLAU wrote:
On Tuesday 18 January 2011 20:13:36 Renaud AUBIN wrote:
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
There appears to be a problem (for quite some time, not just for the current beta) of the robustness of \loadmarkfile{}. Indeed, quite often, the conditions under which this occurs is not clear to me, but almost systematically, updating the minimals gives such errors. The update succeeds by entering the full path, such as /usr/local/context/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv and repeating the whole path many times, for all instances of unfound markfiles (under X11, right-click is your friend)!
It is strange, as the update process finds the other files perfectly well, and is really annoying.
it is probably a side effect of files being renamed, i.e. the new name is mkiv while the old one is tex; in that case the file database has to be rebuild (using mtxrun --generate); it can also happen when there are new files added so, the main question is: why isn't "mtxrun --generate" run? Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On Wed, 26 Jan 2011 09:26:18 +0100
Hans Hagen
On 26-1-2011 8:50, Alan BRASLAU wrote:
On Tuesday 18 January 2011 20:13:36 Renaud AUBIN wrote:
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
There appears to be a problem (for quite some time, not just for the current beta) of the robustness of \loadmarkfile{}. Indeed, quite often, the conditions under which this occurs is not clear to me, but almost systematically, updating the minimals gives such errors. The update succeeds by entering the full path, such as /usr/local/context/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv and repeating the whole path many times, for all instances of unfound markfiles (under X11, right-click is your friend)!
It is strange, as the update process finds the other files perfectly well, and is really annoying.
it is probably a side effect of files being renamed, i.e. the new name is mkiv while the old one is tex; in that case the file database has to be rebuild (using mtxrun --generate); it can also happen when there are new files added
so, the main question is: why isn't "mtxrun --generate" run?
Hans
Yes, I have encountered this problem many times during the last weeks. Aborting the format-making, rerunning mtxrun --generate works too, after that, all the files that have been flagged before are found. So the question is: is mtxrun --generate maybe run too early in the process, before all the files are copied into their place? Thomas
On 26-1-2011 9:41, Thomas Schmitz wrote:
On Wed, 26 Jan 2011 09:26:18 +0100 Hans Hagen
wrote: On 26-1-2011 8:50, Alan BRASLAU wrote:
On Tuesday 18 January 2011 20:13:36 Renaud AUBIN wrote:
Hi Hans,
On Linux with debian minimals:
l.42 \loadmarkfile{catc-ini}
Please type another input file name: ! Emergency stop. <to be read again> \relax l.42 \loadmarkfile{catc-ini}
There appears to be a problem (for quite some time, not just for the current beta) of the robustness of \loadmarkfile{}. Indeed, quite often, the conditions under which this occurs is not clear to me, but almost systematically, updating the minimals gives such errors. The update succeeds by entering the full path, such as /usr/local/context/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv and repeating the whole path many times, for all instances of unfound markfiles (under X11, right-click is your friend)!
It is strange, as the update process finds the other files perfectly well, and is really annoying.
it is probably a side effect of files being renamed, i.e. the new name is mkiv while the old one is tex; in that case the file database has to be rebuild (using mtxrun --generate); it can also happen when there are new files added
so, the main question is: why isn't "mtxrun --generate" run?
Hans
Yes, I have encountered this problem many times during the last weeks. Aborting the format-making, rerunning mtxrun --generate works too, after that, all the files that have been flagged before are found. So the question is: is mtxrun --generate maybe run too early in the process, before all the files are copied into their place?
I just synced the minimals on a xubuntu virtual machine and it works ok. I never saw that problem, and recently I updated 6 machines that ran the minimals without problems (and i see mtxrun --generate being run). Maybe next time capture all the messages on the console so that we can see what happens at your end. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
so, the main question is: why isn't "mtxrun --generate" run?
