[PATCH] performance history specification

[Robert Collins]

On Thu, 2006-07-13 at 23:21 -0500, John Arbash Meinel wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Robert Collins wrote:
> > On Wed, 2006-07-12 at 14:29 -0500, John Arbash Meinel wrote:
> >> -----BEGIN PGP SIGNED MESSAGE-----
> >> Hash: SHA1
> >>
> 
> ...
> 
> >> Seems like you are writing a timestamp from epoch time here, rather than
> >> a date. It would look nice if you wrote:
> >>
> >> bench_history.write(time.strftime('--date %Y-%m-%d %H:%M:%S\n'))
> >>
> >> Or if you want something more official, maybe:
> >> bench_history.write('--date %s\n' %
> >> 		    (datetime.datetime.utcnow().isoformat(),))
> >>
> >> But in general, if it is for human viewing time.time() is not very good.
> > 
> > The file is not for human viewing - its a log file for data gathering.
> > time.time() gives seconds since epoch, in UTC, and can be trivially
> > formatted into any date when the file is parsed.
> 
> Well, I probably wouldn't call it '--date' if you are writing a
> timestamp. But I understand its purpose.
> 
> ...
> 
> >> It looks pretty good. The only thing it isn't doing is checking to make
> >> sure .perf-history isn't being overwritten.
> >> Also, it wasn't clear where it was creating the .perf-history file. I
> >> assume you were shooting for putting it inside the project (not the test
> >> temp dirs)
> >> Which means that .bzrignore should be updated to ignore the newly
> >> generated file.
> > 
> > Good point, +1 on adding an ignore to .bzrignore for .perf-history.
> > 
> > Yes, it puts it in cwd, which is what the tests test. And you are right,
> > we don't test that it appends always. That should be a TODO for the next
> > set of performance work related to this.
> > 
> > That said John, are you happy for this to merge as is (with .bzrignore
> > changed)?
> > 
> > Cheers,
> > Rob
> 
> Sure. I'd prefer --timestamp, but the real changes I'm +1 on.
> 
> Do we want to get this in for 0.9, or should we wait until after? It
> seems pretty small, and reasonable to come in.


I'd like it in asap so that we can put a cron job on bazaar-vcs.org
running HEAD with this, so we have object performance stats for all the
benchmarks.

Rob
-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060714/728f1e04/attachment.pgp