Re: [update] piggybank performance profiling

From: <eric_at_prebys.com>
Date: Fri, 9 Sep 2005 17:51:32 -0400 (EDT)

I am excited to try out the new build with the memory leak
fix, but I am not interested in running firefox's beta.

does that mean i have to wait for mozilla to make a new
production release?

for some reason i got the sense that piggybank is having a
problem with the firefox auto-update. this is probably a
misread by me, but just in case... in general, i think that
"versions" are a 20th century concept. at least, they have
very little place in a web2.0 world. i'm fine with having
two branches, but after choosing wether i want the
production or the developer branch, i want my web2.0
software to take care of staying on the freshest available
prod/dev version. every day, i should have new features to
greet me
(http://radar.oreilly.com/archives/2005/09/why_microsoft_cant_best_google.html)

On Fri, 9 Sep 2005, Eric Miller wrote:

> Date: Fri, 9 Sep 2005 12:36:38 -0400
> From: Eric Miller <em_at_w3.org>
> Reply-To: general_at_simile.mit.edu
> To: general_at_simile.mit.edu
> Subject: Re: [update] piggybank performance profiling
>
>
> On Sep 8, 2005, at 9:48 AM, Stefano Mazzocchi wrote:
>
> > David R. Karger wrote:
> >
> >> stefano, if you don't want to wait for a fix to the db, you might
> >> solve the problem by storing each predicate in both directions and
> >> having your interface swap the query to the fast direction if it is
> >> backwards.
> >>
> >
> > Yes, I thought of that. But I think if both Vineet and us need hit
> > this wall, maybe it's time to not work around it and fix it.
> >
> > BTW, with the latest fixes to the cache, PB performance is now
> > acceptable after a little caching happens (say a few pages get
> > loaded and the JVM hotspot starts optimizing for real).
> >
> > So, if we get the OP-index in place before the release great, if
> > not, well, we'll do it in the next one.
>
> I think the latest build, coupled with the firefox 1.5b1 is a
> substantive improvement in stability, functionality and performance.
> Well done!
>
> My take on this is Vineet's OP-index work, and fresnel display be
> incorporated into the subsequent release.
>
> --
> eric miller http://www.w3.org/people/em/
> semantic web activity lead http://www.w3.org/2001/sw/
> w3c world wide web consortium http://www.w3.org/
>
>

-- 
Received on Fri Sep 09 2005 - 21:45:13 EDT

This archive was generated by hypermail 2.3.0 : Thu Aug 09 2012 - 16:39:18 EDT