Some unscientific timing results (on the Python source tree)

[David Cournapeau]

Paul Moore wrote:
>
> Good point, I'd forgotten that. But I repeat that I wasn't trying to
> optimise, rather to "just use" what the people setting up the
> repositories had suggested I use. (I don't even know if there's a bzr
> smart server I have the ability - as a user without commit access - to
> use).
>

I understand the nuisance as a user, but I am not sure you can blame bzr 
if nobody set it up for good performances. Maybe bzr server is harder to 
set up than mercurial, at which point it is a bzr problem (I don't say 
it is, I just don't have that experience).

>
> I dispute "reasonable". 

No need to dispute, that was just my broken English :) I meant expected.

> "Unsurprising" or "expected", maybe. But
> "reasonable" implies that you feel that people should not be looking
> at Mercurial and Bazaar, and considering the speed difference as being
> relevant in making a choice. As I said, I am personally very close to
> a point where I would choose (and advocate) Mercurial over Bazaar for
> any significantly-sized project, based on the speed difference.

Well, that's a reasonable (hope I used it right this time:) ) position.

>  And if
> the Bazaar team don't feel a need to improve speed further with "as
> fast as other DVCSs" as a target, then that pushes me even closer to
> deciding on Mercurial.
>

I don't know about bzr team's directions, but some of the differences 
you pointed are known (log, and blame (?) I think are significantly 
slower than on mercurial) . Some other, I don't consider them as 
important (0.5 vs 1s for example: if you don't like latency, you can use 
bzr shell to avoid the overhead of bzr launch; that's what I am using 
myself).

Personally, if using bzr was as bad as it is right now for python 
repository (getting the repository as a big tarball, 10 minutes to 
update a few changes), I would have never considered it. In my own 
settings (local development, mainly, plus branches on launchpad with 
bzr+ssh:// access), it is good enough (I have a few private branches 
from projects of several thousand files/several thousand revisions, that 
is the same order as python I think). So maybe that's just a question of 
setting things up a bit better than they are now.

cheers,

David