It is run. But this has been broken since the very beginning. When you run "mtxrun --generate" from withit mtx-update in creates a different database and updates that one (at least that used to be a problem in past; I didn't check that recently, but it is one of few longstanding nasty bugs (the other one is inability to properly set the platform in a decent way)). Most of the time it is not as nasty since formats for MKIV are rebuild later anyway, so often one would not notice the problem. At the end there are two databases then. It would help to know how you calculate the md5 hash and why it chooses two different hashes ... But I will try to check again. Mojca
On Wed, Jan 26, 2011 at 13:31, Mojca Miklavec
so, the main question is: why isn't "mtxrun --generate" run?
It is run. But this has been broken since the very beginning.
When you run "mtxrun --generate" from withit mtx-update in creates a different database and updates that one (at least that used to be a problem in past; I didn't check that recently, but it is one of few longstanding nasty bugs (the other one is inability to properly set the platform in a decent way)).
Most of the time it is not as nasty since formats for MKIV are rebuild later anyway, so often one would not notice the problem. At the end there are two databases then.
Or maybe not. I need to have a closer look again. I tried to reinstall now and it generated only a single database (for mkiv, but for mkii some files were missing), however there was no mkiv format. Mojca
I think I understand at least part of the problem now. When I run ./first-setup.sh it runs /usr/local/texlive/2010/bin/x86_64-darwin/mktexlsr /usr/local/texlive/2010/bin/x86_64-darwin/mtxrun instead of mktexlsr and mtxrun from the newly installed distribution. Is anyone willing to try to change part of first-setup.sh into Option A (more bullet-proof): --------------------------------------------- CONTEXTROOT="$PWD/tex" OLDPATH=$PATH # download or rsync the latest scripts first rsync -rlptv rsync://contextgarden.net/minimals/setup/$platform/bin . # download or update the distribution # you may remove the --context=beta switch if you want to use "current" export PATH=$PWD/bin:"$CONTEXTROOT/texmf-$platform/bin":$OLDPATH mtxrun --script ./bin/mtx-update.lua --force --update --engine=all --context=beta --platform=$platform --texroot="$CONTEXTROOT" $@ export PATH="$CONTEXTROOT/texmf-$platform/bin":$OLDPATH mtxrun --script ./bin/mtx-update.lua --force --make --engine=all $@ --------------------------------------------- Option B: --------------------------------------------- CONTEXTROOT="$PWD/tex" # download or rsync the latest scripts first rsync -rlptv rsync://contextgarden.net/minimals/setup/$platform/bin . # download or update the distribution # you may remove the --context=beta switch if you want to use "current" export PATH=$PWD/bin:"$CONTEXTROOT/texmf-$platform/bin":$PATH mtxrun --script ./bin/mtx-update.lua --force --update --make --engine=all --context=beta --platform=$platform --texroot="$CONTEXTROOT" $@ --------------------------------------------- and test if that behaves any better? The problem is that in the first step one doesn't necessary have mtxrun & luatex at all or they might be too recent, so we have to put "$PWD/bin/" in front of PATH. However it might be much better to use (now existing) luatex and mtxrun from $CONTEXTROOT/texmf-$platform/bin when updating the database and making formats. Later I decided that one also needs $CONTEXTROOT/texmf-$platform/bin before calling --update since mktexlsr is also called at the end of update. I will rewrite some of the scripts from scratch anyway, but I hope that this should work now. I would have to test whether there is a more elegant way for enforcing some specific type of platform when uname lies. Mojca
On Wednesday 26 January 2011 19:56:04 Mojca Miklavec wrote:
I will rewrite some of the scripts from scratch anyway, but I hope that this should work now. I would have to test whether there is a more elegant way for enforcing some specific type of platform when uname lies.
Maybe "bash --version" would be more reliable than uname? Alan
On 01/26/11 22:01, Alan BRASLAU wrote:
On Wednesday 26 January 2011 19:56:04 Mojca Miklavec wrote:
I will rewrite some of the scripts from scratch anyway, but I hope that this should work now. I would have to test whether there is a more elegant way for enforcing some specific type of platform when uname lies.
Maybe "bash --version" would be more reliable than uname?
