> These late adopters will be late adopters of a future version of James
too.
> Which is fine. They are free to be as cautious as they see fit.
Not necessarily, If they are late adopters because of the ££Cost of
upgrading commercial products, and associated support contracts, the same
constr
> On Wednesday 16 June 2004 21:21, Serge Knystautas wrote:
> > Steve Brewin wrote:
> > > Anyway, before we dwell too much on the future, a big thanks to everyone
> > > who helped in getting this release out, especially Noel who has spent so
> > > much time painstakingly putting it all together.
> >
On Wednesday 16 June 2004 21:21, Serge Knystautas wrote:
> Steve Brewin wrote:
> > Anyway, before we dwell too much on the future, a big thanks to everyone
> > who helped in getting this release out, especially Noel who has spent so
> > much time painstakingly putting it all together.
>
> +1!
Actu
Steve Brewin wrote:
Anyway, before we dwell too much on the future, a big thanks to everyone who
helped in getting this release out, especially Noel who has spent so much
time painstakingly putting it all together.
+1!
--
Serge Knystautas
Lokitech >> software . strategy . design >> http://www.lokit
> From: Danny Angus wrote:
> > Back in December 2002 the primary J2EE servers (BEA WebLogic and IBM
> > WebSphere) used by the larger institutions required JDK 1.3.x and I
> argued
> > we should too. Current releases of BEA WebLogic and IBM
> WebSphere require
> > JDK 1.4.x and I would now argue t
Danny Angus wrote:
I'm in favour of making this change, but only if our users don't raise any
objection. Which we will only find out if we ask them.
There's been one response from a user to this thread so far. I don't
think anyone would object if you moved this thread to the user list.
--
Serge Kn
> Back in December 2002 the primary J2EE servers (BEA WebLogic and IBM
> WebSphere) used by the larger institutions required JDK 1.3.x and I
argued
> we should too. Current releases of BEA WebLogic and IBM WebSphere require
> JDK 1.4.x and I would now argue that this should be our minimum.
Steve Brewin wrote:
Back in December 2002 the primary J2EE servers (BEA WebLogic and IBM
WebSphere) used by the larger institutions required JDK 1.3.x and I argued
we should too. Current releases of BEA WebLogic and IBM WebSphere require
JDK 1.4.x and I would now argue that this should be our mini
(Ought to be JDK 1.3 or 1.4 redux)
This was an issue thrashed over in December 2002. My main argument then
remains my main argument now. Larger institutions are very conservative in
moving JDKs. If we feel that they are or could be a significant part of the
user base we should not freeze them out
> > Our code currently presumes something of a pull-model, and would
> > need some changes. The best approach would be to support both
> > socket I/O and selector I/O.
> Therein lies the problem though... since they are different models,
> it's non-trivial to support both.
If we do the work to s
Noel J. Bergman wrote:
NIO: This is a technology we could have if we upgraded to JDK 1.4. I'm
not sure what you mean... we are not using it since we cannot use it
Architecturally, we would have to change. Our code currently presumes
something of a pull-model, and would need some changes. The bes
> NIO: This is a technology we could have if we upgraded to JDK 1.4. I'm
> not sure what you mean... we are not using it since we cannot use it
Architecturally, we would have to change. Our code currently presumes
something of a pull-model, and would need some changes. The best approach
would b
Noel J. Bergman wrote:
There's NIO, built in JNDI DNS library, TLS capabilities. What about
requiring 1.4 for james 3.0?
We aren't using NIO, yet, there is no support for TLS with NIO in JDK 1.4
(requires JDK 1.*5*), we have better DNS support from dnsjava, and we
already have TLS in the current c
> There's NIO, built in JNDI DNS library, TLS capabilities. What about
> requiring 1.4 for james 3.0?
We aren't using NIO, yet, there is no support for TLS with NIO in JDK 1.4
(requires JDK 1.*5*), we have better DNS support from dnsjava, and we
already have TLS in the current code.
Not seeing a
Hi.
I'm a James user and I think you don't need ask your user. Windows users
are threatened continously with new viruses because 90% of them do never
update their systems. If you think the best for James is updating to JVM
1.4, excellent :) Also, you must remember old JVM versions had security
Danny Angus wrote:
There's NIO, built in JNDI DNS library, TLS capabilities. What about
requiring 1.4 for james 3.0?
+1
What about asking our users to see if we have an issue or not first??
I'm not saying to avoid their input, but not treat it as the basis for
the decision.
Similarly, if we made
> I know Danny is stuck with JDK 1.3. :)
I'm not stuck with 1.3 actually, but I know that people are, and understand
why.
> There's NIO, built in JNDI DNS library, TLS capabilities. What about
> requiring 1.4 for james 3.0?
> +1
What about asking our users to see if we have an issue or no
On Tuesday 15 June 2004 15:19, Serge Knystautas wrote:
> Danny Angus wrote:
> > I think we'd need to poll our users explicitly first.
> > There are still good reasons for not doing so unless we really *need* to.
> >
> > -1 (It is not a good indication)
>
> I know Danny is stuck with JDK 1.3. :)
>
Danny Angus wrote:
I think we'd need to poll our users explicitly first.
There are still good reasons for not doing so unless we really *need* to.
-1 (It is not a good indication)
I know Danny is stuck with JDK 1.3. :)
There's NIO, built in JNDI DNS library, TLS capabilities. What about
requirin
I think we'd need to poll our users explicitly first.
There are still good reasons for not doing so unless we really *need* to.
-1 (It is not a good indication)
d.
> Could it be an indicator that we can move on to requiring 1.4, that would
be
> nice.
>
> --Søren
+1 :-)
Vincenzo
---
> On Monday 14 June 2004 23:05, Noel J. Bergman wrote:
> > As you can probably tell, I just tweaked CVS to include the final release
> > versions of DBCP and Pool. The only change between what we have now and
> > what we used for testing is trivial removal (a few lines, which were
> > carefully ve
On Monday 14 June 2004 23:05, Noel J. Bergman wrote:
> As you can probably tell, I just tweaked CVS to include the final release
> versions of DBCP and Pool. The only change between what we have now and
> what we used for testing is trivial removal (a few lines, which were
> carefully vetted by ha
As you can probably tell, I just tweaked CVS to include the final release
versions of DBCP and Pool. The only change between what we have now and
what we used for testing is trivial removal (a few lines, which were
carefully vetted by hand) of some JDK 1.4 dependencies. It is interesting
to note
23 matches
Mail list logo
