mailing-list for TeXmacs Users

[Jan Kuhfeld <address@hidden>]

Hello Jeroen,

Thanks a lot. Removing interactive.scm solved the problem (actually 
commenting the line mentioned in the error message was enough). Here is 
the (shortened) content of the file:

- start interactive.scm -

((load-buffer
   (("0"
     system->url
     "/home/jk/somepath/somefile.tm")))
 (recent-buffer
   (("0"
     .
     "tmfs://help/article/tm/doc/about/welcome/new-welcome.en.tm")))
 (execute-command
   (("0" . #<eof>))     <-------- problematic line
   (("0" . "w"))
   (("0" . "q")))
 (save-buffer-as
   (("0" system->url "/home/jk/Documents/scratch.tm"))))

- end interactive.scm -

I removed all items under recent-buffer except for one, therefore the 
line numbers do not correlate anymore with the error message. But the 
problematic line is marked.

Thanks again,
Jan


On 15.09.21 21:30, Jeroen Wouters wrote:
> Hello Jan,
>
> There was a similar report recently on the forum:
> http://forum.texmacs.cn/t/texmacs-falling-into-infinite-crash-loop/416/3 
> <http://forum.texmacs.cn/t/texmacs-falling-into-infinite-crash-loop/416/3>
>
> The file ~/.TeXmacs/system/interactive.scm probably got corrupted somehow.
>
> You can move that file to another location and retry running TeXmacs.
>
> It would be good if you could post that file's contents, perhaps it will 
> reveal something about how it got corrupted.
>
> Best wishes,
> Jeroen
>
> On 15 September 2021 09:26:43 BST, Jan Kuhfeld <address@hidden> wrote:
>
>
>     Hello everyone,
>
>     for a while now, I am happily using TeXmacs under Archlinux (thank you
>     for this nice program).
>
>     Sadly, since a recent system update I am not able to start TeXmacs
>     anymore. Running texmacs -d in the terminal results in following error
>     message:
>
>     -Start error message-
>
>           Installing internal plug-ins...
>           Welcome to TeXmacs 2.1
>                                         
>           kpsepath works with your TeX distribution
>           kpsewhich works with your TeX distribution
>           mktextfm works with your TeX distribution
>           mktexpk works with your TeX distribution
>           texhash works with your TeX distribution
>           attempt install of inkscape extension
>
>           Installation completed successfully !
>           I will now start up the editor
>
>           Opening display...
>           Screen extents: 2736 x 1824
>           Starting server...
>           TeXmacs] debug-std, Evaluating
>     /usr/share/TeXmacs/progs/init-texmacs.scm...
>           /usr/share/TeXmacs/progs/kernel/boot/abbrevs.scm:81:12: In
>     procedure scm_lreadr in expression (read (open-file # OPEN_READ)):
>           /usr/share/TeXmacs/progs/kernel/boot/abbrevs.scm:81:12:
>     /home/jk/.TeXmacs/system/interactive.scm:59:14: Unknown # object: #\<
>           Loading welcome message...
>           Creating 'no name' buffer...
>           TeXmacs] With linked TrueType support
>           TeXmacs] debug-std, Redefined color dark slate
>           TeXmacs] debug-std, #2F4F4F replaced by #483D8B
>           TeXmacs] debug-std, Redefined color light slate
>           TeXmacs] debug-std, #778899 replaced by #8470FF
>           TeXmacs] debug-std, Redefined color light goldenrod
>           TeXmacs] debug-std, #FAFAD2 replaced by #EEDD82
>           TeXmacs] debug-std, Evaluating
>     /usr/share/TeXmacs/progs/init-buffer.scm...
>           /usr/share/TeXmacs/progs/init-buffer.scm:15:9: In expression
>     (buffer-missing-style?):
>           /usr/share/TeXmacs/progs/init-buffer.scm:15:9: Unbound variable:
>     buffer-missing-style?
>           Throwing widget should be attached
>     ------------------------------------------------------------------------
>           Error message:
>             widget should be attached
>
>           System information:
>             TeXmacs version  : 2.1
>             Built by         : builduser
>             Building date    : 2021-06-23T11:29:57
>             Operating system : linux-gnu
>             Vendor           : pc
>             Processor        : x86_64
>             Crash date       : Mi 15. Sep 09:52:39 CEST 2021
>
>           TeXmacs does not yet have a current view
>           Backtrace of C++ stack:
>             texmacs.bin : () + 0x467e8f
>             texmacs.bin : () + 0x2a9494
>             texmacs.bin : () + 0x453ced
>             texmacs.bin : () + 0x45447b
>             texmacs.bin : () + 0x45c4ba
>             texmacs.bin : () + 0x45e181
>             texmacs.bin : () + 0x257dba
>             /usr/lib/libguile.so.17 : () + 0x5dd06
>             /usr/lib/libguile.so.17 : () + 0x35fe5
>             /usr/lib/libguile.so.17 : scm_c_catch() + 0x163
>             /usr/lib/libguile.so.17 : scm_i_with_continuation_barrier() + 0xb0
>             /usr/lib/libguile.so.17 : scm_c_with_continuation_barrier() + 0x44
>             /usr/lib/libguile.so.17 : scm_i_with_guile_and_parent() + 0x37
>             /usr/lib/libguile.so.17 : scm_with_guile() + 0x11
>             /usr/lib/libguile.so.17 : scm_boot_guile() + 0x36
>             texmacs.bin : () + 0x23ad47
>             /usr/lib/libc.so.6 : __libc_start_main() + 0xd5
>             texmacs.bin : () + 0x256a6e
>     ------------------------------------------------------------------------
>           terminate called after throwing an instance of 'string'
>           zsh: abort (core dumped)  texmacs -d
>
>     -End error message-
>
>     My first guess is that this is related to an update of Qt5. Both,
>     because I remember that Qt5 was in the list of updated packages (there
>     were multiple updates to the qt5-base package and probably other qt5
>     related packages in the last days) and because of the "widget should be
>     attached" message. I already tried rebuilding TeXmacs (with the PKGBUILD
>     of the official arch repo), but the problem stayed. I am willing to do
>     more tests, if you need additional information, and of course I can test
>     any patches on my system.
>
>     Thank you again for all your work on TeXmacs!
>
>     Jan
>
>     PS: Please let me know, if these kind of bug reports should better be
>     posted on this mailing list or on the dev list.