2.8 seconds is locking the browser

3.3 seconds is not locking browser

When dealing with web it is unacceptable to lock the browser. Also with core / worker distribution you may be able to make the worker approach faster :).

That's something we disagree on then, in my humble opinion, on the web it is unacceptable to make users wait longer then necessary. And it's not a small, 1 or 2 percent difference. And certainly not longer then 1 second.

And why is the example then not setup with core/worker distribution so it actually does as you write 'potentially' work faster?

Nice i can still use my browser while i wait longer for the stuff i _want_ to see. But what is the use? I want to see that data/page i click the link or button for, nothing else.

Or please elaborate why it is in _your opinion_ unacceptable to lock a browser but get the data on screen faster?

Its not just something me, and you disagree on. I would say it's me, plus a good amount of other, and also most browsers.

If the browser is locked over X seconds it will actually tell you that the script is unresponsive, and ask if it should stop it. But you can of course work around this, and disable it in your browser of choice (http://www.reginout.com/nflp/fix-unresponsive-script-error-i...). Or you could just use a webworker, and not lock the UI.

If my opinion differs from you or anybody else, that would be considered a disagreement.

I never see my browser being locked, and if it is locked waiting for the information i desperately want to see, i honestly couldn't care of my UI locks up during that. I don't care about anything else at that point.

I understand this is not how most people look at it and to you there is just one way and thats the responsive UI way. But i see it as they rather have nice spinning wheels but a car that won't drive anyway..

And if that test uses a sleep, which is specifically says it does not, it would have even been a bad showcase. Now it performs bad (people suggest using multiple workers.. lol for a bubble sort)...

Be critical about a non locking UI that doesn't show your information anyway. Fix the application it self so it wouldn't need to lockup in the first place.

This demo is not a demo of performance, it's a demo of the UI not being locked up. It could've used a 3 second sleep for all that the sorting itself mattered.

That's not how you should think about this.

It's 2.8 seconds of unresponsiveness vs. 0 seconds of unresponsiveness.