Gossamer Forum
Home : Products : Gossamer Forum : Discussion :

GForum flexibility & performance compared to other bulletin boards or forums

Quote Reply
GForum flexibility & performance compared to other bulletin boards or forums
I searched the forum for performance & flexibility related threads, but I did not find too much of them.
I know, this was already discussed on the forum, but I did not find any final conclusions.
Just wondering, if there was detailed discussions about flexibility or GForum & compared to other bulletin boards...

Would be fine to get some info about flexibility, performance. Feature and speed comparisons to other BBS & forum software would be also fine.

Just a few ideas:

1) What is the maximum performance (without & with mod_perl) we can await from GForum?
- for small sites (50-500 users a day)
- for medium sites (500-2000 users a day)
- for big sites (2000-10.000+ users a day)
(not sure if the intervals are valid)

2) What are the advantages & features of GForum comparing to other BBS & forum scripts?

I know most answers to feature questions, but I have not a clue how would GForum perform under heavy usage... I think many new users would like a such comparative article.

Best regards,
Webmaster33


Paid Support
from Webmaster33. Expert in Perl programming & Gossamer Threads applications. (click here for prices)
Webmaster33's products (upd.2004.09.26) | Private message | Contact me | Was my post helpful? Donate my help...
Quote Reply
Re: [webmaster33] GForum flexibility & performance compared to other bulletin boards or forums In reply to
Regarding your first point of how well this forum copes under a server without mod_perl. I'd be very interested in this too, and would suspect with a default setup a PHP board would outperform this forum.

Of course, there are ways around this. You might want to contact ellipsiiis for more on information on her surviiivor boards. I believe she had to reconfigure her forum to work with PerlEx for it to sustain otherwise she would have consdered another forum.

Here's a link to the relevent thread:

http://www.gossamer-threads.com/...orum.cgi?post=184751

Hope this helps.

- wil
Quote Reply
Re: [Wil] GForum flexibility & performance compared to other bulletin boards or forums In reply to
Unfortunately, a CGI script simply cannot compete with a PHP script in terms of speed. PHP has the advantage of being preloaded and therefore doesn't have to worry about load time. A fair speed comparison here would be comparing Gossamer Forum under mod_perl to a PHP forum.

Unfortunately, mod_php is much more common on servers than mod_perl, partly because the mod_perl configuration is "unsafe" - meaning, essentially, that any process can grow uncontrollably and take all a server's memory and/or CPU resources. mod_perl 2 is supposed to do something about this problem to make mod_perl a more viable solutions for ISP's of shared servers to offer. I suppose the best we can do for now is hope!

Jason Rhinelander
Gossamer Threads
jason@gossamer-threads.com
Quote Reply
Re: [Jagerman] GForum flexibility & performance compared to other bulletin boards or forums In reply to
4 kind of tests would be useful:

Feature comparison tests:
1) Feature comparison for several (5-6) most known forum scripts.
(As this is a feature comparison, would not depend if it is written in Perl or Perl with mod_perl or PHP, but these would be listed as features.)

Speed comparison tests (Gforum speed compared to several other (5-6) known competitive forums):
2) Gforum compared to Perl forums (without mod_perl)
3) Gforum under mod_perl, compared to Perl forums using mod_perl
4) Gforum under mod_perl, compared to known PHP forums

I think this kind of comparison would be a fair comparison (same type to same or similar type).
All speed tests should be done in same or very similar environment.

Suggested environment for speed tests:
- Mysql databases (I don't think would be fair to compare to low performance flat text databases...)
- if possible, same content in each database
- 10 separate forums
- overall 50.000 posts in database (if possible each post should contain different text)
- 5000 post in each of the 10 forum
- 5.000 registered users in database
- no categories
- forum scripts must run using their own design (this affects only the download speed)

Suggested speed tests:
- page display time of the main or category page
- page display time of the forum thread listing page
- page display time of a thread page with 10 posts (or the maximum common post number, what the compared forums can display)
- elapsed time after posting a new post
- elapsed time after a search was started (e.g.: for word "test").
- how much thread page display requests are possible to be served without error in a second or minute?

All the result of feature & speed comparison should be listed in tables (4 tests, 4 tables).

I think would not be an easy task to do all the tests above, but would definitely worth for GT and the users, too.
If GForum would get the best results, then GT can be happy, they did a good job. If the results shows, that there are faster forums than Gforum, then could see the competitive products and could work on it to be better than them.

Anyway features, performance, flexibility, ease of use, price, together shows the worth of a forum product, so there would be some fine conclusions, which forum & when is good solution.
There would be one selected "Best product", based on feature/performance/price, and there would be Top winners of each category: "feature", "performance", "flexibility", "ease of use".

I don't know how the speed tests could be done to get comparative results. Using profiling script or by a separate timer tool?

I think a feature comparison would be fine anyway, but at least a few speed tests would be useful, too.

Ideas? Opinions?

Best regards,
Webmaster33


Paid Support
from Webmaster33. Expert in Perl programming & Gossamer Threads applications. (click here for prices)
Webmaster33's products (upd.2004.09.26) | Private message | Contact me | Was my post helpful? Donate my help...
Quote Reply
Re: [Jagerman] GForum flexibility & performance compared to other bulletin boards or forums In reply to
Thanks for the explanation of why ISPs do not load mod_perl by default. That is actually very logical and makes perfect sense. How does this PerlEx compare with mod_perl then?

- wil