Comments on: OMG! as3-signals are faster than events!!!!!1

By: Yanis Benson

Yanis Benson — Sat, 06 Mar 2010 09:37:22 +0000

Haha! No, callbacks can be slower than signals.
You forgot one essential part – to check “if (callback != null)”. In a real program, half of the times callbacks are null, and for null cases my signals code is 3+ times faster than callbacks.

By: Дмитрий

Дмитрий — Tue, 16 Feb 2010 15:11:32 +0000

http://jacksondunstan.com/turbosignals

By: Diomas

Diomas — Wed, 27 Jan 2010 07:03:01 +0000

С колбэками это хорошо, Вы, Валентин, стебок :)

By: Valentin

Valentin — Tue, 26 Jan 2010 20:12:02 +0000

Indeed. But callbacks will be faster anywhere.

By: Robert Penner

Robert Penner — Tue, 26 Jan 2010 19:50:16 +0000

The Debug Player isn’t the best for measuring performance.

By: Alec McEachran

Alec McEachran — Tue, 26 Jan 2010 18:27:13 +0000

As it happens, I wrote the performance test code at the behest of a couple of other coders who were interested in the results. It might have been useful to compare it with callbacks, but my difficulty with callbacks are that they are less manageable over large projects. Thank you for adding a comparison with callbacks.

The purpose of my producing graphs was to demonstrate the relative speed differences at a casual glance. You might have poured scorn over my choice of graph: 3D? I was careful to say explicitly that you shouldn’t take my results as scientific, so the graphs are purely illustrative. They are easier to read, and allow busy developers (who are the vast majority of people who look at my blog) to understand the results as quickly as possible.

Other than that, I find your tone unreasonable and offensive. Certainly I didn’t say anything stupid, even if it is trivial. I didn’t claim that my results were original, that they were important or that they deserved attention. Of all the posts on the internet that your bile could have been targeted towards, I find it really strange that you should aim it at mine.