taijinlee
commented
12 years ago i think this is good enough for r0 (it runs). Will beef it up in the future as needed. Will probably need to add:
- logging of sorts so we can track down bugs when stuff breaks
- make the cron detect previous runs taking too long and killing them
In the long run, this is probably a bad idea, but it seems a lot nicer to not need to manage crons at the moment (for production and dev), and it's very little amounts of code. At some point in the future, we might need smarter dynamic crons, which we could leverage this anyways, so I don't think it's a complete waste.
Idea is register a cron and it'll call itself. I'm not sure if there's a memory leak doing it this way or not... we'll see.