Re: object oriented Squid ?

From: Wayne Salamonsen <wayne@dont-contact.us>
Date: Sat, 29 Nov 1997 14:27:48 +0800

--MimeMultipartBoundary
Content-Type: text/plain; charset="us-ascii"

Just to throw in my 2c worth.

I think that all of the afore mentioned changes would be fantastic. As a C,
C++ and now Java programmer I reckon OO design is the way to go. I believe
that most people out there who are writing code for Squid would be fairly
conversant with C++. It seems to me that coping with C++ and OO is probably
easier than learning an uncommented Squid source. :-)

I would also love to see threading introduced to Squid....and comments in
the source. WOW! That'd make life a whole lot easier, especially for those
trying to work their way through the code for the first time.

However, to me, all of these changes seem to be things that will take a
large amount of work. Given that Squid 1.2 is comming along, still in C
and unthreaded, to then sit down and convert the lot to C++ and using
threads will take someone (or ones) a long time. Duane has done a great
job in maintaining the code and putting together everyone's work on 1.2.
But I dunno if he would have the time to do or co-ordinate such an effort?
Unless Kostas has a lot of time on his hands..... :-) Heh Heh Heh.

Just my initial thoughts.....

Wayne

At 08:33 PM 11/28/97 -0600, Alex Rousskov wrote:
>On Sat, 29 Nov 1997, Henrik Nordstrom wrote:
>
>> As A sparetime Squid developer that knows both C and C++ (and a bit of
>> Squid too), I must say that there are some other design issues that is
>> fare more interesing than rewriting it in C++..
>
>Agree. I was just trying to make a point that having Squid in C++ can
>speedup the development of other, more interesting issues.
>
>> 1. A Threaded Squid. This would really speed up development since
>> threading makes it possible to have a linear flow of control. Keeping
>> track of 3 (or more) interrelated state machines is not at all easy. All
>> developers makes mistakes here sooner or later (including Duane W).
>
>I agree that having a linear flow of control would help a lot. However, I
>do believe that switching to threads will create far more problems than it
>will solve. Just take a look at the threaded (in one way or another) code
>of Web servers like Apache. I am positive one will spend more time locking
>the stuff and debugging the deadlocks than keeping track of "state
>machines".
>
>On the other hand, it is possible to achieve "logical" threading with a
>proper use of C++. "concurrent" state machines could be made almost
>transparent to a developer using a proper design.
>
>> 2. Comments in the code, describing what it does, and why...
>
>Agree 100%.
>
>> I don't beleive writing it in C++ alone makes
>> Squid much easier to develop.
>
>A good OO design could help adding new modules and algorithms much easier
>though...
>
>Alex.
>
>
-----------------------------------------------------------------
Wayne Salamonsen
R&D Engineer

Internet Research & Development Unit Email: wayne@irdu.nus.sg
National University of Singapore URL: http://irdu.nus.sg
Computer Centre Ph: +65 771 5238
10 Kent Ridge Crescent Fax: +65 872 6205
Singapore
-----------------------------------------------------------------

--MimeMultipartBoundary--
Received on Tue Jul 29 2003 - 13:15:44 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:11:30 MST