Sasvata (Shash) Chatterjee wrote:

Leif,


If we are going to do another RCn then I have a minor change of my own for the excalibur-instrument-mgr-impl jar. If not then it can wait until the next release. (A typo and removing a stack dump if the state file is corrupted or can't be accessed for any reason. It can happen if a previous JVM invocation crashed
at just the wrong moment.)

You tell us. As the release-manager, here's my position: There's no sense doing a release, but leaving the community high and dry if they cannot use the release. So, tell us, if you can trudge along with using most of the new release, but use the previous fortress-meta.

I don't think I really have any choice but to keep using the previous fortress-meta for now. From what I could tell by looking at the code, there is not a way to fix the encoding problem from within Excalibur. It is a QDox problem. I don't think it is reasonable to hold the release until that gets fixed as that could potentially take quite a while. (I have not heard if anyone has yet contacted them about this issue. If nobody in Excalibur has any contacts over there, then I will go ahead and do it.)

If you cannot do that, we'll do a RC4 and fix Vadim's (I think important, but non-blocker) issue, and you can fix the minor issue you mention here. But, here's the catch. I would not do the RC4 for another week, instead really *encouraging* all users to complete testing with RC3. Once we do RC4, the only things that will stop the release will be things broken between RC3 and RC4, but nothing else.

Does that sound kosher?

Yes, my changes to instrument are quite minor. I would only put them in if we were
already going to do an RC4 anyway.

By the way. Thank you for all the energy you are putting into this release cycle.

Cheers,
Leif


Shash


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]




---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to