I think that some work can be done investigating where the jelly lies.
We can then add sub-tasks to track their conversion. All of this can be
done w/out interfering with the M5 release.
Regards,
Alan
Geir Magnusson Jr. wrote, On 8/17/2005 9:54 AM:
Thx.
I think we should consider choosing appropriate timing - it sounds
like a big effort, and how does this fit into what we want to do w/
M5, 1.0, etc I was going to suggest voting on even doing this, but I
can't imagine that at *some* point in the future, we don't want to be
on M2, so we will go there eventually. it's just a matter of when...
geir
On Aug 17, 2005, at 12:42 PM, Alan D. Cabrera wrote:
To be a little more specific, maven 1 and maven 2 can both work with
maven 2 POMs. It's our non-standard jelly code that impedes our
conversion. We need to enumerate where that code is, determine if
there is an M2 replacement for it, it not decide if we need it and
if we do, make a plug in or convert the code.
It is my understanding that much of the jelly code is optimizations,
e.g. don't run tests if they've already been run. The maven team is
happy to make such optimizations a priority to include in M2.
After the conversion, I hope that we have no more jelly.
Regards,
Alan
Alan D. Cabrera wrote, On 8/17/2005 9:34 AM:
Normally, yes. Unfortunately there is a lot of jelly code out there.
Geir Magnusson Jr. wrote, On 8/17/2005 9:15 AM:
Is it backwards compatible?
IOW, can people that use M1 also still work?
