pouët.net

jsRocket

category: code [glöplog]
mog: Huh, did my websocket-branch break compatibility with the C++ code? I didn't intend to...
added on the 2013-03-07 13:31:51 by kusma kusma
Well, it doesn't work without websocket - thus the exe can't be used with the cpp example, or am I off track there?

And for the example demo link from the previous side:
3ln - Feliz Navis ASD

:)
added on the 2013-03-07 17:10:02 by mog mog
mog: The intention of the code on the websocket branch is to add support for WebSocket without removing support for normal sockets. So at least in theory, the editor that you built should be able to speak to both JavaScript and C/C++ clients. But I didn't actually test that the C/C++ clients can still connect. The websocket-hack branch were less helpful to C/C++ clients, though :P
added on the 2013-03-07 22:22:05 by kusma kusma
Oh! That's even cooler then! :D
added on the 2013-03-07 22:30:55 by mog mog
Turns out, I had slightly broken the non-websocket code-path. Pushed out a fix, and considering taking this upstream soon.
added on the 2013-03-11 23:28:10 by kusma kusma
Damn. And here I was watching the Excess Dropbox coming to life was the early signs of a new demo ;)
added on the 2013-03-12 09:01:42 by gloom gloom
Good job guys.
added on the 2013-03-12 10:40:30 by p01 p01
This is awesomeness, thanks guys!
added on the 2013-04-17 02:05:34 by Fell Fell
Hmmm -- all was going wonderfully; with mog's jsRocket-RocketWebsocket and kusma's rocket-work-websocket I found myself in a veritable nirvana of synchronicity and was merrily hacking away on some manner of glorious Rocket-enriched scene management framework when SUDDENLY!

I chanced to look upon the window, and spied a large barn owl sitting on the windowsill gazing beadily back at me. At least, I think it was a barn owl. I'm not an owl expert.

"What the fuck!" thought I, when the awkward silence was broken by the owl: "Dude -- You should probably try this in Chrome".

Aghast, I turned to Chrome and loaded the prod; Chrome said no!

For the threeJS_Cube example, for instance:

XMLHttpRequest cannot load file:///D:/Code/jsRocket-RocketWebsocket/jsRocket-RocketWebsocket/example/threeJS_Cube/cube.rocket. Cross origin requests are only supported for HTTP.

Heart racing, I turned back to the window. The owl was gone.

Any thoughts?
added on the 2013-04-18 02:06:34 by Fell Fell
That darn owl should have told you to either run the stuff on your localhost, or feed --allow-file-access-from-files as parameter to Chrome.

For the latter one it's crucial to close all Chromes before, otherwise two owls may appear :)

The cause for why this is necessary, imagine any website being able to read stuff from your harddisk - jupp, would suck :)
added on the 2013-04-18 08:51:33 by mog mog
Mog, thanks very much for the ornithological clarification :)
added on the 2013-04-18 10:21:23 by Fell Fell
Quote:
Hmmm -- all was going wonderfully; with mog's jsRocket-RocketWebsocket and kusma's rocket-work-websocket I found myself in a veritable nirvana of synchronicity and was merrily hacking away on some manner of glorious Rocket-enriched scene management framework when SUDDENLY!

I chanced to look upon the window, and spied a large barn owl sitting on the windowsill gazing beadily back at me. At least, I think it was a barn owl. I'm not an owl expert.

"What the fuck!" thought I, when the awkward silence was broken by the owl: "Dude -- You should probably try this in Chrome".

Aghast, I turned to Chrome and loaded the prod; Chrome said no!

For the threeJS_Cube example, for instance:

XMLHttpRequest cannot load file:///D:/Code/jsRocket-RocketWebsocket/jsRocket-RocketWebsocket/example/threeJ S_Cube/cube.rocket. Cross origin requests are only supported for HTTP.

Heart racing, I turned back to the window. The owl was gone.

Any thoughts?

Yes. You win at life with that post right there :)
added on the 2013-04-18 11:10:39 by gloom gloom
BB Image
(no Wacom at work though ;_;)
added on the 2013-04-18 12:41:10 by mog mog
Another error that bugged me for quite a while until I put some thought into, is the Rocket limbo between two rows when one moves quickly in Rocket.
You usually render when you get a row update from Rocket, though you also want rocket to know where you are in the demo when you play it. So one usually updates the row towards Rocket in render as well.

However if you send the row update while not playing the demo, you might end up in the row limbo. The problem seems to be JS not giving accurate enough position data on the audio, so you send row 4 on getting row 5 from Rocket - and that keeps going on forever, until you manage to slip out by hitting spacebar like a mad man in Rocket.

The current fix is to only update the row towards Rocket when playing the demo.

The Rocket spasm fix as source

(This isn't a Rocket error, but might be something I can address in jsRocket - if I bring more thought into it O_o)
added on the 2013-04-30 23:12:59 by mog mog
Just thought I'd let you know that I've *finally* gotten my act together and pushed out websocket support to the master-branch.
added on the 2013-05-07 11:36:14 by kusma kusma
Cool! =)
added on the 2013-05-07 11:58:05 by mog mog
And now I've finally gotten my act together, and merged jsRocket into the main GNU Rocket repo! Thanks a lot for the great work, mog!
added on the 2016-01-15 22:15:39 by kusma kusma
fuck yea kusma :)
added on the 2016-01-15 22:41:15 by visy visy
<3
Also there's now a list of alternate ports!
added on the 2016-01-15 22:43:19 by mog mog
wooot :)
added on the 2016-01-17 18:51:50 by wullon wullon
cool!
added on the 2016-01-17 20:41:28 by psenough psenough

login