On 2026/07/02 23:43, Daniel Dickman wrote: > > > On Thu, 2 Jul 2026, Stuart Henderson wrote: > > > On 2026/07/01 17:21, Daniel Dickman wrote: > > > > mailman2 and mailman3 need to be installed in parallel to manage the > > migration. when I read your earlier mail "porting mailman 2 to 3" I > > thought you were talking about porting it to python 3... (there is > > also https://github.com/jaredmauch/mailman2-python3 which while > > still labelled 'alpha' is very active). > > > > mailman2 and mailman3 are not really comparable anyway, mailman3 is > > a very different beast and most list admins seem to move to other > > software (mlmmj etc) rather than to mailman3. > > so let me know which route you'd like to go and I'll try to do the work > for it as time permits. I don't really have a preference as long as we > remove the Python 2 dep which is my only goal.
I honestly don't know. Maybe a better option would be to remove it from ports. Is anyone reading that actually uses mailman from OpenBSD ports/packages? Any comments? > > also it really seems > > intended to run in a venv. I'm a bit concerned about keeping all of > > those moving parts synchronized otherwise... > > Some of the below would be useful no matter which way we go. So would be > good to get oks to import. I've added comments below on those. agreed. I'll try to have a look through these over the next few days. > > > devel/py-zopehookable > > -> missing RDEP from devel/py-zopecomponent (which is already in > ports) > > > > devel/py-zopei18nmessageid > > > devel/py-zopeschema > > > devel/py-zopeconfiguration > > > devel/py-atpublic > > -> needed by aiosmtpd > > > > devel/py-flufl.bounce > > > devel/py-flufl.i18n > > > devel/py-flufl.lock > > > devel/py-lazr.delegates > > > devel/py-lazr.config > > > devel/py-publicsuffix2 > > -> needed by authheaders > > > > mail/py-authheaders > > -> useful for DKIM/ARC related mail tooling > > > > mail/py-aiosmtpd > > -> useful aio smtpd module > > > > mail/py-standard-nntplib news/ for this one, I think > > > www/py-falcon > > -> nice web framework, looks independently useful >
