Hi All, I'd have some concerns about the scale of impact on a MINA upgrade to the M2 branch. My own feeling is that this is too significant a change to introduce now, unless it's a bug fix we really need ?
Would certainly need quite a bit of testing, but perhaps more than that it also offers 'opportunites' to do some stuff with throttling etc. I'm not sure we ought best to do the upgrade and yet not make the associated enhancements .... We're encountered quite a bit of trouble with the MINA stuff in the past, owing mainly to the complex bits of code it affects afaik, so I'd suggest we hold off and do this upgrade on trunk for M3. Hth ! Bfn, Marnie On 5/22/07, Rupert Smith <[EMAIL PROTECTED]> wrote:
There are some tests for the retrotranslated client code, or rather the client code and anything in the integrationtests module are both retrotranslated (but not systemtests as it has a broker dependency, does they use an in VM broker?) so can be run together to check the build is still 1.4 compatible. Retrotranslator seems pretty good, but there are still some gaps, so I hope Mina 1.1 doesn't discover any of them. +1 Rupert On 22/05/07, Robert Godfrey <[EMAIL PROTECTED]> wrote: > > I'm +1 for moving to Mina 1.1 > > The client code already requires Java 5 to compile, and our 1.4 > strategy is to use the retrotranslator. > > -- Rob > > On 22/05/07, Martin Ritchie <[EMAIL PROTECTED]> wrote: > > Hi. > > > > Just wanted to get the lists thoughts on upgrading our dependencies > > such as mina for the M2 release. Currently the mina we use requires > > backport utils but we could upgrade to the fully java 5 version. Does > > anyone have any thoughts or reasons why we shouldn't do this now? It > > would be good to upgrade as a number of bugs have been fixed since the > > 1.0.0 we downgraded to. > > > > -- > > Martin Ritchie > > >
