Mochabot log - CommonJS IRC channel: #commonjs on irc.freenode.net

2010-01-29:

[0:31] <isaacs> lulz... i was just looking up how to do a CommonJS proposal, and found this: http://wiki.commonjs.org/wiki/ProposalProcess
[0:32] <isaacs> it sounded really familiar as i was reading it. pretty sure i wrote the first draft of that page.
[0:34] <dmachi> the first four versions it seams :)
[0:34] <dmachi> seems
[0:52] <isaacs> yeah
[0:52] <isaacs> i was like, "whoa, who wrote this!? it's so well written, it's like it's exactly how i'm thinking it! they must be the best writer ever!
[0:59] <MisterN> isaacs: you need to throw in more ISO numbers
[1:00] <isaacs> iso numbers?
[1:00] <MisterN> like ISO 9000
[1:00] <MisterN> i know my jokes are bad :P
[1:26] <Wes-_> We could make the specification process ISO-8859-1 certified
[1:26] <Wes-_> That would impress all the MBAs
[1:27] <MisterN> lol
[1:27] <MisterN> Wes-_: is unicode an ISO standard? i'd prefer the number of it, then :)
[1:43] <Wes-_> MisterN: It is, actually, I forget the number, 646 maybe?
[1:58] <isaacs> So, I don't have ISO numbers for it, but go weigh in on this: http://wiki.commonjs.org/wiki/JSGI/StreamExtension
[8:46] <Dantman> *sigh* I wonder if we're underestimating the JavaScript community and in a indirectly rude sort of way trying to handhold the evolution of good APIs... Taking a quick look back, I believe most of the best APIs out on the web are the product of pure-js libraries made by developers with drives for practical applications and communities... Not standardization bodies and working groups..... In fact generally the working groups come out with the stuff that
[8:46] <Dantman> actual implementations of the spec get wrong and let the web suffer bloating code with compatibility layers (not as much of a problem for the server), and the stuff we love to hate and normally abstract over top of completely with cleaner api... Then again perhaps the hated api I'm thinking of are actually just stuff written in the pre-standardized era of the web. In which case the mandra would instead be "Low-level programmers who don't actually spend
[8:46] <Dantman> much time trenched in JS code made bad api decisions that actual JS programmers generally dislike..."
[8:47] <Dantman> *twitch* X-Chat needs a character limit warning...
[8:58] <Dantman> Heck... Rack isn't even part of Ruby...
[9:08] * oberhamsi is happy with the commonjs stuff that reached him as app dev so far.
[9:10] <Dantman> I'm talking about the bikeshedding of high-level api, instead of just standardizing the lowest necessary levels and allowing the high level api to evolve out in the real world where they are being used in practical application, and allowing the capabilities and api to be determined based on trial and error instead of theory and conjecture.
[9:11] <Dantman> *twitch* I'm getting to philosophical... To much excitement and energy drink at work, plus a good discussion with a friend...
[9:15] <Dantman> Who was it doing the stuff on running JS in the GPU? I thought one of our members was involved with that...
[9:24] <Dantman> Hmmmm.... my SpiderMonkey version of any canvas impl could allow hardware acceleration.
[13:21] <jammi> so wtf is going on?
[13:22] <DanielFriesen> Global spam, looks like...
[14:08] <DanielFriesen> I found out something interesting, yet disappointing about e4x
[14:10] <DanielFriesen> Long as you omit the surrounding () you can write something like <xml>{function() { /* large code block */ return someVar; }()}</xml> meaning you can use return to put the result of a big block of code into the xml tree.
[14:10] <DanielFriesen> And that includes other xml nodes.
[14:11] <DanielFriesen> However e4x can't parse any e4x inside the e4x, it can only be outside.
[14:11] <DanielFriesen> In other words this'll work `var a = <xml>...</xml>; var b = <xml>{function() { /* do stuff then */ return a; })()}</xml>;
[14:12] <ashb> DanielFriesen: e4x has all kinds of spec bugs like that
[14:12] <DanielFriesen> However this'll give a syntax error: `var b = <xml>(function() { /* do stuff then */ return <xml>...</xml>; }()}</xml>;
[14:12] <DanielFriesen> ashb, ^_^ e4x also makes string comparison /fun/...
[14:13] <ashb> :)
[14:13] <ashb> e4x is a really great *idea*. shame the spec sucks
[14:13] <DanielFriesen> Aye...
[14:13] <ashb> e4x is basiclaly LINQ for xml
[14:13] <ashb> building gcc takes for ever.
