May I ask, how many people work on DBmail? I cannot help directly, but
if it would assist I'd be happy to donate to the project, given how much
easier DBMail will be making my life. I can't afford to sponsor the
project, but I would be happy to help in any way I can.
Thanks,
- Naz.
It's on the target list for 2.4 alright. As to a timeframe however - I
wish I could be specific there. I've got my hands full at the moment
rewriting the server core and sql layers: event-driven, threaded
database connection pools, prepared statements, etc. Most of the work
has been done already. What remains is making more imap commands
non-blocking (by delegating them to worker threads), but a nice pattern
is already emerging that can be expanded to the rest of the codebase.
But since the native sql layer has been replaced by libzdb, any bugs in
libzdb that are exposed by dbmail will have to be fixed by the libzdb
people. They have been very good at that, so I'm not worried. But it
also means I have to wait with releasing 2.3.3 (the first release with
libevent, libzdb and threading) until all supported backends pass the
tests. At this moment only the postgres driver is bug-free, but the bugs
remaining in the mysql and sqlite drivers are very minor.
Once the server-core stabilizes we will have a much better architecture
for dealing with scalability issues. That will also free my hands to
deal with the other main todos for 2.4:
- native SSL/TLS support (John Guthrie has done the groundwork there
already)
- IMAP THREAD=REFERENCES
- selected IMAP extensions (CONDSTORE, LITERAL+, MULTIAPPEND, UIDPLUS)
- archive mode (this is what you want)
Since I'm doing this on my own, don't expect a final 2.4 release before
october, unless a sponsorship deal comes along...
_______________________________________________
DBmail mailing list
[email protected]
https://mailman.fastxs.nl/mailman/listinfo/dbmail