If you do something like that, then you may as well do $ file /bin/ls Best wishes, Taco
Hans Hagen
- If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem: %%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt] \starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%% It fails because itemize seems to require some math characters (the bullets) and the libertine font seems not to have them. lmodern don't kick in, it just fails: ! Math error: parameter \Umathquad\displaystyle is not set. \mathematics ...rmalstartimath #1\normalstopimath \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi <to be read again> } ... l.7 \item o ne ? Bests and thanks for all your work -- Marco
Am 25.01.2011 um 21:18 schrieb Marco Pessotto:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
%%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt]
\starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%%
It fails because itemize seems to require some math characters (the bullets) and the libertine font seems not to have them. lmodern don't kick in, it just fails:
Works for me with the latest minimals. For Libertine bullets you can use the following: \setupbodyfont[libertine] \definesymbol[1][•] % otherwise LMMathSymbols10-Regular \starttext \startitemize \item one \item two \stopitemize \stoptext Greeting Andreas
Andreas Harder
Am 25.01.2011 um 21:18 schrieb Marco Pessotto:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
%%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt]
\starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%%
It fails because itemize seems to require some math characters (the bullets) and the libertine font seems not to have them. lmodern don't kick in, it just fails:
Works for me with the latest minimals. For Libertine bullets you can use the following:
\setupbodyfont[libertine] \definesymbol[1][•] % otherwise LMMathSymbols10-Regular
\starttext \startitemize \item one \item two \stopitemize \stoptext
Greeting Andreas
Just to be sure I've updated the Minimals, but with \enablemode[*nofonts] in cont-new.mkiv it fails. I don't need a workaround, because I can comment out the *nofonts in the cont-new.mkiv (which was added by myself). I was just reporting a problem, as asked (IIRC). $ context prova.tex MTXrun | run 1: luatex --fmt="/home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3/formats/cont-en" --lua="/home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3/formats/cont-en.lui" --backend=pdf "./prova.tex"This is LuaTeX, Version beta-0.65.0-2010121316 \write18 enabled. (prova.tex jobcontrol > resuming randomizer with 0.71793062133618 ConTeXt ver: 2011.01.25 13:17 MKIV fmt: 2011.1.25 int: english/english system : cont-new.mkiv loaded (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/cont-new.mkiv systems : beware: some patches loaded from cont-new.mkiv ) system : prova.top loaded (prova.top) fonts : latin modern fonts are not preloaded language : language en is active (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/type-siz.mkiv) (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/type-otf.mkiv){/home/melmoth/usr/context/tex/texmf-context/fonts/map/pdftex/context/mkiv-base.map} systems : begin file prova.tex at line 5 ! Math error: parameter \Umathquad\displaystyle is not set. \mathematics ...rmalstartimath #1\normalstopimath \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi <to be read again> } ... l.7 \item o ne ? ? Type <return> to proceed, S to scroll future error messages, R to run without stopping, Q to run quietly, I to insert something, E to edit your file, 1 or ... or 9 to ignore the next 1 to 9 tokens of input, H for help, X to quit. ? X mkiv lua stats : used config file - /home/melmoth/usr/context/tex/texmf/web2c/texmfcnf.lua mkiv lua stats : used cache path - /home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3 mkiv lua stats : input load time - 0.015 seconds mkiv lua stats : stored bytecode data - 265 modules, 56 tables, 321 chunks mkiv lua stats : cleaned up reserved nodes - 30 nodes, 9 lists of 412 mkiv lua stats : node memory usage - 1 hlist, 1 math, 2 glue, 1 noad, 1 math_char, 19 attribute, 20 glue_spec, 4 attribute_list, 2 temp, 7 if_stack, 1 local_par, 3 dir mkiv lua stats : node list callback tasks - 6 unique task lists, 2 instances (re)created, 1773 calls mkiv lua stats : used backend - pdf (backend for directly generating pdf output) mkiv lua stats : loaded patterns - en::2 mkiv lua stats : callbacks - direct: 1788, indirect: 2937, total: 4725 mkiv lua stats : result saved in file - prova.pdf mkiv lua stats : loaded fonts - 1 files: fxlr.otf mkiv lua stats : fonts load time - 0.101 seconds mkiv lua stats : luatex banner - this is luatex, version beta-0.65.0-2010121316 mkiv lua stats : control sequences - 29335 of 165536 mkiv lua stats : current memory usage - 18 MB (ctx: 19 MB) mkiv lua stats : runtime - 6.489 seconds Cheers -- Marco
On 25-1-2011 10:18, Marco Pessotto wrote:
I don't need a workaround, because I can comment out the *nofonts in the cont-new.mkiv (which was added by myself). I was just reporting a problem, as asked (IIRC).
