Re: [egit-dev] Expected performance profile of staging changes and commi

[Robert Munteanu <robert.munteanu@xxxxxxxxx>]

Hi Markus,

On Wed, Nov 7, 2012 at 9:57 AM, Markus Duft <markus.duft@xxxxxxxxxx> wrote:
> On 11/06/2012 06:00 PM, Matthias Sohn wrote:
>> 2012/11/6 Markus Duft <markus.duft@xxxxxxxxxx <mailto:markus.duft@xxxxxxxxxx>>
>>
> [snip]
>>
>>
>> we need to keep the ability to do a full refresh also for the case when some git operations are done by
>> another process than the Eclipse instance we are looking at. E.g. you may want to do some advanced
>> git operations not yet available in EGit using native git which won't send these events.
>
> I understand the problem you're talking about, but i don't see how this would interfere with this code? you mean if the index does not change, but only files on disc change externally? This would only be detected by a full refresh then, right? is there another way to detect it?

Manual refresh ( from the staging view ) still works, I still drop to
the cli for some tasks.

>
> I now did a second (little different) implementation, which i like better, since it does not actually change JGit, but only the IndexDiffCacheEntry to fix the local performance issue. Please tell me which one you like better, and i will try to get production quality on that version then :D
>
> Either these two (from yesterday):
>         [1] https://git.eclipse.org/r/#/c/8535/
>         [2] https://git.eclipse.org/r/#/c/8536/
> Or this single on (from today):
>         [1] https://git.eclipse.org/r/#/c/8558/
>
> For the second one i have a slightly better feeling about things like synchronization, and stability/reliability.

I'm now running with jgit master and egit master + gerrit #8558 and
staging/unstaging are a lot faster, so thanks for that! I'll let you
know if I run into some unintended side effects but this change should
help a lot.

Robert

>
> Markus
>
>>
>> --
>> Matthias



--
Sent from my (old) computer