On Mon, 31 Jan 2011 18:33:57 +0100 Otared Kavian wrote:

Hi Hans,

I uploaded the new beta with the Minimals, but neither mkii nor mkiv work anymore… I get MTXrun | unknown script 'context.lua' or 'mtx-context.lua' and luatools --generate gives the same result.

This has been deprecated for quite a while now; you have to run mtxrun --generate

Has something changed in first-setup.sh for Mac OS X and other machines (I followed the thread about that file being changed for Windows by Mojca, but I don't know about other operating systems).

I see the same thing on linux: you have to run mtxrun --generate and mktexlsr and texexec --make by hand now; after that, mkiv and mkii will run. I assume that's a result of our discussion last week, but not quite what Hans and Mojca are aiming at... Thomas

Surprised this works somewhere. I uploaded the beta with the script I am using for sometime now: #!/bin/bash mkdir -p $TEXMINIMAL/tex cd $TEXMINIMAL ln -s $TEXPRIVATE tex/texmf-local curl -o first-setup.sh http://minimals.contextgarden.net/setup/first-setup.sh sh ./first-setup.sh >first-setup.log cd tex echo "luatools --selfupdate" luatools --selfupdate echo "mtxrun -selfupdate" mtxrun -selfupdate echo "mtxrun -selfupdate" echo "installation context minimals done" But now I get an error in running my TeXShop after installing: MTXrun | unknown script 'context.lua' or 'mtx-context.lua' Confirmed that something is broken by regressing to my previous installation of ConTeXt ver: 2011.01.26 11:02 MKIV fmt: 2011.1.31 int: english/english Changing the script to mtxrun --generate do not improve things here. Thanks for changing things that way between one and the other. I am starting to get a bit choleric about all things that constantly seem broken with the next update. Sorry for the sour note but these are costing me hours to iron out, time I do not have at the moment. Hans van der Meer On 31 jan 2011, at 18:45, Thomas Schmitz wrote:

On Mon, 31 Jan 2011 18:33:57 +0100 Otared Kavian wrote:

...

Hi Hans, I uploaded the new beta with the Minimals, but neither mkii nor mkiv work anymore… I get MTXrun | unknown script 'context.lua' or 'mtx-context.lua' and luatools --generate gives the same result.

This has been deprecated for quite a while now; you have to run

mtxrun --generate

...

Has something changed in first-setup.sh for Mac OS X and other machines (I followed the thread about that file being changed for Windows by Mojca, but I don't know about other operating systems).

I see the same thing on linux: you have to run mtxrun --generate and mktexlsr and texexec --make by hand now; after that, mkiv and mkii will run.

I assume that's a result of our discussion last week, but not quite what Hans and Mojca are aiming at...

Thomas ___________________________________________________________________________________ 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 31-1-2011 7:11, Hans van der Meer wrote:

Thanks for changing things that way between one and the other. I am starting to get a bit choleric about all things that constantly seem broken with the next update. Sorry for the sour note but these are costing me hours to iron out, time I do not have at the moment.

well, no one forces you to update betas also, as long as we're in transition to mkiv you can expect an occasional hickup, after all, i have only a few machines to test maybe it gives you comfort to know that I spend more hours to iron out problems -) Hans ps. recently i spent a week ironing out nfs4 problems just to find out (by googling) that i'd best revert al machines back to nfs3 and i can give more examples; believe it or not, it makes me feel less guilty ----------------------------------------------------------------- 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 1-2-2011 6:53, Vnpenguin wrote:

On Mon, Jan 31, 2011 at 22:28, Hans Hagen wrote:

...

On 31-1-2011 7:11, Hans van der Meer wrote:

...

