This content has been marked as final.
Show 20 replies
-
15. Re: JBDEPLOY-3 issues
adrian.brock Mar 14, 2008 8:13 AM (in response to alesj)"alesj" wrote:
A fair warning: all AS5 deployers that define type should get that removed from xml, since setType is not longer there ;-)
Fair warning is a message to the dev-list with a title that has block capitals
and an explanation of what people need to do to get the correct behaviour.
In this case, the change to managed object should have gone into jbossas first
before we removed this from the deployers. -
16. Re: JBDEPLOY-3 issues
alesj Mar 14, 2008 8:16 AM (in response to alesj)"adrian@jboss.org" wrote:
That is unless you want to go through all the deployers that exist
to get jboss-head to compile/boot. :-)
Hehe, thought this might come up.
That's why I didn't want to remove type on the deployers in the first place. :-)
OK, will do that.
btw: What's the mechanism?
I'll just put 'Managed* api' for now. -
17. Re: JBDEPLOY-3 issues
adrian.brock Mar 14, 2008 8:30 AM (in response to alesj)"alesj" wrote:
btw: What's the mechanism?
I'll just put 'Managed* api' for now.
For now, just leave the type mechanism in the code, but deprecate it
and get scott to update the warning once he has done the work in the profile service
to make it work. We can then remove the code but NOT the method signature
until we know everybody has made the switch.log.warn("Types should be set using <Scott to provide mechansim/> for " + this);
:-) -
18. Re: JBDEPLOY-3 issues
adrian.brock Mar 14, 2008 9:20 AM (in response to alesj)By the way, this is holding me up.
I built a snapshot last night that included
some of these changes and now I can't get JBossAS to build
So if you can revert it asap that would be nice. ;-) -
19. Re: JBDEPLOY-3 issues
alesj Mar 14, 2008 9:29 AM (in response to alesj)"adrian@jboss.org" wrote:
So if you can revert it asap that would be nice. ;-)
Done --> Reverting back to 70804, one commit before type removal. -
20. Re: JBDEPLOY-3 issues
adrian.brock Mar 14, 2008 11:31 AM (in response to alesj)I don't know what you reverted but it was very incomplete,
don't worry about I'll fix it.
P.S. The stuff that Scott actually uses looks very broken to me anyway.
There is an AbstractDeployment::types field which isn't included in the serialization
contract????? :-)