well, at some point it will be default so we need to catch these cases (or fix the typescripts) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On 25-1-2011 9:35, Andreas Harder wrote:
Am 25.01.2011 um 21:18 schrieb Marco Pessotto:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
%%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt]
\starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%%
It fails because itemize seems to require some math characters (the bullets) and the libertine font seems not to have them. lmodern don't kick in, it just fails:
Works for me with the latest minimals. For Libertine bullets you can use the following:
\setupbodyfont[libertine] \definesymbol[1][•] % otherwise LMMathSymbols10-Regular
\starttext \startitemize \item one \item two \stopitemize \stoptext
we can consider adding some suitable math font \starttypescript [libertine] \definetypeface [libertine] [rm] [serif] [libertine] [default] \definetypeface [libertine] [ss] [sans] [biolinum] [default] \definetypeface [libertine] [tt] [mono] [default] [default] %definetypeface [libertine] [mm] [math] [times] [default] \quittypescriptscanning \stoptypescript at some point i want to redefine the bullets to be more clever (it is a pitty that regular text bullets are often so small) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
On Tue, 25 Jan 2011, Marco Pessotto wrote:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
%%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt]
\starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%%
Can you test if \setupbodyfont[libertine,12pt] \starttext $ABC \bullet \star$ \stoptext also fails?
It fails because itemize seems to require some math characters (the bullets) and the libertine font seems not to have them. lmodern don't kick in, it just fails:
! Math error: parameter \Umathquad\displaystyle is not set. \mathematics ...rmalstartimath #1\normalstopimath \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi \dodosymbol ...bol \csname \??ss :#1:#2\endcsname \relax }\relax \donormalsymbol ...bol \redosymbol \currentsymbol \fi \fi <to be read again> } ... l.7 \item o ne
Seems like a bug. If no math typescripts are defined, the fallback should have been loaded. (In any case, the libertine typescript should use either txfonts or perhaps xits as the default math fonts). Aditya
Aditya Mahajan
On Tue, 25 Jan 2011, Marco Pessotto wrote:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
%%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt]
\starttext \startitemize[1] \item one \item two \stopitemize \stoptext %%%% end %%%%
Can you test if
\setupbodyfont[libertine,12pt] \starttext $ABC \bullet \star$ \stoptext
also fails?