Thanks for changing things that way between one and the other. I am starting to get a bit choleric about all things that constantly seem broken with the next update. Sorry for the sour note but these are costing me hours to iron out, time I do not have at the moment.

well, no one forces you to update betas

also, as long as we're in transition to mkiv you can expect an occasional hickup, after all, i have only a few machines to test

It's better to have someone could give you a hand. In the long term, something like "Release Team" would be a must for ConTeXt development. With this team, we can do some "Unit testing" for each release in order to minimize problem.

to some extend this is what the beta is for ... because mkiv is a moving target it sort of replaces the current, but if i can imagine that users wait a few days with updating till there are no problems reported concerning testing ... there is a slowly growing test suite, and i guess that at some point Luigi will run it alongside his module documentation runs .. you can offer him help with that (sanjoy also has made some test setup) in this particular case (adaption of mtxrun) testing is somewhat problematic as we need a test-minimal alongside the regular one and that might happen when mojca has set up the new machine 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 -----------------------------------------------------------------

My update proceedure systematically maintains a previous, presumably working, copy of the minimals. For this, I run a script: #!/bin/sh cd /usr/local/context cd beta echo "*** move current to previous ***" rsync -av --delete --progress . ../previous echo "*** update beta ***" rsync -ptv --progress rsync://contextgarden.net/minimals/setup/first-setup.sh . ./first-setup.sh --extras=all | tee first-setup.log This way I limit problems while still working with the latest, greatest beta version of ConTeXt. When things go wrong with the update, I simply move "beta" to "new" (or something else) and "previous" to "beta", my PATH being set to use "beta". Perhaps others can suggest a better working proceedure. Of course, when pressed for time, I postpone updating to Hans' latest so as not to be distracted from the work at hand. Alan On Tuesday 01 February 2011 09:43:00 Hans Hagen wrote:

On 1-2-2011 6:53, Vnpenguin wrote:

...

On Mon, Jan 31, 2011 at 22:28, Hans Hagen wrote:

...

On 31-1-2011 7:11, Hans van der Meer wrote:

...

Thanks for changing things that way between one and the other. I am starting to get a bit choleric about all things that constantly seem broken with the next update. Sorry for the sour note but these are costing me hours to iron out, time I do not have at the moment.

well, no one forces you to update betas

also, as long as we're in transition to mkiv you can expect an occasional hickup, after all, i have only a few machines to test

It's better to have someone could give you a hand. In the long term, something like "Release Team" would be a must for ConTeXt development. With this team, we can do some "Unit testing" for each release in order to minimize problem.

to some extend this is what the beta is for ... because mkiv is a moving target it sort of replaces the current, but if i can imagine that users wait a few days with updating till there are no problems reported

concerning testing ... there is a slowly growing test suite, and i guess that at some point Luigi will run it alongside his module documentation runs .. you can offer him help with that (sanjoy also has made some test setup)

in this particular case (adaption of mtxrun) testing is somewhat problematic as we need a test-minimal alongside the regular one and that might happen when mojca has set up the new machine

Hans

On 1/31/11 6:45 PM, Thomas Schmitz wrote:

...

and luatools --generate gives the same result.

This has been deprecated for quite a while now; you have to run

mtxrun --generate Does that mean that the instructions (http://wiki.contextgarden.net/ConTeXt_Minimals#Remaking_formats) have to be updated? There, both "mtxrun --generate" and "luatools --generate" are listed for making the mkiv format.

