-----BEGIN PGP SIGNED MESSAGE-----
Am Sa den 16. Feb 2008 um 20:13 schrieb Sven Neumann:
> > Ehem, why didn't you do that before?
> Because it is very unlikely that the user directory as specified in the
> passwd is not present.
Oh, that is very common. It could be intentional or not but it is very
It is intentional if you generate the users $HOME in pam session state
where the home directory in passwd is just a dummy.
It might unintentional if the file serve is gone but you want to work
either using /tmp/username as $HOME.
> This is different for the value of the HOME environment variable. The
> user might easily mess this up.
The variable can be set by pam or similar thinks or can be fixed by the
user to keep working if the normal home is not available. More over it
can be set to a other directory if you have need to. Eg. you have no
space left in your home or you want to test a new version without
poisoning your current gimp settings.
And, by the way, not respecting $HOME was always a bug as it is
- From "man gimp":
Most GIMP configuration is read in from the user's init file,
$HOME/.gimp-2.4/gimprc. The system wide
equivalent is in /etc/gimprc. The system wide file is parsed first
and the user gimprc can override the
system settings. /etc/gimprc_user is the default gimprc placed in
users' home directories the first time
GIMP is run.
$HOME/.gimp-2.4/devicerc - holds settings for input devices together
with the tool, colors, brush, pattern
and gradient associated to that device.
$HOME/.gimp-2.4/documents - lists all images that have been opened or
saved using GIMP.
> Why are you arguing at all? The problem you pointed out hasn't bothered
> anyone but you in the last twelve years.
Oh, it has. I know many people who told me that this is completely shit. I just
get so bothered that I try to fix the bug.
> The natural response to your > request would have been to turn it down.
Ah yes. Sorry, but aren't you able to be open for constructive critic?
Maybe you note that I not only told that this is shit. I provide you
with a working and tested patch and I gave some reasons why this is a
bug. Can you give me just one reason why to use the passwd entry and not
accepting user choice? (Except of that it is implemented this way in
> In particular because it is based on the wrong assumption that every
> other application would behave differently. That is wrong as can be
> easily seen by doing a code search for g_get_home_dir().
Oh, sorry, I just checked the applications I know about. And not all but
several. And there are graphical applications and non graphical
> Instead of turning your request down we expressed interest to consider
> your patch if you can rework it a little.
Oh, I'm open for it. This patch is just how I get around this bug for
> It would help a lot if you could simply do that.
Ok, I will. Since now I just understand that YOU have to chance some to
put that patch in. Sorry for the misunderstanding.
Ps. I hope not to get away to much from a objective explaining in this
Klaus Ethgen http://www.ethgen.de/
pub 2048R/D1A4EDE5 2000-02-26 Klaus Ethgen <[EMAIL PROTECTED]>
Fingerprint: D7 67 71 C4 99 A6 D4 FE EA 40 30 57 3C 88 26 2B
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
-----END PGP SIGNATURE-----
Gimp-developer mailing list