Publish/Subscribe is not a replacement for Observer based patterns, such as most event systems are implemented.

I am working on an article that shows the difference between the two patterns, and discusses when the use of each of them is appropriate.

Thanks for the feedback, I've updated the first version of the test case to use the raw version of PubSubJS, http://jsperf.com/pubsubjs-vs-jquery-custom-event...

I am currently doing some refactoring of PubSubJS in a private branch, to see if I can speed up the asynchronous publishing of messages. If my hunch is correct, the speed will be much closer to synchronous publishing of messages. Hopefully I will soon find the time to complete this, and if JSPerf testing shows that my research has been successful, I'll merge it into master and do a new release.

I am also working on a little blog post examining the differences between the Observer pattern and Publish/Subscribe.

Interesting.

I think it would absolutely be worthwhile to do a code retreat in Copenhagen. If it turns out to be a success, then I think we could probably do it more than once, and perhaps also do one in Malmö.

The message is passed, so you can use the same subscriber for several different messages. Data is usually the optional argument, but in pubsub / events systems is customary to always pass the message or event.

No, not at all ... it was an imagined scenario, for the sake of argument.

I've not proved that it exists or doesn't, and with no test cases, nor has anyone else ...

Hostdude, thank you for your comments!

Perhaps it IS a contradiction, that I'd like to see other plugins blessed and become part of jQuery core. Perhaps a better solution would be rewriting parts of jQuery core where it could be stronger, just like what's going on with Ajax in 1.5b1 (basically a complete rewrite, which greatly improves error handling).

I agree, it could become a rather lengthy discussion of which plugins should become part of the core offering. Perhaps a better solution would be more official plugins, that would be actively maintained by the jQuery team (or a jQuery plugins team)?

Glad to hear it :)

Just tried it out with a newly installed OSX (new computer) and Safari 5.0.3, it still works :-)

Glad you liked it.

I wasn't really sure if I was being clear enough about the message that I wanted to get out there, but just like writing code, writing good blog posts takes practice :-)

I am planning on writing a blog post about jQuery plugins that I DO think should be merged into core, or blessed and endorsed by the jQuery team. Any nominations would be great :-)