RE: [PROPOSAL] UserDataRepository

2007-01-10 Thread Steve Brewin
Stefano Bagnara wrote: > > > Joachim Draeger wrote: > > Am Dienstag, den 09.01.2007, 20:52 + schrieb Steve Brewin: > >> I don't know about Robert's experiences. I have found that > you can get very > >> good performance with a custom persitence manager tuned to > the needs of the > >> applicati

Re: [PROPOSAL] UserDataRepository

2007-01-10 Thread Norman Maurer
Bernd Fondermann schrieb: > On 1/9/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote: >> PS: I like JCR and I think that JCR+JMS could be the future of a >> modular, transactional, clusterable, distributable James Server. > > Just a quick +1, I am thinking in the same direction for some time now. > >

Re: [PROPOSAL] UserDataRepository

2007-01-10 Thread Stefano Bagnara
Joachim Draeger wrote: Am Dienstag, den 09.01.2007, 20:52 + schrieb Steve Brewin: I don't know about Robert's experiences. I have found that you can get very good performance with a custom persitence manager tuned to the needs of the application and an appropriatley indexed and tuned databas

RE: [PROPOSAL] UserDataRepository

2007-01-10 Thread Joachim Draeger
Am Dienstag, den 09.01.2007, 20:52 + schrieb Steve Brewin: > > robert burrell donkin wrote: > > > (i performed experiments with storing emails in JCRs last year and > > > they seemed *very* quick but don't have any comparative numbers.) > > > > If you don't tell us what JCR implementation you

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread Danny Angus
On 1/9/07, Steve Brewin <[EMAIL PROTECTED]> wrote: Me too! Long term I would hope we can devolve the responsibility for achieving many of these characteristics to standards based components such as a JSR-170 based CR, leaving James to concentrate on mail technologies. Just wanted to add my +1

RE: [PROPOSAL] UserDataRepository

2007-01-09 Thread Steve Brewin
Stefano Bagnara wrote: > > > robert burrell donkin wrote: > > (i performed experiments with storing emails in JCRs last year and > > they seemed *very* quick but don't have any comparative numbers.) > > If you don't tell us what JCR implementation you tested and what > configuration this is meaning

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread Stefano Bagnara
robert burrell donkin wrote: On 1/9/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote: robert burrell donkin wrote: > (i performed experiments with storing emails in JCRs last year and > they seemed *very* quick but don't have any comparative numbers.) If you don't tell us what JCR implementation y

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread robert burrell donkin
On 1/9/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote: robert burrell donkin wrote: > (i performed experiments with storing emails in JCRs last year and > they seemed *very* quick but don't have any comparative numbers.) If you don't tell us what JCR implementation you tested and what configurati

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread Bernd Fondermann
On 1/9/07, Stefano Bagnara <[EMAIL PROTECTED]> wrote: PS: I like JCR and I think that JCR+JMS could be the future of a modular, transactional, clusterable, distributable James Server. Just a quick +1, I am thinking in the same direction for some time now. Bernd --

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread Stefano Bagnara
robert burrell donkin wrote: (i performed experiments with storing emails in JCRs last year and they seemed *very* quick but don't have any comparative numbers.) If you don't tell us what JCR implementation you tested and what configuration this is meaningless ;-) If you used JackRabbit, wha

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread robert burrell donkin
On 1/9/07, Joachim Draeger <[EMAIL PROTECTED]> wrote: Steve Brewin schrieb: > Joachim Draeger wrote: > As a persistent store for user mail its would be rather cool too as the > built in security means it could be used to safely expose the mail to the > alternative services that robert and ot

Re: [PROPOSAL] UserDataRepository

2007-01-09 Thread Joachim Draeger
Steve Brewin schrieb: Joachim Draeger wrote: I had especially text files in mind. XML, properties and scripts could be edited outside of james by different programming languages, tools, web-front-ends or simple text editors. Even in a BIGVARCHAR inside a DB. Using serialized objects would bi

RE: [PROPOSAL] UserDataRepository

2007-01-08 Thread Steve Brewin
Joachim Draeger wrote: > Am Montag, den 08.01.2007, 17:31 +0100 schrieb Stefano Bagnara: > > Joachim Draeger wrote: > > > > I have the idea of a repository that simply deals with > streams. That way > > > everything could be easily stored. Text (like a script), > XML, serialized > > > objects, pro

Re: [PROPOSAL] UserDataRepository

2007-01-08 Thread Joachim Draeger
Am Montag, den 08.01.2007, 17:31 +0100 schrieb Stefano Bagnara: > Joachim Draeger wrote: > > I have the idea of a repository that simply deals with streams. That way > > everything could be easily stored. Text (like a script), XML, serialized > > objects, property files. > > > I think that crea

Re: [PROPOSAL] UserDataRepository

2007-01-08 Thread Stefano Bagnara
Joachim Draeger wrote: Yet another proposal, go James go! ;-) Discussing jSieve usage, the question came up where to put the scripts. There are number of possibilities... File, DB, LDAP etc, etc. At the moment every Mailet that should work on a per user basis has to implement it's own backend.