In any case, I haven't been able to install (update) the minimals in the last 3 months (on both OS X and ubuntu). MkII works, but context --version gives "MTXrun | main context file: unknown, 'context.mkiv' not found" And the log of running a test file contains: ------------- MTXrun | warning: no format found, forcing remake (commandline driven) ... MTXrun | resolvers: using given filetype 'tex' MTXrun | resolvers: resolver: method=concatinators, how=tag, tag=file MTXrun | resolvers: resolver: method=concatinators, how=tag, tag=file MTXrun | resolvers: using given filetype 'tex' MTXrun | no tex source file with name: cont-en (mkiv or tex) MTXrun | error, no format found with name: cont-en, aborting MTXrun | total runtime: 0.347 ----------------- From the installation log: ------------------ MTXrun | forcing cache reload MTXrun | resolvers: unknown configuration file '/Users/bonasia/texmf-local/texmfcnf.lua' MTXrun | resolvers: unknown configuration file '/Users/bonasia/texmf-local/web2c/texmfcnf.lua' MTXrun | resolvers: unknown configuration file '/Users/bonasia/texmf/texmfcnf.lua' MTXrun | resolvers: unknown configuration file '/Users/bonasia/texmf/web2c/texmfcnf.lua' MTXrun | resolvers: unknown configuration file '/Users/bonasia/texmfcnf.lua' MTXrun | resolvers MTXrun | resolvers: warning: no lua configuration files found MTXrun | resolvers MTXrun | resolvers MTXrun | the resolver databases are not present or outdated MTXrun | state: loaded MTXrun | update: start ------------------- Thanks and regards, Jörg

On 2/1/11 1:47 PM, luigi scarso wrote:

On Tue, Feb 1, 2011 at 1:42 PM, Jörg Hagmann wrote:

...

...

...

and luatools --generate gives the same result.

This has been deprecated for quite a while now; you have to run

mtxrun --generate Does that mean that the instructions (http://wiki.contextgarden.net/ConTeXt_Minimals#Remaking_formats) have to be updated? There, both "mtxrun --generate" and "luatools --generate" are

On 1/31/11 6:45 PM, Thomas Schmitz wrote: listed for making the mkiv format.

In any case, I haven't been able to install (update) the minimals in the last 3 months (on both OS X and ubuntu). hm so much time for me. At least on January update was almost retgular for me (mkiv)

...

"MTXrun | main context file: unknown, 'context.mkiv' not found" it can be a problem of cache corrupted; try to erase it and rebuild with mtxrun --generate Didn't help (provided I deleted the right thing). Thanks and regards, Jörg

On 2/1/11 2:26 PM, luigi scarso wrote:

On Tue, Feb 1, 2011 at 2:22 PM, Jörg Hagmann wrote:

...

Didn't help (provided I deleted the right thing). Did you $>. setuptex Sure. mkII is working.

On 2/1/11 2:42 PM, luigi scarso wrote:

On Tue, Feb 1, 2011 at 2:36 PM, Jörg Hagmann wrote:

...

...

Did you $>. setuptex Sure. mkII is working. and $> kpsewhich context.mkiv ? /Users/bonasia/.context-new/tex/texmf-context/tex/context/base/context.mkiv

On 2/1/11 2:54 PM, luigi scarso wrote:

On Tue, Feb 1, 2011 at 2:46 PM, Jörg Hagmann wrote:

...

On 2/1/11 2:42 PM, luigi scarso wrote:

...

On Tue, Feb 1, 2011 at 2:36 PM, Jörg Hagmann wrote:

...

...

Did you $>. setuptex Sure. mkII is working. and $> kpsewhich context.mkiv ? /Users/bonasia/.context-new/tex/texmf-context/tex/context/base/context.mkiv

^^^^^^^^^^^ this dot looks suspect... can you try without it, ie /Users/bonasia/context-new... and then $>. setuptex $>mtxrun --generate ? Ok - Thank you! I keep (kept) my contexts in hidden dirs (my current one from 3 months ago is in ".context" and it works, but I don't remember whether I changed the directory to "hidden" after I installed the minimals or whether I installed them into ".context").

Conclusion: Installing the actual minimals in a hidden file doesn't work, in a "visible" one it does. Thanks a lot for the help, Jörg

On 2/1/11 3:46 PM, luigi scarso wrote:

On Tue, Feb 1, 2011 at 3:38 PM, Jörg Hagmann wrote:

...

Conclusion: Installing the actual minimals in a hidden file doesn't work, in a "visible" one it does.

Not sure if it's a bug or feature. If I understand correctly, mkii works ok in "hidden" folders, but mkiv no. I suspect then a bug in names resolver which in this situation it's ok . (perhaps you can try with /Users/bonasia/a.context-new ) Doesn't work either. Well - my context directory doesn't have to be hidden - why hide a pretty thing?

