(Fwd) Re: Ideas for the future of Squid

------- Forwarded message follows -------
From: Andres Kroonmaa <andre@online.ee>
To: Henrik Nordstrom <hno@hem.passagen.se>
Subject: Re: Ideas for the future of Squid
Copies to: squid-dev@ircache.net
Date sent: Tue, 28 Mar 2000 14:50:50 +0200

 I'm glad to hear that. Also glad to see that Duane sees this as a
 continuation and natural development of Squid project.
 Back in '97 such ideas met very little interest, but I agree with you
 (and always been of that opinion) that any change to squid is possible
 without the need to start another project.

 Perhaps we should start a debate about what should Squid next major
 release (3.0?) "look like" in terms of overall design?

On 27 Mar 00, at 19:54, Henrik Nordstrom <hno@hem.passagen.se> wrote:

> The change today is not as big as it was once in a time. The design is
> slowly moving to a modularized design. The store I/O is mostly finished,
> and next in chain is replacement policies and state logging.
>
> I agree that it is a rather complex task, and it will take a couple of
> iterations to accomplish it. My goal now is to find the design goal, and
> then to see what we can do to reach that goal. I do not agree that it
> wouldn't be Squid. In fact the changes may not be as large as you first
> think and then larger in other areas you don't think about at first.
>
> /Henrik
>
> Andres Kroonmaa wrote:
> >
> > Unfortunately, as you very well know, this is way too radical change,
> > and although many people might agree with principles, it will get stuck
> > due to complexity of the task. Besides, and it have been pointed out
> > few times, that this wouldn't be Squid any more. It would be another
> > project derived from Squid, so NLANR support would probably be quite
> > small. Therefore, radical changes can't happen too fast.
> >
> > I also believe that Squid has not reached its potential as it is, and
> > that we should for now focus on reaching its current limits in terms
> > of performance, scalability and fault tolerance. And while we are at
> > that, we might implement changes with keeping in mind genereal future
> > directions of Squid. In this sense it is ok to try to define a path
> > where the Squid development should aim to. If we can "prepare" Squid
> > project for modular design in such a way that it would be easier in
> > the future to switch to totally modular model, we might have much
> > more supporters for that change to happen.

------- End of forwarded message -------

------------------------------------
 Andres Kroonmaa <andre@online.ee>
 MicroLink Online
 Network Development Manager
 Tel: 6501 731, Fax: 6501 708
 Pärnu mnt. 158, Tallinn,
 11317 Estonia
Received on Tue Mar 28 2000 - 06:13:35 MST