Evolution and 2.5.x

Evolution and 2.5.x

[Eric Blade]

Greetings,
 
  I'm guessing that not too many of the kernel developers use Evolution as
their email program :)   Since I started picking up the 2.5.x series, at around
2.5.34, Evolution does not run anywhere near properly.  I'm not sure if that
is a kernel issue, or a problem with Evolution's code..  But it did improve
quite a bit with all the low-level process management that was in the 2.5.3x 
series.  It still doesn't work right though.  (in 2.5.34, evolution would
just plain halt the system ... in 2.5.42, it mostly works right, as long
as you don't try to compose a message.. composing a message will leave you
with a whole buch of zombie processes). 
 

Re: Evolution and 2.5.x

[Robert Love]

On Sun, 2002-10-13 at 14:54, Eric Blade wrote:

>   I'm guessing that not too many of the kernel developers use Evolution as
> their email program :)   Since I started picking up the 2.5.x series, at around
> 2.5.34, Evolution does not run anywhere near properly.  I'm not sure if that
> is a kernel issue, or a problem with Evolution's code..  But it did improve
> quite a bit with all the low-level process management that was in the 2.5.3x 
> series.  It still doesn't work right though.  (in 2.5.34, evolution would
> just plain halt the system ... in 2.5.42, it mostly works right, as long
> as you don't try to compose a message.. composing a message will leave you
> with a whole buch of zombie processes). 

Hey, I use Evolution ;-)

See this thread:
http://lists.ximian.com/archives/public/evolution-hackers/2002-June/004841.html

It is indeed broken in 2.5 and it is not, for once, our fault.  This
thread and other discussion seem to point out it is a bug in ORBit.

	Robert Love

Re: Evolution and 2.5.x

[Eric Blade]

Robert,
 
  Thanks for the response!  Honestly, after reading through the thread, it 
seems as if it's more of a "we don't care if it's kernel our evolution's 
problem, until after 2.6 happens, but either way we don't plan to track it 
down and fix it".  

Re: Evolution and 2.5.x

[Andy Pfiffer]

On Sun, 2002-10-13 at 12:03, Robert Love wrote:
> On Sun, 2002-10-13 at 14:54, Eric Blade wrote:
> 
> >   I'm guessing that not too many of the kernel developers use Evolution as
> > their email program :)   Since I started picking up the 2.5.x series, at around
> > 2.5.34, Evolution does not run anywhere near properly.  I'm not sure if that
> > is a kernel issue, or a problem with Evolution's code..  But it did improve
> > quite a bit with all the low-level process management that was in the 2.5.3x 
> > series.  It still doesn't work right though.  (in 2.5.34, evolution would
> > just plain halt the system ... in 2.5.42, it mostly works right, as long
> > as you don't try to compose a message.. composing a message will leave you
> > with a whole buch of zombie processes). 
> 
> Hey, I use Evolution ;-)
> 
> See this thread:
> http://lists.ximian.com/archives/public/evolution-hackers/2002-June/004841.html
> 
> It is indeed broken in 2.5 and it is not, for once, our fault.  This
> thread and other discussion seem to point out it is a bug in ORBit.
> 
> 	Robert Love

I spent some time trying to track this problem down, but reached a wall
due to the size and nature of the ChangeSet.

The bitkeeper ChangeSet that made Evolution's address book hang when
trying to compose a new message when run on 2.5.x kernels was 1.262.2.2.

According to my notes from June, that ChangeSet comprised modifications
to 116 files, splitting the socket structure from the protocol.  Up to
that ChangeSet, the problem does not exist.  After applying that
ChangeSet, the problem is manifest.

Sorry I couldn't dig deeper...and I hope this helps someone.

Andy

Re: Evolution and 2.5.x

[Thomas Molina]

> > See this thread:
> > http://lists.ximian.com/archives/public/evolution-hackers/2002-June/004841.html
> > 
> > It is indeed broken in 2.5 and it is not, for once, our fault.  This
> > thread and other discussion seem to point out it is a bug in ORBit.
> > 
> > 	Robert Love
> 
> I spent some time trying to track this problem down, but reached a wall
> due to the size and nature of the ChangeSet.
> 
> The bitkeeper ChangeSet that made Evolution's address book hang when
> trying to compose a new message when run on 2.5.x kernels was 1.262.2.2.

I've read this thread and I'm confused.  Is this seen as a problem with 
Evolution, ORBit, or the 2.5 kernel?  If it is seen as a possible kernel 
problem, I'll add it to my problem report status page and track it.  If I 
track it, Eric Blade will get a weekly email asking whether he's still 
seeing the problem, at least until I'm told to drop it, or no one 
responds.

Re: Evolution and 2.5.x

[Robert Love]

On Mon, 2002-10-14 at 15:41, Thomas Molina wrote:

> I've read this thread and I'm confused.  Is this seen as a problem with 
> Evolution, ORBit, or the 2.5 kernel?  If it is seen as a possible kernel 
> problem, I'll add it to my problem report status page and track it.  If I 
> track it, Eric Blade will get a weekly email asking whether he's still 
> seeing the problem, at least until I'm told to drop it, or no one 
> responds.

It is a problem with ORBit.

Behavior changed somewhere in the kernel - I think a return value is
different but was not standard before and ORBit relied on it.

	Robert Love

Re: Evolution and 2.5.x

[Andy Pfiffer]

On Mon, 2002-10-14 at 12:41, Thomas Molina wrote:
> > > See this thread:
> > > http://lists.ximian.com/archives/public/evolution-hackers/2002-June/004841.html
> > > 
> > > It is indeed broken in 2.5 and it is not, for once, our fault.  This
> > > thread and other discussion seem to point out it is a bug in ORBit.
> > > 
> > > 	Robert Love
> > 
> > I spent some time trying to track this problem down, but reached a wall
> > due to the size and nature of the ChangeSet.
> > 
> > The bitkeeper ChangeSet that made Evolution's address book hang when
> > trying to compose a new message when run on 2.5.x kernels was 1.262.2.2.
> 
> I've read this thread and I'm confused.  Is this seen as a problem with 
> Evolution, ORBit, or the 2.5 kernel?  If it is seen as a possible kernel 
> problem, I'll add it to my problem report status page and track it.  If I 
> track it, Eric Blade will get a weekly email asking whether he's still 
> seeing the problem, at least until I'm told to drop it, or no one 
> responds.

I'll defer to the experts as to the root cause of the problem.

All I know for sure is that before 1.262.2.2 Evolution works, and after
applying 1.262.2.2, it doesn't.

Andy

Re: Evolution and 2.5.x

[Eric Blade]

On Mon, 2002-10-14 at 15:41, Thomas Molina wrote:
> I've read this thread and I'm confused.  Is this seen as a problem with 
> Evolution, ORBit, or the 2.5 kernel?  If it is seen as a possible kernel 
> problem, I'll add it to my problem report status page and track it.  If I 
> track it, Eric Blade will get a weekly email asking whether he's still 
> seeing the problem, at least until I'm told to drop it, or no one 
> responds.


I'm fairly well confused, as well.  I don't know if it's kernel, or
Evolution, or ORBit.  The Evolution people don't seem to care all that
much - it's broken, it's a development kernel, they'll straighten it out
later.  *shrug*  I don't see much point to getting a weekly email about
it, if no one's working on the problem?  I don't understand anywhere
near enough to look into it.
 
 - Eric