Hello: Please find attached my comments below:
El Martes, 30 de Enero de 2007, escribió:
> Package: kword
> Version: 1:1.6.1-2
> Severity: important
I see you were on testing and current koffice version in testing is newer
than the one you are stating in the bug report. So it would be great if you
could test and report back with the current testing (or even unstable)
version.
[...]
>
>
> For this reason, I don't agree at all with bug #404683. It isn't
> possible to have a per-document language setting in the cases above.
>
> kwrite does the right thing, kudos! (step 4 below)
I have checked this in koffice and it seems to store language on a font
style basis. This is, each font style has a property that defines the
language, and the right spell checker will be user for that,
BTW, if you have something to say about that bug, why not commenting it
out? ;)
>
> Alas, the feature is bugged in kedit (step 5 below). And badly broken
> in kword (step 6 below).
>
> Details:
>
[...]
> 2. control centre -> KDE components -> spell checker
> The drop-down tab says "ispell default".
> Change this to "Unknown [ngerman]", encoding 8859-1.
> Apply. Quit.
>
> (BUG: the name should be better than this. End-users will not be
> impressed that they have to pick "unknown" dictionaries.)
>
Is the name corrected on a recent version? if not, please open a separate
bug report.
[...]
> 5. k -> utilities -> editors -> kedit
> Open file "test" saved at step 3.
> On the kedit menu bar, click tools -> spelling. The tab shows british,
> the program objects to "Und". Correct so far.
>
> Click on the language tab and change to "Unknown [ngerman]".
> Click "okay" (You have to restart the dialogue for changes to take
> effect) On the kedit menu bar, again click tools -> spelling.
>
> BUG: The tab shows the changed language (ngerman), but the program is
> still using the previous value (british).
This is working for me on 3.5.8, but if it is not for you on a recent
version, please open a separate bug report.
[...]
> 6. k -> office -> kword
> import file "test" (filter locale [8859-1])
>
> click on spellcheck button (or click tools -> spellcheck -> spelling...)
>
> BUG: The languages shown in the language tab are not related to the
> installed languages. ingerman does not show up at all.
>
> I have a list of about 29 entries matching regexp en.* (en_GB and so on)
> and at the end an entry "he" (for Hebrew, which I've never installed).
>
> It isn't at all clear whether kword is trying to integrate with the
> rest of kde or not. Personally, I think it should behave like kwrite.
Kword relies on other spell checkers like aspell, for instance, so if you
don't have the language you expect, maybe the problem is that you haven't
installed the right spell checker dictionary.
>
> click "he", then "ignore" to advance the checking to the next word...
> BUG: kword crashes with signal 6 (SIGABRT)
>
> I attach a backtrace, in case it helps, but the bug seems perfectly
> reproducible. I also notice a snippet (similar to the attached) at the
> end of ~./xsession-errors
What you did choosing "he" is selecting the hebrew spell checker which is
hspell. According to the backtrace it was hspell that crashed, and not kword.
Thanks,
--
Raúl Sánchez Siles
----->Proud Debian user<-----
Linux registered user #416098
signature.asc
Description: This is a digitally signed message part.