Regards, Jörg

On Tuesday 01 February 2011 16:13:25 Vedran Miletić wrote:

Yes, there is a bug when using dots, someone recently triggered it when using versioned names (context-2011.01.18 or something). Not sure if it can be fixed.

A good guide should be the use of POSIX "Fully portable filenames" Lewine, Donald, POSIX Programmer's Guide: Writing Portable UNIX Programs, 1991, O'Reilly & Associates, Inc., Sebastopol, CA, pp. 63-64. Of course, some perfectly acceptable filenames (even for a NTFS filesystem, according to MicroSoft's own standard) wreak havoc for Windows and some Windows applications, as well as for MacOS. See, http://en.wikipedia.org/wiki/Filename Microsoft Windows: Windows kernel forbids the use of characters in range 1-31 (i.e., 0x01-0x1F) and characters " * : < > ? \ / |. Although NTFS allows each path component (directory or filename) to be 255 characters long and paths up to about 32767 characters long, the Windows kernel only supports paths up to 259 characters long if no UNC is used for addressing. Additionally, Windows forbids the use of the MS-DOS device names AUX, CLOCK$, COM1, COM2, COM3, COM4, COM5, COM6, COM7, COM8, COM9, CON, LPT1, LPT2, LPT3, LPT4, LPT5, LPT6, LPT7, LPT8, LPT9, NUL and PRN, as well as these names with any extension (for example, AUX.txt), except when using Long UNC paths (ex. \\.\C:\nul.txt or \\? \D:\aux\con). (In fact, CLOCK$ may be used if an extension is provided.) These restrictions only apply to Windows - Linux, for example, allows use of " * : <

? \ | even in NTFS.

Alan

On Tue, Feb 1, 2011 at 5:04 PM, Marco wrote:

On 2011-02-01 Vedran Miletić wrote:

...

...

...

...

 Not sure if it's a bug or feature. If I understand correctly, mkii works ok in "hidden" folders, but mkiv no. I suspect then a bug in names resolver which  in this situation it's ok . (perhaps  you can try  with /Users/bonasia/a.context-new )

Doesn't work either. Well - my context directory doesn't have to be hidden - why hide a pretty thing?

Regards, Jörg

Yes, there is a bug when using dots, someone recently triggered it when using versioned names (context-2011.01.18 or something). Not sure if it can be fixed.

I also ran across a bug considering the case of external graphics. The case is not taken into account. And luatex hangs, if a graphic is not found. Example:

\starttext \externalfigure [CoW]     % A file CoW does not exist, only cow.pdf                          % and cow.pdf is included, should output gray box

\externalfigure [GrApHiC] % Does not exist and causes mkiv to hang                          % Should output gray box \stoptext

Regards Marco confirmed -- luigi

On 2011-02-01 pmlists@free.fr (Peter Münster) wrote:

Vedran Miletić writes:

...

Yes, there is a bug when using dots,

Or a feature...?! ;)

See also: - http://www.ntg.nl/pipermail/ntg-context/2010/052339.html ("and such" can mean a dot) - http://pmrb.free.fr/contextref.pdf section 1.7 ;-)

I don't want to start an argue, sice this bug/feature doesn't affect my work. But just my 2 cents: 1) The operating system / file system basically decides which names are permitted. And in times of luatex and unicode it's a pity that file names with international characters aren't supported. All other programs can cope with that. 2) I often use automatically generated graphics like graph.1.pdf graph.2.pdf graph.3.pdf ... According to the documentation and Hans these file names aren't permitted, since dots and numbers aren't allowed. However when I include the complete names with extention it works. Regards Marco

2011/2/1 Hans Hagen

On 1-2-2011 1:40, Marco wrote:

...

