On Wed, 27 Mar 2019, Jonas Meurer wrote: > Hi Alex, > > Alexander Wirt: > > On Tue, 26 Mar 2019, Jonas Meurer wrote: > >> Alexander Wirt: > >>> In my experience as a former mailman admin and listadmin mailman is a > >>> no-go. > >>> Getting our feature set even nearly into mailman is impossible, takes years > >>> and will just get us an unmaintainable thing. I don't want to ever run a > >>> bigger mailman setup again. > >> > >> Can you give an example, what from "our feature set" is missing in > >> mailman? Also, you probably mean mailman2, right? Have you taken a look > >> at mailman3 recently? > > > > Sure, just a few coming into mind: > > > > All those gpg related features we use, our spam removal tools, our special > > archiving hacks, we way we support blacklist through several lists, our > > second line of spamfiltering, crossassassin, the way we can do management on > > several lists and probably a lot more I forgot. It may take man years to do a > > migration (in fact we talked about that a few days ago in our internal IRC > > channel and this is more or less consense about the needed effort). > > Thanks for elaborating on that. At least some of them might be > interesting to submit as mailman3 feature requests as they probably > would be of help for other projects that use mailman3 as well. > > I fully understand that you as the listmasters don't consider to switch > right now though. > > > And I tested hyperkitty some time ago with our archive and it was > > unusable slow. > > Interesting. I wonder how Fedora deals with this. I haven't used their > archives extensively, but from quick tests it appeared to be quite > responsive. One question that came into mind: which problems with our lists do you think mailman would solve? If its list archives: provide (or let someone do so) a hyperkitty plugin for reading our mboxes (bonus points if its able to also remove mails from it afterwards (gpdr, spam). Alex
Attachment:
signature.asc
Description: PGP signature