From: Kragen Sitaker (kragen@pobox.com)
Date: Thu Mar 16 2000 - 15:57:13 PST
Jeff Bone writes:
> > It does sort of defeat the point; aside from the dubious benefit of
> > surreptitiously tunneling through firewalls,
>
> >From experience, there's no dubiousness in that benefit. It can literally mean life
> or death for apps intended to be run by corporate users. Activerse had the
> (probably, then) misguided notion that corps were ready for productive use of
> instant messaging; whatever philosophical problems we had, firewalls *were* always
> an adoption barrier in that market. This isn't an abstract / philosophical problem,
> it's a real-world problem that *all* communicating network applications face.
I didn't mean to imply that not being blocked by firewalls was
unimportant for the application. I meant only that if you run
everything over HTTP, your firewall becomes less useful --- thus my
doubt about the overall benefit.
Mark Baker writes:
> At 09:14 PM 3/15/00 -0600, Jeff Bone wrote:
> >Sure, but again, in loosely-couple wide-area apps: how important is
> performance?
>
> Hold on just a sec. How on earth is a distributed system built with RPCs
> "loosely-coupled"?!
RPC is just syntactic sugar for sending a request and waiting for a
response. It doesn't necessarily have anything to do with the degree
of coupling between the systems at each end.
Treating RPC like local procedure calls is definitely dangerous.
Mark, what do you prefer over RPC?
> But is it worth the cost of breaking the Web's architecture? What exactly
> will it mean to cache a method invocation?
It means memoization. :)
IMHO, caching http://www.google.com/search?q=mp3 is perfectly
reasonable. I would be surprised if Mark disagreed.
> What will it mean to try to transcode one?
The same thing it means for documents --- the result gets translated
into a format you can understand. (Or, if things are broken, into a
format you can't.)
> >WebDAV talks about *application level* abstractions.
>
> It talks about resource abstractions. It's hard to call those "application
> level" when HTTP is also concerned with them.
"resource" is a really fuzzy term. I think it means "file", although
the filesystem may not be in an OS kernel. Is that right?
Eugene Leitl writes:
> Anything which takes longer than 100 ns to communicate through a pipe
> thinner than 10 GBps is loosely coupled.
I guess my CPU is loosely coupled to my RAM, then :)
(Well, it would be if I had RDRAM, anyway.)
-- <kragen@pobox.com> Kragen Sitaker <http://www.pobox.com/~kragen/> The Internet stock bubble didn't burst on 1999-11-08. Hurrah! <URL:http://www.pobox.com/~kragen/bubble.html> The power didn't go out on 2000-01-01 either. :)
This archive was generated by hypermail 2b29 : Thu Mar 16 2000 - 15:57:28 PST