[11:04:44] <jlast> hello
[11:04:49] <arthurvr> hey
[11:05:09] <ariya> good morning/afternoon/evening everyone!
[11:05:26] <gibson042> helllllo
[11:06:43] <ariya> let's get started
[11:07:39] <ariya> we got finally Sauce Labs hooked up: https://saucelabs.com/u/esprima
[11:09:10] <jlast> nice
[11:09:10] <ariya> Travis CI is also running the test suite with the fresh Node.js 4.0
[11:09:24] <ariya> JSCS and Esmangle are included in the downstream test as well
[11:09:41] <ariya> jlast: our test suite doesn't run on IE 8, just in case you want to help
[11:09:51] <ariya> it's Mocha problem, as well as Object.keys usage
[11:10:06] <jlast> what's the problem?
[11:10:06] <ariya> probably just needs to be polyfilled, but only if we care of running the tests on IE8
[11:10:17] <ariya> Array#filter, or something like that
[11:10:32] <jlast> yea, that'd make sense
[11:10:37] <jlast> is there an issue for it?
[11:11:31] <ariya> not yet :-)
[11:11:59] <ariya> last time I tried IE 8, it was not fast enough to do any meaningful AST transformation
[11:12:10] <ariya> so probably it's the point of diminishing return
[11:12:35] <ariya> we're already fine with IE9 all the way to Edge
[11:12:47] <ariya> jlast: your Karma-ification of our test suite pays off!!
[11:13:36] <jlast> got it. yea, that makes sense. Really happy about how karma turned out. great job getting the multi-browser environment figured out
[11:14:01] <ariya> another advantage of doing it a bit slower but careful and thoughtful :-)
[11:14:28] <ariya> looking at http://esprima.org/test/ci.html, I think we're as stress-tested as it could be made :-)
[11:14:34] <ariya> not a lot of project can claim that
[11:15:54] <ariya> I think that's all that is happening last week
[11:15:55] <jlast> awesome. I was thinking about trying to make the testrunner publicly available
[11:16:24] <ariya> jlast: BTW do you want to submit a talk for the upcoming Fluent 2016?
[11:16:25] <jlast> that way, there could be a link to the tests that any developer can see. use case, you're wondering if esprima covers the case you're considering
[11:16:29] <jlast> also, just cool
[11:16:55] <ariya> jlast: or trigger the demo parser with the code
[11:17:06] <ariya> just like http://esprima.org/demo/parse.html?code=%2F%2F%20Life%2C%20Universe%2C%20and%20Everything%0Avar%20answer%20%3D%206%20*%207%3B%0A
[11:17:43] <ariya> jlast: on the subject of Fluent, we need more videos to populate http://esprima.org/doc/videos.html :-)
[11:17:58] <ariya> I think an updated version of "Building JavaScript Tools" is desperately needed
[11:18:18] <jlast> like a re-skin of the page?
[11:19:02] <ariya> no, a new fresh talk
[11:19:12] <ariya> with all ES6-ness and stuff
[11:19:25] <ariya> or e.g. plugin for a linter (eslint)
[11:19:34] <jlast> yea, definitely
[11:19:54] <ariya> CfP deadline is monday, think about it :-)
[11:20:14] <jlast> oh, i see what you're saying. alright - will mull it over
[11:20:20] <ariya> let me know if you need another pair of eyes to review the abstract
[11:20:39] <jlast> thanks, will do
[11:20:56] <ariya> on a different subject, espree will stop being a fork of esprima: https://github.com/eslint/espree/issues/200
[11:21:31] <ariya> we're down to just a few parsers in the world: esprima, acorn(-jsx), babylon
[11:21:50] <jlast> yea, i think there's been an espree-acorn fork for awhile
[11:22:26] <jlast> i guess there's also the browser parsers as well?
[11:22:50] <ariya> only spidermonkey exposes the parser api
[11:23:07] <ariya> jlast: btw, regarding your tree visualization, it seems to make sense to merge it into astexplorer
[11:23:17] <ariya> I plan to link to it, from the demo parser
[11:23:27] <ariya> since astexplorer is much more featureful + fancy
[11:23:35] <jlast> ariya - completely agree
[11:23:48] <jlast> was going to try putting the AST explorer on the esprima homepage
[11:24:16] <jlast> I can try doing that on the demo page first...
[11:24:31] <ariya> isn't that going to be too overwhelming?
[11:24:45] <jlast> ariya, it could be
[11:25:13] <jlast> if it's done right, i hope it would help answer the question - can a parser help me build this tool?
[11:25:20] <ariya> we could keep the simple viz on the landing page, with a strong hint to go to astexplorer for the power user
[11:25:35] <jlast> yea, i like that.
[11:25:48] <jlast> anyways, i like the direction this is going in
[11:26:07] <jlast> and yea, working on top of the AST Explorer is a great suggestion
[11:26:16] <ariya> cool, feel free to come up with a proposed design
[11:27:08] <ariya> jlast: are you using sublime or textmate or another editor?
[11:27:22] <ariya> I'm thinking of an editor plugin using esprima pure tokenizer for the syntax highlighting
[11:27:53] <jlast> sublime
[11:28:03] <ariya> cool
[11:28:14] <ariya> I think a proper ES6 highlighter could be a nice demo
[11:28:25] <ariya> that means we need to have a delegate support in the tokenizer
[11:28:35] <jlast> yea, i can see that being succesfull.
[11:28:40] <ariya> processing the array returned by the tokenizer function is consuming too much memory
[11:28:51] <ariya> I'll write down a summary in a ticket, soon-ish(tm)
[11:29:17] <jlast> have to head out
[11:29:20] * jlast waves
[11:29:27] <ariya> bye jlast
[11:29:44] <ariya> anyhow, that's all I have this week
[11:29:59] <ariya> gibson042: are you planning to go to Fluent?
[11:30:34] <gibson042> actually, it hasn't made my radar
[11:30:54] <gibson042> http://fluentconf.com/javascript-html-2015  ?
[11:30:55] <ariya> gibson042: you should, it's a fun event :-)
[11:31:06] <ariya> in case you do, let's us know
[11:31:11] <ariya> would be fun to hang out in person
[11:31:37] <gibson042> when is it?
[11:31:43] <ariya> early march
[11:32:30] <ariya> think about it and maybe you'll make a decision next week or so
[11:32:36] <ariya> I go to go now
[11:32:42] <ariya> thanks everyone and see you again next time!
[11:32:43] * ariya waves
[11:33:23] * gibson042 waves