Avatar
Paul
aebbe11c4358f6fe8a17bd121c86a972f38e8b1baac733b03eadd05f2c344e82
UK Whovian, railway & politics geek and event organiser. Freelance PHP developer and Linux system administrator. He/him/Paul Most toots deleted after 3 months Book reviews and updates: @pwaring #NoSearch #NoBot #NoArchive #NoIndex #NoBridge

Couple of recent changes that fellow Fosstodoners may have missed:

1. English-only public toots policy has been dropped (which I am pleased by).

2. Sign-ups are now invite-only (slows growth, which I think is a good thing - I would rather be on a mid-sized instance than something like mastodon.social).

#Fosstodon

Replying to Avatar Paul

nostr:npub19lt4284mghqxekzm6n5njxurnxrxhqhrva2leusdsuu5ja5jeycq66qfjk If you want a rough and ready figure you can call microtime before and after and compare the results, likewise memory_get_usage. They aren't great but they will probably tell you if one function is taking 10x as long as another.

If they're not enough then really Xdebug + an analyser such as KCacheGrind are the way to go.

https://3v4l.org/ allows you to run code online but I don't think it would help with profiling.

nostr:npub19lt4284mghqxekzm6n5njxurnxrxhqhrva2leusdsuu5ja5jeycq66qfjk In my experience though, the things that really kill performance are function calls and database queries, especially if performed in loops. There's not usually a huge gain from tweaking other parts of the code.

nostr:npub19lt4284mghqxekzm6n5njxurnxrxhqhrva2leusdsuu5ja5jeycq66qfjk If you want a rough and ready figure you can call microtime before and after and compare the results, likewise memory_get_usage. They aren't great but they will probably tell you if one function is taking 10x as long as another.

If they're not enough then really Xdebug + an analyser such as KCacheGrind are the way to go.

https://3v4l.org/ allows you to run code online but I don't think it would help with profiling.