Hi,

I get the following error while updating minimals:

MTXrun | run: texexec --make --all --fast --xetex cont-nl cont-enTeXExec | using search method 'kpsewhich' TeXExec | using existing database TeXExec | using tex engine xetex TeXExec | using tex format path /nonesuch/texmf/web2c/xetex TeXExec | generating tex format cont-nl.mkii This is XeTeX, Version 3.1415926-2.2-0.9995.2 (Web2C 7.5.6) (INITEX) entering extended mode ! I can't find file `cont-nl.mkii'. <*> *cont-nl.mkii

(Press Enter to retry, or Control-D to exit) Please type another input file name:

can you run mktexlsr manually?

Hans

Same here. Running mktexlsr manually works, but it doesn't help next ./first-setup run. Regards, Vedran Miletić

On Tue, Feb 1, 2011 at 12:35 PM, Hans Hagen wrote:

On 1-2-2011 11:36, Vedran Miletić wrote:

...

Same here.

Running mktexlsr manually works, but it doesn't help next ./first-setup run..

i'll have a look at it (i only tested a luatex only tree sofar) seem related only to pdftex and xetex. luatex seems ok. -- luigi

On 2011-02-01 Hans Hagen wrote:

On 1-2-2011 1:40, Marco wrote:

...

Hi,

I get the following error while updating minimals:

MTXrun | run: texexec --make --all --fast --xetex cont-nl cont-enTeXExec | using search method 'kpsewhich' TeXExec | using existing database TeXExec | using tex engine xetex TeXExec | using tex format path /nonesuch/texmf/web2c/xetex TeXExec | generating tex format cont-nl.mkii This is XeTeX, Version 3.1415926-2.2-0.9995.2 (Web2C 7.5.6) (INITEX) entering extended mode ! I can't find file `cont-nl.mkii'. <*> *cont-nl.mkii

(Press Enter to retry, or Control-D to exit) Please type another input file name:

can you run mktexlsr manually?

Yes. This fixes mkii. But as Vedran and Luigi already pointed out first-setup.sh is broken. mkiv works ok. BTW: Is it possible to only build mkiv and the english interface? Maybe by passing some options to first-setup.sh? I never need other interfaces and hardly ever mkii. Marco

On 1-2-2011 4:41, luigi scarso wrote:

On Tue, Feb 1, 2011 at 4:34 PM, Marco wrote:

...

BTW: Is it possible to only build mkiv and the english interface? Maybe by passing some options to first-setup.sh? I never need other interfaces and hardly ever mkii.

Marco In first-septup.sh , near the bottom mtxrun --verbose --script ./bin/mtx-update.lua --update --make --force --engine=all --context=beta --platform=$platform --texroot=$PWD/tex $@

I believe that --engine=luatex is the switch you want

indeed, but one needs a fresh minimal because the current state is stored and kept ----------------------------------------------------------------- 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 -----------------------------------------------------------------

Am 01.02.2011 um 16:47 schrieb Hans Hagen:

On 1-2-2011 4:41, luigi scarso wrote:

...

On Tue, Feb 1, 2011 at 4:34 PM, Marco wrote:

...

BTW: Is it possible to only build mkiv and the english interface? Maybe by passing some options to first-setup.sh? I never need other interfaces and hardly ever mkii.

Marco In first-septup.sh , near the bottom mtxrun --verbose --script ./bin/mtx-update.lua --update --make --force --engine=all --context=beta --platform=$platform --texroot=$PWD/tex $@

I believe that --engine=luatex is the switch you want

indeed, but one needs a fresh minimal because the current state is stored and kept

But one need the other formats to use TeX Gyre fonts with math? If I try the –engine=luatex switch this \setupbodyfont[palatino] \starttext $a+b$ \stoptext returns: !LuaTeX error: cannot open Type 1 font file for reading (qplri.pfb) Greeting Andreas

Just uploaded latest beta and on "context --make en" got the error: ! I can't find file `catc-ini.mkiv'. <to be read again> \relax l.76 \loadmarkfile{catc-ini} Are there still missing files? Hans van der Meer

Confirmed - I got the same error yesterday evening on my linux system (after a successful installation in the afternoon on OS X). Jörg On 2/2/11 11:25 AM, luigi scarso wrote:

On Wed, Feb 2, 2011 at 11:18 AM, Hans van der Meer wrote:

...

