Re: Request bodies.

From: Henrik Nordstrom <hno@dont-contact.us>
Date: Sun, 26 Apr 1998 05:45:47 +0200

--MimeMultipartBoundary
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Alex Rousskov wrote:

> I assume you are talking about pump.c replacement. Why do you
> want to replace pump.c?

Because I had so many problems with it that it was easier to start from
scratch
* Assertions failing on aborted transfers
* Pipelined request cut short
* Messy flow of control

> Are you positive it is worth the effort? The current version may have
> a few problems left. However, it may be easier to fix those compared
> to starting fixing fresh problems that a new huge patch always has.

Yes, I am positive that it is worth the effort.

+ A clean isolating interface between client and protocol code.
+ Normal flow of control (in terms of Squid flow of control)
+ Rather small change by code logics.
+ The old code had several problems
- Partly new untested code
- A new flow of control, which may reveal other problems.

> When there is virtually no a priory discussions, we are facing a
> tough choice: two huge pieces of code, one we were working on for
> a while and a brand new one. Both are supposed to work, but both
> usually require bug fixing. There are a lot of _new_ features that
> should be added to Squid. Why rewrite something that is working
> already?? Just trying to understand what I am missing here...

I have said several times the last months (or at least tried to) that
I'd like to have request processing rewritten in a way similar to this,
but I do admit that I never asked for a open discussion on the subject.

I am open for discussions on every piece of code I write, and welcome
any comments or even refusals (if motivated).

I do expect the people at NLANR to carefully read the changes I send,
and if something is unlclear ask me about it. Writing code is a way of
expression..

On a side point: My Squid coding is purely for the fun of it and has
always been, and my coding is done in respect of that.

/Henrik

--MimeMultipartBoundary--
Received on Tue Jul 29 2003 - 13:15:48 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:11:45 MST