Yes. Full log here melmoth@universe:/tmp$ context --version MTXrun | main context file: /home/melmoth/usr/context/tex/texmf-context/tex/context/base/context.mkiv MTXrun | current version: 2011.01.25 13:17 melmoth@universe:/tmp$ cat prova.tex %%%% minimal example %%%% \usetypescript[libertine] \setupbodyfont[libertine,12pt] \starttext $ABC \bullet \star$ \stoptext %%%% end %%%% melmoth@universe:/tmp$ context prova.tex MTXrun | run 1: luatex --fmt="/home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3/formats/cont-en" --lua="/home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3/formats/cont-en.lui" --backend=pdf "./prova.tex"This is LuaTeX, Version beta-0.65.0-2010121316 \write18 enabled. (prova.tex jobcontrol > resuming randomizer with 0.71793062133618 ConTeXt ver: 2011.01.25 13:17 MKIV fmt: 2011.1.25 int: english/english system : cont-new.mkiv loaded (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/cont-new.mkiv systems : beware: some patches loaded from cont-new.mkiv ) system : prova.top loaded (prova.top) fonts : latin modern fonts are not preloaded language : language en is active (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/type-siz.mkiv) (/home/melmoth/usr/context/tex/texmf-context/tex/context/base/type-otf.mkiv){/home/melmoth/usr/context/tex/texmf-context/fonts/map/pdftex/context/mkiv-base.map} systems : begin file prova.tex at line 4 ! Math error: parameter \Umathquad\displaystyle is not set. l.5 $ABC \bullet \star$ ? ? Type <return> to proceed, S to scroll future error messages, R to run without stopping, Q to run quietly, I to insert something, E to edit your file, 1 or ... or 9 to ignore the next 1 to 9 tokens of input, H for help, X to quit. ? H Sorry, but I can't typeset math unless various parameters have been set. This is normally done by loading special math fonts into the math family slots. Your font set is lacking at least the parameter mentioned earlier. ? X mkiv lua stats : used config file - /home/melmoth/usr/context/tex/texmf/web2c/texmfcnf.lua mkiv lua stats : used cache path - /home/melmoth/usr/context/tex/texmf-cache/luatex-cache/context/161a8cd09bbec9213c251f73843e59e3 mkiv lua stats : input load time - 0.015 seconds mkiv lua stats : stored bytecode data - 265 modules, 56 tables, 321 chunks mkiv lua stats : cleaned up reserved nodes - 30 nodes, 9 lists of 412 mkiv lua stats : node memory usage - 1 hlist, 1 math, 1 glue, 5 noad, 5 math_char, 57 attribute, 18 glue_spec, 10 attribute_list, 1 temp, 1 local_par, 2 dir mkiv lua stats : node list callback tasks - 6 unique task lists, 2 instances (re)created, 1767 calls mkiv lua stats : used backend - pdf (backend for directly generating pdf output) mkiv lua stats : loaded patterns - en::2 mkiv lua stats : callbacks - direct: 1782, indirect: 2932, total: 4714 mkiv lua stats : result saved in file - prova.pdf mkiv lua stats : loaded fonts - 1 files: fxlr.otf mkiv lua stats : fonts load time - 0.102 seconds mkiv lua stats : luatex banner - this is luatex, version beta-0.65.0-2010121316 mkiv lua stats : control sequences - 29324 of 165536 mkiv lua stats : current memory usage - 18 MB (ctx: 19 MB) mkiv lua stats : runtime - 14.364 seconds
Seems like a bug. If no math typescripts are defined, the fallback should have been loaded. (In any case, the libertine typescript should use either txfonts or perhaps xits as the default math fonts).
Aditya
Cheers -- Marco
On 25-1-2011 9:18, Marco Pessotto wrote:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
new beta that checks for math and mono being loaded and if not it loads modern fallbacks ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | voip: 087 875 68 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------
Hans Hagen
On 25-1-2011 9:18, Marco Pessotto wrote:
Hans Hagen
writes: - If you are adventurous you can add the following to cont-new.mkiv
\enablemode[*nofonts]
This will postpone font loading and will be default soon. This saves runtime when you use another font than latin modern. When no font is loaded before the first \starttext, modern will kick in. A side effect is that any text typeset before \starttext will not show up in the output. If you observe that, you need to setup a bodyfont explicitly beforehand. In practice there should be no problems. The delayed loading is somewhat tricky but has been tested for a while.
If felt adventurous and I found the following problem:
new beta that checks for math and mono being loaded and if not it loads modern fallbacks
Works here. Thanks -- Marco
participants (10)
-
Aditya Mahajan
-
Alan BRASLAU
-
Andreas Harder
-
Hans Hagen
-
Marco Pessotto
-
Mojca Miklavec
-
pmlists@free.fr
-
Renaud AUBIN
-
Taco Hoekwater
-
Thomas Schmitz