I was just looking at this problem today. magit-status is getting slower
and slower each day for one of my repos. Sign me up to be a tester.


On Thu, May 2, 2013 at 3:42 PM, <[email protected]> wrote:

> Hello all,
>
> I use magit to interact with 
> github.com/mozilla/mozilla-**central<http://github.com/mozilla/mozilla-central>,
> a fairly large repository, and it is a poor experience.  My beefy laptop
> with an SSD regularly janks for 10-30 seconds on synchronous git
> invocations.  I would like to improve the magit experience.
>
> I have several patches under development to make magit performant, and I
> would love testers and reviewers.  For example, [1] speeds up
> branch-guessing significantly for me.  I also have patches in the works to
>
> * add a mode-line indicator during refresh;
> * cache some frequent git invocations;
> * make displaying tree-graph output in logs an option (this is very slow
> on large repositories);
> * log git invocations, outputs, and record running times (so that we can
> profile magit more effectively);
> * make sections not be refreshed automatically.
>
> These almost make magit useable with mozilla-central.  If you're
> interested in any of these approaches, or have other requests or ideas for
> making magit speedier, please let me know [2].
>
> Yours,
> Nick Alexander
>
> [1] 
> https://github.com/magit/**magit/pull/636<https://github.com/magit/magit/pull/636>
>
> [2] A few things I think are not feasible:
>
> It appears to me that magit essentially does not cache.  This is most
> evident in `magit-refresh-buffer', which tears down all sections and
> synchronously invokes git to repopulate everything.  (There is a small sop
> to maintaining `needs-refresh' state for diffs.)  I was surprised, at
> first, but this seems like an excellent decision: git can cache better than
> magit can, and this makes running magit in multiple repositories
> transparent.
>
> magit does not do much refresh work asynchronously.  Changing refresh to
> invoke git asynchronously would be a significant reworking of the section
> internals, to my eye, so I would not propose such changes.
>
> --
>
> ---
> You received this message because you are subscribed to the Google Groups
> "magit" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"magit" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to