Re: [PATCH 0/4] Honor core.deltaBaseCacheLimit during index-pack

[Nicolas Pitre <nico@cam.org>]

On Thu, 17 Jul 2008, Stephan Hennig wrote:

> Shawn O. Pearce schrieb:
> 
> > We're better off keeping our memory usage low and recomputing
> > the delta base when we need to return to it to process a sibling.
> 
> Thanks to all who have had a look at this issue!  From a user's
> perspective I have one more suggestions and a question:
> 
> First, it would have helped me to bring this issue onto the list if I
> had earlier known that this was no misconfiguration, but a memory
> problem.

Well, if we had known before that this could be a problem, we'd have 
fixed it earlier.  In other words, sh*t happens.

> Even though Git now makes some efforts to substitute runtime
> for memory to be able to operate with low(er) memory, I think it would
> still be informative for a user that repository and hardware, resp.
> core.deltaBaseCacheLimit, are, say, incompatible.  If valuable objects
> have to be discarded due to memory restrictions a warning could be
> issued to make the user aware of this fact, e.g.,
> 
>   Warning! Low memory. Git might be slowing down.

Well, I disagree.  First we don't know how slow git would effectively be 
since all (my) concerns so far were totally theoretical.  It will still 
work better than, say, 'git verify-pack' nevertheless. And git should 
just do its best regardless and avoid being needlessly verbose.

> Second, while there have been some changes to Git now, as a poor user,
> how can I make use of that changes?  I think, updating my client should
> only help with pushing.  For pulling, I have to wait for repo.or.cz to
> update to Git 1.6.0, right?

Actually that's the other way around.  This change will help the 
receiving side of a transfer.  So it should help you when pulling.


Nicolas