[2025-01-10 12:52] Omar Polo <o...@omarpolo.com> > On 06/12/24 12:44, Philipp wrote: > > [2024-11-27 11:12] Philipp <phil...@bureaucracy.de> > >> I would like to understand whats the reason for this quite low limit. > >> Also I'm currently unsure what a reasonable default limit would be. > >> An option to change this limit would also be nice. I check if I can > >> write a patch for this on the weekend. > > I have attached a patch for this. I haven't had the time to test it. > > > > I'm not opposed to something like this. My only concern is that you're > providing a knob that applies to all the user in a system, while I'd prefer > to err on the safe side and have a cap per user. This way, you only need to > bump the limit per the user(s) you're using to deliver the mail. Maybe we > could also provide a global knob though.
I can check how to implement a per user setting. Some sort of global fallback would also be nesesary. > (another option would be to use smtp(1) and connect to localhost, this will > avoid the limits, but if you're not mentioned it I assum > e you've discarded this option for some reason) > > > One nitpick is about the "control" keyword. maybe we could augment the > `listen on socket` with a `connection limit user $N` argument instead. I'll check how to add this to the listener. Have you an idea how the syntax should look like for a per user setting? Philipp
