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

2009-11-16:

[6:21] <Dantman> Aaaagh, Rhino isn't giving me file and line info
[6:42] <ondras> re
[21:03] <okito> tlrobinson yt?
[21:28] <tlrobinson> okito: hey
[21:29] <okito> hi. so I got Narwhal to install as a gem this morning
[21:29] <okito> to make it easier to bootstrap into using the environment
[21:29] <okito> so I can publish this gem now but before I did I wanted to get your input on how you want me to do it
[21:30] <okito> It is currently published as sproutit-narwhal: http://gemcutter.org/gems/sproutit-narwhal
[21:30] <ashb> bah. most markdown parsers suck. by which i mean very few ofthem have unit tests
[21:30] <okito> I know
[21:30] <okito> I published at sproutit-narwhal so narwhal would remain available if you wanted it.
[21:31] <ashb> interesting idea
[21:31] <okito> but I can do narwhal and try to keep it up to date if you prefer
[21:31] <okito> yes- the idea is to bootstrap narwhal btw. not to replace tusk.
[21:31] <okito> tusk is actually installed as part of this.
[21:31] <ashb> does the narwhall tree end up living under gems dirs or not?
[21:31] <okito> yes
[21:32] <okito> is that a problem?
[21:32] <tlrobinson> ok, yeah, i should probably snag "narwhal" on gemcutter
[21:32] <tlrobinson> not that there's any other open source narwhal projects (that i know of)
[21:32] <ashb> okito: probably not a problem, just lsightly odd
[21:32] <okito> tlrobinson: so I pushed my changes to make it publish at http://github.com/sproutit/narwhal.git
[21:33] <ashb> cna you not use a glob rather than a hardcoded list of files?
[21:33] <okito> if you want, you can pull those changes, then replace all mentions of 'sproutit-narwhal' with 'narwhal' in the Rakefile
[21:33] <okito> then rake release:all
[21:33] <okito> and that will claim narwhal for you
[21:33] <ashb> okito: s/ServerJs/CommonJS/ in s.summary
[21:33] <tlrobinson> cool, thanks
[21:34] <okito> ashb - that gemspec in narwhal right now is out of date
[21:34] <ashb> fair enough
[21:34] <tlrobinson> yeah that's really old, i should take it out
[21:34] <okito> tlrobinson - so do you want to publish this gem or should I keep a sproutit-narwhal published also?
[21:34] <okito> yeah - since github doens't do gems anymore. :(
[21:35] <okito> ashb - also, the Rakefile I added does use a glob. It also will attempt to build and use the JSC engine if you are on a Mac
[21:36] <tlrobinson> i can start publishing it
[21:36] <ashb> whats the state of hte JSC engine?
[21:36] <tlrobinson> thanks for setting it up
[21:36] <tlrobinson> JSC is pretty far along
[21:36] <tlrobinson> it's missing a few modules required for tusk though
[21:37] <tlrobinson> http and zip
[21:37] <ashb> happen to know off the topf of you head any docs about how its rooting stuffworks?
[21:37] <ashb> (rooting = protecting from GC)
[21:37] <okito> k cool. sure. you can also publish to rubyforge if you want. the command is in there. I will keep the Rakefile updated.
[21:37] <okito> in my fork and send u pull req if that's OK. it is the same Rakefile as I use for SproutCore so they should stay largely in sync
[21:39] <tlrobinson> ashb: JSC just has JSValueProtect and JSValueUnprotect
[21:39] <ashb> ah right. so just simple - root htis value, unroot this value
[21:39] <ashb> any sort of 'always root things created in this scope'?
[21:39] <tlrobinson> narwhal-jsc has some macros that wrap it
[21:40] <tlrobinson> (and theoretically do something similar in v8 though currently the v8 version is very leaky)
[21:40] <tlrobinson> ashb: hm i don't think so
[21:40] <tlrobinson> JSC's API is pretty simple
[21:40] <okito> it looks like you plan on having the JSC run on non-macs as well.
[21:40] <ashb> yeah - we're re-looking at our root<> api for flusspferd
[21:40] <okito> do you have it running on Windows?
[21:40] <MisterN> ashb: thinking whether we have to drop local_root_scope?
[21:40] <ashb> cos we just hit a bug in apple's gcc
[21:40] <tlrobinson> it should work with minimal effort on other unixes
[21:40] <ashb> MisterN: investigting yeah
[21:40] <MisterN> and because we break a lot API currently
[21:41] <tlrobinson> windows will probably be trickier
[21:41] <tlrobinson> at some point i'll take a look at mingw
[21:41] <MisterN> tlrobinson: i probably can't get your envy with this, but i try anyways: http://flusspferd.org/docs/group__create.html
[21:41] <ashb> MisterN: so we can write a version of v8's HandleScope for flusspferd and v8
[21:41] <ashb> er, SM and JSC
[21:42] <ashb> i wonder if HandleScope<> affects rooting of things in the JS space
[21:42] <tlrobinson> MisterN ashb I'm still open to using flusspferd in JSC/V8/SM versions of narwhal
[21:42] <ashb> (which spidermonkey's local root scope does
[21:42] <MisterN> ashb: do you think the examples would be more impressive to JS programmers if i didn't use the hard namespacing? :)
[21:42] <ashb> MisterN: yeah - probably
[21:43] <ashb> MisterN: did oyu ever create that amalgamation namespace?
[21:43] <tlrobinson> what's the status of JSC/V8 in flusspferd?
[21:43] <MisterN> ashb: no
[21:43] <MisterN> tlrobinson: nobody's volunteered to do it, yet.
[21:43] <ashb> tlrobinson: non existant, but http://redmine.flusspferd.org/issues/show/165 should make the introduction of them easier
[21:49] <ashb> tlrobinson: i might get bored and try writing an engine for one of them over christmas
[21:49] <ashb> or i might find something else ot hack on. who knows
[21:50] <ashb> i certainly want to try soon
[21:50] <ashb> (<3 months)
[21:57] <tlrobinson> I <3 months too
[21:58] <ashb> heh
[23:38] <ashb> kriskowal: "we embrace argument".
[23:38] <ashb> no we don't
[23:38] <ashb> ;)
[23:52] <kriskowal> yes we do
[23:52] <deanlandolt> now we don't!
[23:52] <deanlandolt> heh :)
[23:53] <kriskowal> yeah, meta
[23:54] <ashb> i resisted doing that on the ML at least

 

 

Logs by date :