[Pharo-dev] Komitter

Stéphane Ducasse stephane.ducasse at inria.fr
Wed Dec 4 14:15:34 EST 2013


Nicolas

From what I understand Kommitter just filters the entities that will be passed as modified and as such does not change the 
ancestry of a package. 

Stef


On Dec 4, 2013, at 8:12 PM, Nicolas Cellier <nicolas.cellier.aka.nice at gmail.com> wrote:

> Does Kommitter enables partial commits (cherry picking) for a package?
> If so, my question is how does it deal with MC ancestry?
> 
> Imagine I have a package- A.247 from my-trunk.
> I merge some A.312 from my-work repository with work in progress.
> Now I want to cherry pick the changes which are OK, and publish only those to my-trunk.
> I do so in my-trunk/A.313.
> Will A.312 be in the ancestry of A.313?
> Logically as MC is working, yes, because it is in the ancestry of in-image working copy.
> 
> What if I later download A.313 from my-trunk in a fresh image and try to merge A.312 to finish the work in progress?
> IMO, this is not possible, A312 being an ancestor, no change will happen.
> 
> So, it sounds like when I committed my-trunk/A.313 with partial changes, I should have commited my-work/A.314 immediately after with the rest of pending changes...
> If I forgot to do so and threw my image away, from a fresh image, it is still possible to load A.312, merge A.313, and publish A.314 in my-work repo.
> 
> But then it rather complicates the flow... If we commit/merge carelessly, we might throw away some valuable work in progress.
> The fear of such scenario prevented us to integrate a change from Bert in Squeak/trunk enabling such partial commits (directly from MC UI in Squeak case).
> 
> 
> 2013/12/4 Stéphane Ducasse <stephane.ducasse at inria.fr>
> +1
> 
> > Thanks once more for your effort. This is a really important tool for me.
> >
> > Uko
> >
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.pharo.org/pipermail/pharo-dev_lists.pharo.org/attachments/20131204/b9c1cd15/attachment-0002.html>


More information about the Pharo-dev mailing list