[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Monotone-devel] Re: results of mercurial user survey
From: |
Nathaniel Smith |
Subject: |
Re: [Monotone-devel] Re: results of mercurial user survey |
Date: |
Fri, 28 Apr 2006 10:48:47 -0700 |
User-agent: |
Mutt/1.5.11 |
On Fri, Apr 28, 2006 at 03:50:05PM +0200, Markus Schiltknecht wrote:
> Why not simply keep a black-list of suspect revisions? Every node can
> recheck the suspected revision and decide itself, if it really is
> erroneous.
So, there's obviously a lot of interest in this thread in coming up
with a general solution to this problem, and a lot of people thinking
up different things to do, suggesting things to work on, etc.
Now, if you read my original post, you'll notice that I didn't say we
_shouldn't_ come up with some tricky not-as-safe-but-at-least-its-fast
sort of approach; I just argued that we shouldn't do it _yet_.
So, a modest proposal -- maybe some of the energy in this thread would
be good to direct towards speeding up the netsync we already have?
Remember, right now, _no-one knows_ whether we can make netsync fast
_and_ safe -- maybe it can be done! Wouldn't _everyone_ prefer that,
if we can manage it?
And, you know, if we try and fail, then of course we'll consider other
options. But right now it feels like we're rejecting the best
solution, because we're worried it _might_ not work.
Optimizing netsync is basically manpower-bound right now. It's
basically just me and Graydon and Matt Johnston whenever we manage to
put a little time in. There are lots of avenues to explore, that just
need someone with interest to work on them.
Some starting points:
http://venge.net/monotone/wiki/PerformanceWork
(the performance test harness on this page would itself be a huge
help, and requires no knowledge of monotone internals at all)
http://venge.net/monotone/wiki/DeltaStorageStrategies
http://venge.net/monotone/wiki/DeltaStorageStrategies/ShootOut
and, of course, profiling, profiling, profiling.
(Recommended: Oprofile, callgrind, and Patrick has been helping us
figure out how to use DTrace on OpenSolaris to profile IO patterns.
We can totally help anyone get started using these things on IRC or
whatever.)
-- Nathaniel
--
"Lull'd in the countless chambers of the brain,
Our thoughts are link'd by many a hidden chain:
Awake but one, and lo! what myriads rise!
Each stamps its image as the other flies"
-- Ann Ward Radcliffe, The Mysteries of Udolpho
- Re: [Monotone-devel] Re: results of mercurial user survey, (continued)
- Re: [Monotone-devel] Re: results of mercurial user survey, Richard Levitte - VMS Whacker, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Markus Schiltknecht, 2006/04/28
- [Monotone-devel] Re: results of mercurial user survey, Bruce Stephens, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Thomas Keller, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Christof Petig, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Christof Petig, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Jon Bright, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Nuno Lucas, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Markus Schiltknecht, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Ethan Blanton, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey,
Nathaniel Smith <=
- Re: [Monotone-devel] Re: results of mercurial user survey, Justin Patrin, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Zbynek Winkler, 2006/04/28
- Re: [Monotone-devel] Re: results of mercurial user survey, Johan Bolmsjö, 2006/04/27
Re: [Monotone-devel] results of mercurial user survey, Zbynek Winkler, 2006/04/27
Re: [Monotone-devel] results of mercurial user survey, Joe Wilson, 2006/04/28