[texworks] external read-access causes TeXWorks to revert open file to old version

[Paul A Norman]

Stefan,

We may need a TeXWorks developer to step in here, but I know they are
very cautious and thorough in their development procedures and I
respect them for that, so it would seem a little unlikely that
TeXWorks by itself would be the issue here ...

Do you have a completely different machine you can try and replicate this from?

This part of your original posting makes me think that it is possibly
a caching issue of commits from RAM to your hardrive.

" most of the times TeXWorks restores an version of the open file
which is older than the saved document"

I assume that the apparent "version" of the document has exactly the same name?

The bit about TeXWorks closing itself - makes me think about RAM
pirarcy going on in the memory processes.

At a guess, a few things point to some sort of possible memory
violation on your machine perhaps?

Has there been any history of RAM memory problems on your machine at all please?

Also, there are newer  Windows versions of TexWorks available from
Alian Delmot's  here (marked development but quite stable):

http://www.leliseron.org/texworks/      You'll notice that the number
is v0.3  r606

And if you are happy to build it for other OS's, links are here:
http://code.google.com/p/texworks/

Paul




On 22 September 2010 22:31, Stefan Müller <warrence.stm at gmx.de> wrote:
> Thanks for the fast reply,
>
> I'm working on Linux Fedora 13 (it happened in older Fedora versions, too) and
> Windows Vista. I encounter the problem on both systems. Version of TeXWorks is
> 0.2.3 r466 (most current version I found).
>
> During my daily work I have some applications open. Firefox, Google Chrome,
> Pidgin, Kontact and Terminals are normally running. None of them messes with
> the files under consideration. TortoiseHG is my Mercurial client, under Linux
> its name is "hgtk", so naturally "hgtk commit" is running and that's how I
> commit things. Maybe this client is causing the problems, but it's hard for me
> to test with the console client (hg) because of some missing feedback and
> pretty inconvenient handling. But I'll try that.
>
> Stefan.
>
>> Thanks for that Stefan,
>>
>> Just the regular things please -- what OS are you on - version and
>> revision of TexWorks please.
>>
>> Is any other application "open" when this happens?
>>
>> What client are you using for Mercurial please - how do you 'comit'
>> your work/files?
>>
>> Thanks,
>>
>> paul
>>
>> On 22 September 2010 05:27, Stefan M?ller <warrence.stm at gmx.de> wrote:
>> > Hi list,
>> >
>> > I am experiencing some strange problems when using TeXWorks together with
>> > the subversion tool Mercurial. Currently, my opinion is that TeXWorks is
>> > causing the trouble, but I may be wrong.
>> >
>> > I open a .tex-document in TeXWorks, change something, save the file and
>> > commit the change in Mercurial while TeXWorks is still open. The
>> > committed changes are stored in separate files, the .tex-file itself is
>> > not changed during a commit. In about half of the commits something goes
>> > wrong in TeXWorks [Why? The commit is in Mercurial, TeXWorks should have
>> > nothing to do with it...]. Sometimes TeXWorks just closes itself, which
>> > is annoying but not destructive. Unfortunately most of the times
>> > TeXWorks restores an version of the open file which is older than the
>> > saved document. If I close TeXWorks at this point no harm is done,
>> > because this "revert" is not saved [and I am not asked if I want to
>> > save]. If I don't discover this "revert" in TeXWorks and continue with
>> > more changes and save, I wreck my .tex-file with old parts of the
>> > document.
>> >
>> > This issue sometimes even eats new stuff in my document; when I am not
>> > ready to commit the new part but commit a different section, the new
>> > part may be "reverted" to an old version where the new part is not
>> > there.
>> >
>> > I hope I clearly outlined the issue. I am in desperate hope that someone
>> > can give me advice on how to fix this, as I already lost several days of
>> > work because of this.
>> >
>> > Best regards,
>> > Stefan.
>