On Tue 30 Aug 2016 at 18:45:34 (+0100), Brian wrote: > On Tue 30 Aug 2016 at 11:18:10 -0500, David Wright wrote: > > > On Tue 30 Aug 2016 at 09:59:42 (-0400), Henning Follmann wrote: > > > > > > However, why email is still reliable, because a proper setup provides you > > > with a well defined error messages (in case it is not delivered). > > > > There are occasions when this is several days later, unfortunately. > > Some of the retry intervals seem to have been set in the days when > > people/institutions dialled up the internet on a daily schedule. > > I think you have moved from unreliability (whatever that means) to > timeliness of delivery. If you put your mail under the control of > a third party you presumably accept their conditions. Anyway, what, > without drowning the internet in frequent retries, would be suitable > as a sequence of retry intervals? Exim on Debian uses > > # This single retry rule applies to all domains and all errors. It specifies > # retries every 15 minutes for 2 hours, then increasing retry intervals, > # starting at 1 hour and increasing each time by a factor of 1.5, up to 16 > # hours, then retries every 6 hours until 4 days have passed since the first > # failed delivery.
Much of that is reasonable, but I think (without any evidence) that most people would prefer to receive a message of abandonment after 16 hours (perhaps only six) rather that sit on their hands for four days. For comparison, look at the time people here wait for a posting to appear before submitting a second try, and a third... Cheers, David.