Just uploaded latest beta and on "context --make en" got the error:

! I can't find file `catc-ini.mkiv'. <to be read again> \relax l.76 \loadmarkfile{catc-ini}

Are there still missing files?

Hans van der Meer not here from minimals

On Wed, Feb 2, 2011 at 11:43 AM, Hans van der Meer wrote:

Could it be because it is named catc-ini.lua instead of catc-ini.mkiv? Or has it to be loaded as an explicit lua-file? no. What about

$> mktexlsr $> kpsewhich catc-ini.mkiv ? -- luigi

On Wed, Feb 2, 2011 at 11:57 AM, Hans van der Meer wrote:

Strange, doing (running MacOSX 10.6.6, beta is the directory in which context is installed):

102 beta: mktexlsr mktexlsr: Updating /Users/hans/TeX/ConTeXt/beta/ls-R... mktexlsr: Done. 103 beta: kpsewhich catc-ini.mkiv /Users/hans/TeX/ConTeXt/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv 104 beta:

File is there. Checked again that "context --make en" runs into the same error. delete the cache and $> mtxrun --generate -- luigi

More weird things I am afraid. Experimenting somewhat, I removed the texmf-cahce and did mtxrun --generate. The first messages is: "/Users/hans/TeX/ConTeXt/beta" and its permissions are: "drwxrwxr-x 7 hans staff 238 2 feb 17:48 beta/" idem of the parent directory: "drwxr-xr-x 13 hans staff 442 2 feb 10:36 ../" 120 beta: mtxrun --generate shell-init: error retrieving current directory: getcwd: cannot access parent directories: No such file or directory Something wrong here? Something I can/should repair? Hans van der Meer On 2 feb 2011, at 12:24, Hans van der Meer wrote:

Did remove texmf-cache and ran mtxrun --generate. Still get on context --make en the error: ! I can't find file `catc-ini.mkiv'. <to be read again> \relax l.76 \loadmarkfile{catc-ini}

Please type another input file name:

(will be away 1 or 2 hours, sorry not able to answer immediately) Hans van der Meer

On 2 feb 2011, at 12:05, luigi scarso wrote:

...

On Wed, Feb 2, 2011 at 11:57 AM, Hans van der Meer wrote:

...

Strange, doing (running MacOSX 10.6.6, beta is the directory in which context is installed):

102 beta: mktexlsr mktexlsr: Updating /Users/hans/TeX/ConTeXt/beta/ls-R... mktexlsr: Done. 103 beta: kpsewhich catc-ini.mkiv /Users/hans/TeX/ConTeXt/beta/tex/texmf-context/tex/context/base/catc-ini.mkiv 104 beta:

File is there. Checked again that "context --make en" runs into the same error. delete the cache and $> mtxrun --generate --

On Wed, Feb 2, 2011 at 8:07 PM, Hans van der Meer wrote:

More experiments, I am not one to be so easily defeated by software. Retrieved some old beta's, each time I start a context run then it results in the loadfilemarkerror. But I can find several formats lying around: /Users/hans/TeX/ConTeXt/beta/tex/texmf-cache/luatex-cache/context/c180539b697aa06758debbc1182817e4/formats/cont-en.fmt /Users/hans/TeX/ConTeXt/beta/tex/texmf-cache/luatex-cache/context/c180539b697aa06758debbc1182817e4/formats/cont-nl.fmt /Users/hans/TeX/ConTeXt/beta/tex/texmf-osx-64/web2c/pdftex/cont-en.fmt /Users/hans/TeX/ConTeXt/beta/tex/texmf-osx-64/web2c/pdftex/cont-nl.fmt /Users/hans/TeX/ConTeXt/beta/tex/texmf-osx-64/web2c/xetex/cont-en.fmt /Users/hans/TeX/ConTeXt/beta/tex/texmf-osx-64/web2c/xetex/cont-nl.fmt Apparently none of them found! Vide this first message: MTXrun | warning: no format found, forcing remake (commandline driven) It's a Mac OS system (sorry, I though it was linux).

