Smartys wrote:For a couple reasons:
A. What happens when I have two identically named forums?
B. It makes the actual text after forum# irrelevant: you can change the name of the forum and keep the same link
C. If we're given the ID in the URL like that, we don't have to do a scan through the forums table for all forums matching that name.
I addressed this in my version of punbb.
A. If you have a non-unique forum title. append the forum id. That will automatically make it unique (obviously still need to check it ... but 99 times out of 100 it will be true)
-- Forum named: The Main Forum => "the-main-forum" (id 1)
-- Forum named: The Main Forum => "the-main-forum-2" (id 2)
C. My solution was to create an md5 hash of the url part and save the hash in a char(32) with a unique index
-- id = 1 -- 288c7cc23607ce5096d62bbe7c7afb1e
-- id = 2 -- 3f93a8178bb0650749b12ee235c15f98
C. b) I dont really see this becoming a problem for the forum link lookup (unless there are millions of forums hehe), but it could become problematic for the topic link lookup (both work the same way), but we can go ahead and eliminate a considerable amount of threads if we know the forum id. Millions of topics is reasonable, but isnt that why we are able to build db clusters with quad-core processors and scsi drives?