You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@batchee.apache.org by Scott Kurz <sk...@gmail.com> on 2014/10/30 22:16:40 UTC
Interesting in contribution of Chunk controller refactoring for RI
Bugs 6457 and others?
Hi all,
I just contributed a fix to our recently-opened source RI & TCK (moved
to GitHub) which you might be interested in.
The biggest thing it fixes is the problem with retry-with-rollback
where, after going to one-at-a-time processing, we
never revert back to "normal" chunk processing.
See:
https://java.net/bugzilla/show_bug.cgi?id=6457
Along the way I addressed a few other bugs on the RI (6456, 5873, at
least part of 5875).
There were some duplicate listener calls and other cleanup as well.
I'd be happy to merge the changes as a whole and send a patch. I
don't think I have time to break apart every little
change though and map each back to an external bug with separate
patches. I'd answer questions though.
Sound interested enough that it's worth me doing the patch?
The change is at:
https://github.com/WASdev/standards.jsr352.jbatch/commit/1679181c55a5267a15feade1bba35fffd4d401c4
It's a pretty decent-sized refactoring of ChunkStepControllerImpl... I
moved the state to the ChunkStepControllerImpl instance itself
instead of trying to pass it around.
Scott
Re: Interesting in contribution of Chunk controller refactoring for
RI Bugs 6457 and others?
Posted by Romain Manni-Bucau <rm...@gmail.com>.
Hi Scott,
+1 globally, we'll surely clean up the patch then (we removed
logger.entering/exiting almost everywhere) but we can handle it on our
side for sure.
Romain Manni-Bucau
@rmannibucau
http://www.tomitribe.com
http://rmannibucau.wordpress.com
https://github.com/rmannibucau
2014-10-30 22:16 GMT+01:00 Scott Kurz <sk...@gmail.com>:
> Hi all,
>
> I just contributed a fix to our recently-opened source RI & TCK (moved
> to GitHub) which you might be interested in.
>
> The biggest thing it fixes is the problem with retry-with-rollback
> where, after going to one-at-a-time processing, we
> never revert back to "normal" chunk processing.
> See:
> https://java.net/bugzilla/show_bug.cgi?id=6457
>
> Along the way I addressed a few other bugs on the RI (6456, 5873, at
> least part of 5875).
> There were some duplicate listener calls and other cleanup as well.
>
> I'd be happy to merge the changes as a whole and send a patch. I
> don't think I have time to break apart every little
> change though and map each back to an external bug with separate
> patches. I'd answer questions though.
>
> Sound interested enough that it's worth me doing the patch?
>
> The change is at:
> https://github.com/WASdev/standards.jsr352.jbatch/commit/1679181c55a5267a15feade1bba35fffd4d401c4
>
> It's a pretty decent-sized refactoring of ChunkStepControllerImpl... I
> moved the state to the ChunkStepControllerImpl instance itself
> instead of trying to pass it around.
>
> Scott