Re: [Evolution] attachments issue

[Patrick O'Callaghan <poc usb ve>]

On Thu, 2013-05-16 at 22:48 +0200, Thomas Prost wrote:
> Am Donnerstag, den 16.05.2013, 13:13 -0430 schrieb Patrick O'Callaghan: 
> > On Thu, 2013-05-16 at 09:52 -0400, Adam Tauno Williams wrote:
> > > On Thu, 2013-05-16 at 10:30 +0100, Pete Biggs wrote:
> > > > >         
> > > > > >  the .pdf source is indeed  there (and the mail file is huge)  
> > > > > Define huge.  1Mb? 5Mb? 100Mb? 10Gb?
> > > > > around 15Mb 
> > > > Personally I wouldn't trust any mail system with that size of
> > > > attachment.
> > >
> > > Personally I, and 600+ other users here, deal with attachments
> > > multiple
> > > times that size every day all day with no issues what-so-ever.  
> > >
> > > This is not 1991.
> > >
> > > According to my notes I increased the maximum message size at my site
> > > from 20Mb to 75Mb in 2003... with no measurable increase in issues.
> >
> > I seem to remember a recent thread on just this topic. To recap what I
> > said then: your mail system may be happy with large attachments, but you
> > can't assume that every relay in the path to any random destination is
> > equally happy.
>
> Why should they care ? Your mail doesn't come there at a stretch anyway.
> Or does packet switching need a lull from time to time ;-)

Think again. "Relay" does not mean "packet switch". It means an
intermediate host which stores and forwards complete email messages.
Take a look at the headers of some message to see the relays it passes
through before getting to you. For example at a quick glance your
message passed through:

pD9548C44.dip0.t-ipconnect.de (probably your own machine)
mrelayeu.kundenserver.de (your local mail server?)
moutng.kundenserver.de (some intermediate relay)
restaurant.gnome.org (presumably where the evolution-list is stored)
and thence via IMAP to my machine.

Each of these is administered independently and has its own idea of how
large a message it will accept for storing and forwarding. The largest
message that will get through is determined by the *least* upper limit
on any of these machines.

poc