Evil Trout Posted:

I didn’t respond to this because I already addressed how streaks would work on the blog. I went with McBanhammer’s solution.

If you want feedback on your post, I find it a poor solution. Encoding things into strings make it way more complicated for no obvious reason. The disk space savings aren’t worth the difficulty in querying or complicated and error prone nature of the programming.

I just used a streaks table. It’s (forum_id, character_id). If there’s a duplicate constraint the streak is reset. If you die, the next time you enter a forum the streak is reset. It’s working well now!

I… really don’t see how it can make things “more complicated”. In fact, it significantly simplifies the implementation! No need to make any queries at all to check for anything… you just run simple string-search and string-chopping functions, which by themselves are even faster than preg_match (perl regular expressions – also not necessary here), and ITSELF is even faster than a single MySQL query! It’s like an all-you-can-eat buffet of optimizations and simplified code, and it really quite boggles my mind as to how you can say it’s “more complicated”...

Perhaps we’re not understanding it the same way?