Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [egit-dev] Logical model support in EGit

On Thu, May 30, 2013 at 4:41 PM, Laurent Goubet <laurent.goubet@xxxxxxx> wrote:
Hi,

Okay, I've finally gotten back to this amidst the holidays and release/build/... tasks. I've tried to limit the size of all this by splitting the tasks :

- Factorizing the comparison code, making sure we use the same code for each of the 17 actions to launch comparisons. I have submitted a second patch set on https://git.eclipse.org/r/#/c/11966/ for review.
- Enabling restrained synchronizations : I am making these changes to support the logical model; this means that EGit will launch synchronizations not only on the whole repository as it does today, but also on smaller sets of files. I have rebased https://git.eclipse.org/r/#/c/11928/ (which makes sure EGit does not fetch "too much" from the repository when synchronizing) and submitted https://git.eclipse.org/r/#/c/13404/ that allows for finer-grained restrictions on the resources included in a synchronize operation.

This second action would also allow us to provide "synchronize" actions on folders or projects instead of the sole "synchronize workspace" we have now.

I have not modified the comparisons from the commit dialog since these open comparison dialogs instead of editors. I haven't made the changes to have less "magic" comparisons with the index (compare index with head, compare with index) for now either since those will need a much more intrusive change. I'll explore Robin's suggestion on that, but it will be another commit :p.

Hudson does not seem to have been triggered by these reviews. Has it been disabled? I could not launch the SWTBot tests on my machine, so was expecting its results :).

Hudson was triggered but currently fails all EGit builds since jgit's packaging build is
packing some wrong older jgit versions coming from the eclipse platform referenced
by the packaging build. https://git.eclipse.org/r/#/c/13465/ should fix this.

--
MatthiasĀ 

Back to the top