Re: GNOME apps crash with non-root user

 > Is there a reason *why* you chose to install beta code _on top of_
 > GA code?

For the same reason that anyone else would. To try it out and test it. 
Isn't that the purpose of a beta program?

 > 1) Beta code is unsupported. (See the RELEASE NOTES about on using
 > on mission-critical systems).

I'm not asking for support, and I'm not using a mission-critical system. 
I just wanted to know if anybody had experienced the same problem. 
Besides, I thought we were supposed to report bugs. You make it sound 
like I should have just kept my mouth shut. I don't think so, since a 
bug in Phoebe could be a bug in 8.1, as well, and nobody wants that.

 > 2) If you install applications on a pristine beta system, it helps
 > the developers to see how these applications interact

I haven't installed any new applications. I'm using only the standard 
GNOME stuff that was upgraded by Phoebe.

 > 3) OTH, installing beta code on top of stable code, only "muddies
 > the waters", and merely wastes the developers time hunting for bugs

I am not mixing beta code with stable code. I am not using Rawhide 
packages on top of 8.0. I simply upgraded the entire system with Phoebe. 
I realize that it is a "beta", but if 8.1 is coming out in April, then 
8.0.94 (which I used) is really a release candidate rather than a beta. 
Upgrading from 8.0 to 8.0.94 should not cause segfaults, as long as only 
the standard packages are used, which is what I'm doing. Or are you 
saying that 8.0 cannot be upgraded to 8.1 and should be installed from 
scratch, lest it "muddy the waters"? I don't think that's going to be a 
restriction in 8.1, so it shouldn't be a restriction in 8.0.94, either.

 > Your post appears to imply that your initial login was _as root_,
 > rather than logging into a user account.

No, I always log in as a standard user. I only tried logging in as root 
when I was unable to work around the crashes.

Perhaps I'm being unforgiving here, but I believe any segfault is a bug. 
A corrupted executable or hardware failure are clear exceptions to this 
rule, but otherwise crashes should never be blamed on the user. In my 
experience, they're always the result of lazy programming -- not 
checking for NULL after calling malloc, for example, or making system 
calls without proper type checking of the parameters. I know I make 
those mistakes myself on occasion. :)

Please understand that I am not asking for support or demanding that 
these bugs be fixed. I just wanted to see if my problems were unique, 
and I thought I should point them out anyway, since a crash is a fairly 
critical error.

Trevor



-- 
Phoebe-list mailing list
[email protected]
https://listman.redhat.com/mailman/listinfo/phoebe-list