Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [tools-pmc] Requesting exception for patch feature shipped by Object Teams

Collection of answers below:

On Tuesday, November 16, 2010 05:46:26 pm Jeff McAffer wrote:
> My 2c.  We must be sure that when someone goes to the train repo and says (directly or indirectly) "I want JDT", they get the unpatched JDT.  If they choose (directly or indirectly) to install OT and OT needs a patch, great, but they should not get it "for free".

Thanks, Jeff. I 100% agree.
> If there are bugs in p2/b3/... that prevent the above from being true then they need to be fixed 

I'd be happy to help investigate..

> or the train repos must contain patches from the JDT team (in this case) that they intend all users to have.

JDT /Core team currently has no patches, so I'm not sure if this applies in any way, currently.


On Tuesday, November 16, 2010 06:33:18 pm Ian Bull wrote:
> I'm also curious as to what would happen if two people wanted to patch the
> same project?

not possible with current technology (and p2 will not offer such a solution)

> Or, if I installed the OT (and thus the patched JDT came
> along), and then JDT released a critical fix of their own? Would that patch
> be applied?

If only a JDT patch is provided: no. But I gave promise to my team-mates
at JDT/Core that I will provide a "same-minute" variant of such a patch that
would install on top of the patch feature. (Chances are that I authored that
JDT patch in the first place :)


On Tuesday, November 16, 2010 06:43:11 pm Chris Aniszczyk wrote:
> Here are my two cents:
> I'm fine with OT forking JDT as they have a use case for it.


> We should
> grant them an exemption for this time around with the EXPECTATION that
> they would work with the JDT to fix any issues by the next release
> (Stephan has even become a JDT committer which is great). This can
> serve as motivation for projects to fix issues as they wouldn't be
> allowed on the train next time around.

We'll for sure discuss this.

If "fixing" means that no patch feature will be allowed on the next train,
then this will perhaps be a short visit. But if we could get the exception
for now, we would gain time for convincing you (and the community as
such) that we're not causing any harm but actually deliver higher
quality code than some other projects on the train.
And I'm certainly most open to collaborate on any process issues 
should they arise.
> In terms of p2, to ensure that no one installs the JDT "fork" by
> accident... we would explicitly hide the feature when we organize the
> content (this is relatively easy to do).

I don't understand how this would help, frankly.


Back to the top