Re: [Pan-users] disabling spell-checking at run-time?

[Duncan]

On Friday 25 October 2002 00:57, Søren Boll Overgaard wrote:

> It's Debian, and I am the one building the packages. I have had requests
> for packages with as well as without gtkspell support, and instead of
> building two packages, it would be nice if spellchecking could be
> disabled at runtime. If not, I guess I will have to look into building
> two separate packages.

Is it possible part of the requests for nospell versions is because the spell 
enabled versions crash when people attempt to post?  I know both LeMay and I 
(as well as others) have had issues with gtkspell (on Mandrake) where that's 
the problem, and I had nospell versions installed for a time as a result.  
I'm guessing that if Debian users are experiencing the same issues, it's 
possible part of your nospell requests are based on spell enabled versions 
crashing, rather than on any specific dislike of spell checking...

BTW, despite the so-called dependency on gtkspell 2.0.2, I have 0.13.1 running 
apparently fine on Mdk-Cooker/9.0 urpmi updated libgtkspell0-2.0.0, while it 
crashes on attempt to create a post, with the gtkspell-2.0.2 available thru 
rebelbase/download/extras.

Also, since I've got your attention Soren (what a coincidence, this involves 
you, too), I've been experiencing the spacebar scrolling bug (that you filed 
the bugzilla report on) myself -- on posts that look to be 100% us-ascii 
coded, with that specified in the text/encoding headers, and nothing stranger 
than @ (at) and " (quote) symbols in the From header, nothing other than 
spaces and ASCII letters in the subject, and nothing unusual anywhere else I 
could see.  If it's an encoding mismatch, it seems to affect usascii posts on 
iso-8859-1 systems, as well, and that's pretty generic, AFAIK. (I'd be 
perfectly happy with us-ascii here, but Mdk installed iso-8859-1, and like 
you, I don't know enough about it to worry about it to much, as long as 
pretty much everything keeps working, as it has.)

Just a note to say it doesn't seem to be ONLY strangely coded umlautes or 
whatever, doing it.  Unfortunately, the particular messages I'm looking at 
right now, with the problem, are on Cox internal only newsgroups.  However, 
I'll save a sample, if you or the developers think it'd contribute anything 
to that bug report.

-- 
Duncan
"They that can give up essential liberty to obtain a little
temporary safety, deserve neither liberty nor safety." --
Benjamin Franklin