Adrian Chadd wrote:
> On Sat, Aug 05, 2000, dancer@zeor.simegen.com wrote:
>
> [snip adrian's commloops stuff]
>
> > > This is based upon ideas from Andreas and Dancer, and initial coding from
> > > Benno. If you want the details, squid-dev mailing list archives are your
> > > friend.
> > >
> >
> > And in fact I've been doing some experiments of my own (in the few nanoseconds
> > that my job allows me at the moment...grrr). Just to let you know that I'm still
> > tinkering down that path.
>
> Got a sourceforge account? commloops exists to play with optimising/modifying
> the network communication code, it would be a shame to duplicate work.
I do. (Lucked out with dancer, which was not yet taken)
But like I said, work has me screwed for time. Like totally. Every second I spend on
things like Open Source....(or eating) takes time from my sleep cycles. I actually
think I've damaged myself :/ (which would explain the heart-failures that started 12
months ago).
At the moment, I'm working with a (non-caching) proxy framework of my own. It's been
built from the ground up in C++, and I can manage all sorts of funk in it in very
little time. Which is kinda good, really, because very little time is exactly what I
have. Results are mixed, though. I haven't been able to generate any good numbers...I
haven't had time to generate any _useful_ numbers, moreover. Although the first
couple gimmicks (which were the simplest things I can think of) fell flat. Primarily
because they were the simplest things I could think of, and I hadn't thought them
through completely.
Lesson learned: Outside of particle physics, if a system appears like it would behave
in a counter-intuitive fashion, then it won't. It'll behave opposite to how you
expect, because you missed an important concept.
D
Received on Sat Aug 05 2000 - 09:51:39 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:12:34 MST