Re: Reverting the staging branch.

David Bélanger <[email protected]> Sun, 16 Jan 2005 15:46:37 -0500
Newsgroups gmane.comp.java.vm.sablevm.devel
Message-ID <[email protected]>
On Sun, Jan 16, 2005 at 12:00:04PM -0500, Etienne Gagnon wrote:
> 
> Note to David:  I know that SableJIT was also badly broken by this new 
> code.  I recommend that you back off the merge, and work with Greg to 
> first solve the normal staging merge with latest classpath.
> 

Well, it was not that badly broken :)  I now have a fix for it and
everything is fine.

I don't know about the other broken apps but on my side everything is ok
and I would have actually prefered to avoid the revert or at least have
extra time.

> I know that breakage can happen; so do not consider this as a critic of 
> anybody's work.  It's the objective of staging to catch these things 
> early, before they hit end users.  Yet, contrary to CVS trunks, staging 
> is not supposed to be remain in a "break everybody's build" state once a 
> problem is identified.  It is supposed to be either:
> a) fixed immediately, if possible, or
> b) reverted to last stable state, and work on a fix moved to private 
> sandbox.
> 

Agree 100% but again I would have appreciate more time to fix... like a
few extra days, not much.

In any case, please provide svn URL of the branch I should switch to for
the latest SableVM-CP.


David

---

David Bélanger
Web page:   http://www.cs.mcgill.ca/~dbelan2/
Public key: http://www.cs.mcgill.ca/~dbelan2/public_key.txt