I didn't catch the java api note... are you considering Apache James integration? James is a very interesting project, but w/o IMAP support.. seams like dbmail is a good fit to help out on a james implementation 'til they wrap up the imap support. It's been years so it might never happen for them.
I'd be very interested in your work regardless. > On Mon, Aug 1, 2005, Brandon Mercer <[EMAIL PROTECTED]> > said: > >> Kevin Baker wrote: >> >>>It is my understanding that the biggest issue with IMAP >>>web interfaces is connection handling. Basically there >>> is >>>no connection pool for IMAP on most, at least PHP, web >>>clients. >>> >>>This can be handled easily by implementing something >>> like >>>Perdition IMAP Proxy that has its own connection >>> handling, >>>in addition to other great scalability features, that >>> will >>>help significantly with this IMAP *Client* issue. >>> >>>As pointed out by Mike, I'm pretty sure it is not a >>>protocol or server issue... it is really a client >>>connection handling issue. >>> >>> >> I have just done some "testing" and I can much better >> attack the >> information in the database through my own connections >> rather than >> IMAP. Case closed. :-) I'll be writing my own API in >> both java and C >> to your program, it will be release under a BSD style >> license. It will >> account for changes in the database tables should there >> be any. If >> anyone is interested please let me know. >> Brandon > > You're welcome to go this route, and indeed it has already > been travelled > by webDBmail, and it works fine. The deal is that the 2.1 > series is > development. It would be a burden on the development > effort to have a live > library out there that people want us to remain tied to > and not make any > database changes. > > But I figure it'll take you a while to write your library, > and at some > point we'll have frozen the database and have stable > releases of DBMail on > that. So you'll probably be safe. > > But we're also going to be writing a libDBMail to abstract > the interface > to the database so that *if we do need to change a stable > database* we can > safely do it without third parties screaming at us. > > So, from a certain point of view, you're basically writing > libDBMail-Java. > > Aaron > > _______________________________________________ > Dbmail-dev mailing list > Dbmail-dev@dbmail.org > http://twister.fastxs.net/mailman/listinfo/dbmail-dev >