Can you try the latest minimals under /Users/hans/LuaTeX (just to be sure to avoid conflicts) ? -- luigi

On 2-2-2011 11:18, Hans van der Meer wrote:

Just uploaded latest beta and on "context --make en" got the error:

! I can't find file `catc-ini.mkiv'. <to be read again> \relax l.76 \loadmarkfile{catc-ini}

Are there still missing files?

no, i'm still trying to get mkii and mkiv working at the same time with a fundamental different setup strategy .. getting there 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 2-2-2011 1:41, Peter Münster wrote:

Hans Hagen writes:

...

no, i'm still trying to get mkii and mkiv working at the same time with a fundamental different setup strategy .. getting there

Or perhaps 2 completely different distributions? mkii: frozen, only some fixes from time to time mkiv: no more mkii files, no more pdftex, just pure mkiv

this can already be done (just install with the --engine=luatex flag) the main issue is that mtxrun also replaces the ruby texmfstart script and texexec is launched via mtxrun; so mtxrun needs to be able to deal with mkii situations where kpse is used for lookups later on; as the installer uses the --tree option, and as mkiv has an abstract notion of the texroot, some hoops have to be jumped through to get it working for both machineries ----------------------------------------------------------------- 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 2-2-2011 3:01, Vedran Miletić wrote:

2011/2/2 Hans Hagen

...

On 2-2-2011 1:41, Peter Münster wrote:

...

Hans Hagen writes:

no, i'm still trying to get mkii and mkiv working at the same time with a

...

fundamental different setup strategy .. getting there

Or perhaps 2 completely different distributions? mkii: frozen, only some fixes from time to time mkiv: no more mkii files, no more pdftex, just pure mkiv

this can already be done (just install with the --engine=luatex flag)

the main issue is that mtxrun also replaces the ruby texmfstart script and texexec is launched via mtxrun; so mtxrun needs to be able to deal with mkii situations where kpse is used for lookups later on; as the installer uses the --tree option, and as mkiv has an abstract notion of the texroot, some hoops have to be jumped through to get it working for both machineries

You mentioned earlier that texexec could be replaced by mtxrun completely, that only a bit is missing. Would this solve the problem?

it would solve the ruby dependency but pdftex/xetex need kpse 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, Feb 2, 2011 at 13:41, Peter Münster wrote:

...

no, i'm still trying to get mkii and mkiv working at the same time with a fundamental different setup strategy .. getting there

Or perhaps 2 completely different distributions? mkii: frozen, only some fixes from time to time

mkiv: no more mkii files, no more pdftex, just pure mkiv

Yes, I like this idea. I don't care about mkii, nor pdftex.

On 2 feb 2011, at 20:08, Peter Münster wrote:

Vnpenguin writes:

...

On Wed, Feb 2, 2011 at 13:41, Peter Münster wrote:

...

...

no, i'm still trying to get mkii and mkiv working at the same time with a fundamental different setup strategy .. getting there

Or perhaps 2 completely different distributions? mkii: frozen, only some fixes from time to time mkiv: no more mkii files, no more pdftex, just pure mkiv

Yes, I like this idea. I don't care about mkii, nor pdftex.

In the first place, this idea was meant to make Hans' life easier. (not the user's life... ;-)

I strongly support the idea of freezing mkii, apart from repairing obvious bugs in the existing codebase. It will free precious resources and save valuable time. Moreover imho concentrating on mkiv not hampered by compatability issues because of mkii, can only add to its stability. (It is just like evolution or maybe call it progress. Imagine both our primate ancestors and the dinosaurs being maintained at the same time? Where would we be? ;-) Hans van der Meer