On Sat, 2008-09-27 at 17:14 -0700, Nifty Fedora Mitch wrote:
On Mon, Sep 22, 2008 at 07:43:22PM -0430, Patrick O'Callaghan
wrote:
> On Mon, 2008-09-22 at 23:49 +0000, Paul W. Frields wrote:
> > > PS "Works" is a relative term here. "Mostly works" is
closer to the
> > > truth because Evo occasionally loses contact with Gmail and has to
> > be
> > > restarted to get it back. I don't know if this is an Evo problem or
> > a
> > > Gmail one.
> >
> > I don't have the problem using Evo with any other of my IMAP sources,
> > so
> > I'm guessing it's Gmail related.
>
> I tend to agree. I just wish I could figure out what to do about it
> since it's pretty annoying. I have to restart Evo several times a day
> because of this (I normally leave it running permanently).
I suspect but have not yet proved to myself that 'your' server moves and
the connection fails because the google box is different. Given the size
of the google farm and the numerous servers and gateways and network
magic that 'your' server moves and the connection fails because the
google box is different. Some NAT tricks come to mind that make this possible
but very hard to track from this side of the gateway.
I often see gmail connections die even with 'mutt' and have started to
sync my mail box prior to a reply. Often after a well considered email
(=lots of time editing) I find that my connection to the mailbox is gone.
You may find this interesting:
http://groups.google.com/group/comp.mail.mutt/browse_thread/thread/d24f6f...
Executive summary: for Mutt, reduce imap_keepalive to under 10 minutes.
For Evolution I don't know what the equivalent is, but I've set the
"check for new mail" timeout to 5 minutes to see what happens.
poc