[14:16] <DanielFriesen> Biggest mistake was making it non-integrated with the existing DOM in browsers.
[14:17] <DanielFriesen> Or rather impossible to simply convert between them.
[14:17] <ashb> yeah
[14:17] <ashb> agreed
[14:17] <ashb> but the DOM isn't TC39 etc. etc.
[14:18] <DanielFriesen> One of e4x idea's straight up most interesting possible uses is the idea that you could embed part of your markup inside your script in a way you could flexibly modify it without piles of annoying escaping and whatnot...
[14:19] <ashb> yeah
[14:19] <ashb> <3 that
[14:19] <ashb> in fact thats what i do for the built in 404 template for Juice
[14:19] <ashb> its an e4x literal
[14:20] <DanielFriesen> I'm doing that for a very minimal part of my new website.
[14:20] <DanielFriesen> (The server-side generation)
[14:20] <DanielFriesen> ^_^ '<a href="'+attrescape(href)+'">'+htmlescape(text)+'</a>'; vs. <a href={href}>{text}</a>;
[14:21] <DanielFriesen> ;) or worse... "<a href=\""+attrescape(href)+"\">"+htmlescape(text)+"</a>";
[14:21] <ashb> yup
[15:17] <Wes-> DanielFriesen: var b = <xml>{(function() { /* do stuff then */ return <xml>...</xml>; }()}}</xml>;
[15:17] <Wes-> should work
[15:17] <DanielFriesen> Heh... too late
[15:21] <ashb> i appear to be writing crazy makefile rules:
[15:21] <ashb> cd $(dir $@) && ../$</configure $(value configure_args_$(subst -,_,$<))
[15:26] <dmachi> ugg, i hate doing make files
[15:27] <ashb> tyr doing a crossed-native gcc build
[15:28] <ashb> you'll hate that even more!
[15:28] <dmachi> i have unfortunately, luckily not anytime in the past 10 years though :)
[15:29] <dmachi> did it to get freebsd running on a DEC alpha back in the day, i think it may have caused problems in my marriage.
[15:30] <dmachi> ;)
[15:30] <ashb> i think i've managed to get gcc building
[15:30] <ashb> but libstdc++ is causing issues now
[16:08] <Wes-> dmachi: Ha! Should have used redhat, 4.2 ran on DEC out of the box
[16:09] <Wes-> ashb: wtf is $</configure syntax??
[16:10] <ashb> $< is the first dep
[16:10] <dmachi> Wes-: :) Well, I could be wrong about the dates, but I think this predated most standard linux distro's support of it, though I did have linux running on it. Was mainly just trying to make the freebsd install work to see if i could do it....i think i still even have that dec motherboard/cpu around here somewhere.
[16:11] <dmachi> Wes-: this was ~1996 and I was a slackware junkie...tried to avoid redhat :P
[16:13] <Wes-> dmachi: Ah, I was doing this in early 1998 with RedHat 5... although the previous owner the machine had had slackware on it IIRC
[16:13] <Wes-> ashb: Oh geez, LOL, it just doesn't "read" that way (thought you were doing kind of wierd XMLish thing)
[16:14] <dmachi> I still have a special place in my heart for slack, though I've not used it for 5 years or so.
[16:14] <Wes-> I don't
[16:14] <Wes-> Installing from floppies ONCE was enough for me
[16:14] <dmachi> hehe
[16:15] <dmachi> well we had ~350 machines with slack on them, so we worked the install part out :)
[16:15] <dmachi> big windows 3.11 fan too i suppose :P
[16:16] <Wes-> Windows 3.11 was only 6 floppies
[16:16] <Wes-> And I installed *that* sucker *WAY* more than 350 times
[16:17] <Wes-> Five or six machines a day for a couple of years, adds up quick
[16:17] <dmachi> true, though you weren't required to install slack via floppy (though a boot floppy was frequently required)
[16:17] <Wes-> dmachi: Well, NO, but CD burners were frigging $2500 back then!!!!!
[16:18] <Wes-> "Hmm.. Tuition or a CD burner" -- tough choice, but I didn't get the burner. :(
[16:18] <dmachi> yah, we did net installs almost exclusively
[16:18] <dmachi> yeah, and it was also work for me, so that part wasnt' an issue :)
[16:18] <Wes-> Heh. I was still using SLIP, often at 2400 bps
[16:18] <Wes-> The install would still be running
[16:18] <dmachi> ouch, yah
[16:19] <dmachi> man SLIP, haven't thought of that term for awhile
[16:19] <Wes-> The 14.4k modem banks where in such high demand that they would limit you to 8 hours per connection
[16:19] <Wes-> Here's another term: "Van Jacobsen Compression"
[16:19] <dmachi> hehe
[16:19] <Wes-> and "Trumpet Winsock"
[16:19] <dmachi> man, thats too funny
[16:20] <Wes-> "Kermit", "tn3270"
[16:20] <Wes-> Where was all the vocabulary gone?
[16:20] <dmachi> all replaced with acronyms

 

 

Logs by date :