Re: [evolution-patches] fix for bug #40074 (and the newer #42170)

[Mike Kestner <mkestner ximian com>]

On Wed, 2003-04-30 at 20:53, Not Zed wrote:

> we're gonna dump a whole tree's state to disk before rebuilding the
> tree, then reload it from disk again?

Technically you are dumping a record of which nodes are different from
the default state.  It wouldn't be hard to add an API to dump and load
from string if that would help. 

> there has to be a better way ...

Yeah, not deleting the whole tree and rebuilding it.  And you're worried
about loading a state file?
 
> etable used to keep this state itself, but it doesn't now :-/

I hear it used to crash a lot more, too.  And keeping expanded state
across a clear operation isn't exactly behavior I anticipated having to
provide.  :-/

> we've always been rebuilding the whole tree because etable is/has always
> been too buggy to make incremental changes to the tree.
> 
> why was it that we didn't port to gtktreeview again?

I know this is passive aggressive wiseass shit,  but why was it that you
didn't try turning off the BROKEN_ETREE crap and work with me while I've
been trying to fix its problems?
  
Mike