CommonJS - Integration with more JS platforms

[kvz]

I asked Felix (he does a lot of nodejs stuff nowadays) how we could
close the gap between php.js and serverside js platforms like nodejs,
and it seems this is the answer: http://wiki.commonjs.org/wiki/Modules/1.1
Felix: "that just means that you need to put all php.js functions in 1
file and have each of them declared on the 'exports' object in that
file your file would be called php.js and I could include it via:
require('php');"

I imagine the compiler could create a 3rd version next to our normal
and namespaced: 'commonjs'.
Additionally we could provide a preset with just functions that make
sense to have on the server: sha1, md5, that kind of thing.
In that preset we'd have to drop the 'require' function in any case to
avoid confusion between commonjs' require & ours.

Changing the compiler is probably a day work. There may be more
problems in getting our functions compatible, I don't know. We could
have to investigate that if you guys like the idea.

[Brett]

I was told that we might be able to get php.js added as a module with
Cuddlefish--a modular system for Jetpack (which I think I remember
reading used CommonJS too, albeit on the client-side), thus allowing
people to build Firefox extensions with only a knowledge of HTML, CSS,
and php.js!

There's also something called FUEL for those building full-blown
Firefox extensions to which we could possibly get php.js added, but I
think Jetpack would be a higher priority if this module system
develops.

[Kevin van Zonneveld]

Ok that sounds pretty cool. Let me see how easy it is for me to
implement CommonJS in the compiler

--
Met vriendelijke groet / Kind regards,

Kevin van Zonneveld
http://kevin.vanzonneveld.net

http://twitter.com/kvz

[Philip Peterson]

I know this project's rather dead these days, but the code should at least be accessible without going through and copy-pasting all the functions into a giant text file. Is there something I can do to help fix the compiler?

Philip Peterson

[Kevin van Zonneveld]

Hey Philip,

You are right. In between work, bootstrapping transloadit.com, and
bugfixing system_daemon, php.js is not getting the time it deservers
anymore.
I'm willing to give the leadership to someone else if that's what it takes.
But that won't fix the compiler.

I have plans of migrating the website to a github pages site, so that
it everyone with commit access to php.js can generate & push a new
site live without my intervention.
In the meantime I will try to squeeze in some time to fix the
compiler. Cause obviously we the aforementioned plan will take a bit
longer.

PS
The actual compiling code is also part of the phpjs repo itself..

--

[Kevin van Zonneveld]

Think I fixed the issue, btw.

[Philip Peterson]

I'm taking a look at the source to see if I can start to understand how it works. The issue doesn't seem to be fixed; clicking the Compile button just takes me back to the compile page.

Philip Peterson

[Kevin van Zonneveld]

Getting "Successfully compiled" or "Found an existing package with the
same configuration" every time I try it.
What configuration do you use that results in a reproducible failure
on your end?

[Philip Peterson]

Functions: none, except for echo.
Compression: none
Namespaced: yes


Philip Peterson

[Kevin van Zonneveld]

Also getting succesfully compiled on OSX/Chrome 7.0
What platform does it not work on? What if you clear your caches?

[Philip Peterson]

Odd, I can't get it to work on Firefox, Chrome, or Safari on OS X.6. Tried clearing caches as well. I don't think I'm "just missing it" (the notification that the package has successfully compiled, etc.) but can you post a screenshot just so we're sure that I'm not just blind?

Philip Peterson

[Philip Peterson]

Also, even the default configuration fails to build.

Philip Peterson

[Kevin van Zonneveld]

http://dl.dropbox.com/u/343579/Screen%20shot%202010-11-27%20at%204.33.19%20PM.png

[Philip Peterson]

Nope, I don't get that at all. It just takes me back to /packages/compile.
Philip Peterson

[Kevin van Zonneveld]

No errors or nothing?

[carnio...@gmail.com]

No.
Sent on the Sprint® Now Network from my BlackBerry®

[Philip Peterson]

It may be of use to know that accessing the URL from the screenshot works fine; I see the summary and download link, and everything works fine.

Philip Peterson

[Kevin van Zonneveld]

That's the page you get forwarded to after a successful compile, but
the page remains available for a while as well.

I haven't found any errors in the logs since I pushed the fix either.

Would it be possible for you to figure out what's happening in terms
of HTTP requests, JavaScript error console, etc?
Does it appear to do a page reload? Or is nothing happening at all?
Does the POST get through?

Any more info on these matters would be helpful in finding the cause.

[Philip Peterson]

No JavaScript errors; seems there's one "failure" I found by logging the HTTP headers:

http://cl.ly/362C2D1a1r0V2c2o172C

"Client closed connection before receiving entire response"



Philip Peterson

[Kevin van Zonneveld]

Could it be a proxy times out or something? Have you tried another environment?

[Philip Peterson]

I've tried several public computers, all which result in the same failure. The only proxy being used was the Charles proxy (to capture the HTTP headers). I still get the error when not using the proxy, so it doesn't seem like that would be the issue.

Philip Peterson

[Kevin van Zonneveld]

I'm still looking for a machine where I can reproduce this on, Philip.
It's impossible to fix otherwise.

[Philip Peterson]

@KvZ:

It's really odd that I can't find a computer that doesn't reproduce it, and you can't find one that can :P. Perhaps (weird/unlikely as it sounds) it's a geographical thing... maybe you could check that your client's HTTP request headers (if they aren't already) include the following?

Accept-Language: en-us,en;q=0.5
Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7


Philip Peterson