[09:00:50] hellloooo! [09:01:06] hellooo! [09:01:12] okay, we got someone! [09:01:22] hey arthurvr! [09:01:25] sometwo, even [09:01:27] thanks for your help on the docs [09:01:36] can we go for somethree? [09:01:42] yes, thank you [09:02:17] like all good software projects, we tend to abuse our docs :) [09:02:32] i really do want to stop doing that tho [09:02:32] haha, Hope I can help out ;-) [09:03:34] m_gol you around? [09:03:34] present [09:03:36] gibson042: what did you think about https://github.com/jquery/jquery/issues/1969 [09:04:34] seems like there is at least the potential for problems if we throw early out of that block [09:04:38] and miss some feature detects [09:04:45] but none of them affect firefox afaict [09:04:55] yes, that's true [09:05:00] so it doesn't cause problems at the moment [09:05:10] it's probably worth documenting with a comment [09:06:22] and a mozilla bug report [09:06:56] yeah i wonder if it is xhtml related [09:07:04] damn screwed up the link to sizzle [09:07:43] there we go [09:08:18] maybe \f isn't legal in xhtml values? [09:08:29] might need to be encoded [09:08:29] hmmmm [09:08:50] oh did i mention I HATE XHTML [09:09:29] ugh [09:10:11] we really do need some sort of sanity check for XHTML [09:10:56] if we want to work in XHTML, this sanity check will probably mean most of the test suite [09:11:05] altho in this case it wouldn't have helped [09:11:21] actually, is there anything that'd work in XHTML and not work in HTML? [09:11:22] well just something to keep us from looking stupid when it won't even load in xhtml :) [09:11:47] hmm [09:12:08] i can't recall, there are issues in oldIE but that may not matter [09:12:15] i think ie11 added xhtml support [09:12:24] and actually has the right doctype [09:12:36] the doctype doesn't matter for XHTML [09:12:38] formally [09:12:45] at least not for XHTML5 [09:13:04] only the mime type served is taken into account [09:13:18] unless it's a local file [09:13:24] then the extension matters [09:13:30] at least for ff [09:13:43] but that's not relevant for our swarm testing [09:14:35] local files are not served by a web serwer so Fx is probably proxying them and guessing the mime type in the process :) [09:14:37] but yeah [09:15:12] at any rate, I don't see the message when loading our test suite in Firefox, which does lend credence to the XHTML hypothesis [09:15:12] the only reason I wouldn't like running XHTML is that it's not the way 90% of jQuery users run it [09:15:46] yeah, if there are any edge-case bugs we wouldn't catch them, that'd be bad [09:15:58] maybe it's time for another weekly run ;) [09:16:11] might be the best way to fix it [09:16:26] can we change the mime type for just one run? [09:17:10] that might be tricky [09:17:26] there's probably some trick where we could serve it from a different place [09:17:29] we can set env variable for that & use it [09:17:35] or pass in something on the url [09:17:40] yeah [09:17:45] we'd need to change the start template to PHP [09:17:49] m_gol can you take a look? [09:18:02] there's an existing ticket for xhtml let me find it [09:18:20] k, just probably not before next week :) [09:18:22] https://github.com/jquery/jquery/issues/1731 [09:18:29] yeah this isn't a priority [09:18:43] I'll set the milestone to 3.0.0 for now but we'll see how it goes [09:18:46] I still think Sizzle isn't doing anything wrong [09:19:53] so it's okay to have an unescaped formfeed in attribute values in xhtml? i haven't looked at the spec [09:20:01] XML seems to allow it [09:20:08] http://www.w3.org/TR/2000/REC-xml-20001006#NT-AttValue [09:21:02] you sure? [09:21:40] looks like a charref is basically alphanumeric or an entity? [09:22:29] i just love winding my way through these specs [09:22:33] yes, but we're not using a ref [09:22:36] just the literal [09:23:03] ooooh [09:23:18] yeah [09:23:18] \f is a backslash and an f [09:23:22] but that converts to a formfeed [09:23:29] and that is injected as (x)HTML [09:23:37] right [09:23:49] but shouldn't that be a &xxx instead? [09:24:12] well, that's the question [09:24:22] but I don't think it needs such escaping [09:24:54] because it's not a less-than, ampersand, or quote [09:25:49] eventually it comes down to http://www.w3.org/TR/2000/REC-xml-20001006#NT-Char tho doesn't it? [09:26:06] like a tab could be there unescaped but not a FF [09:27:05] the only non-escaped "control chars" allowed are tab, cr, lf by my reading [09:27:42] below that, it says "white space" is space, tab, cr, lf .... but not ff [09:28:19] reading now [09:30:00] the whole mess with https://github.com/jquery/jquery/issues/1967 ... cors [09:30:21] i think there may be some undocumented magic issues there [09:30:22] DaveMethvin: I don't see a direct link between AttValue and Char, but it looks implied, so I think you're right [09:30:45] i followed some link trail gibson042 let me see if i left breadcrumbs [09:31:09] http://www.w3.org/TR/2000/REC-xml-20001006#NT-Reference [09:31:18] http://www.w3.org/TR/2000/REC-xml-20001006#NT-CharRef [09:31:27] http://www.w3.org/TR/2000/REC-xml-20001006#wf-Legalchar [09:31:36] http://www.w3.org/TR/2000/REC-xml-20001006#NT-Char [09:31:51] NT-Reference is the wrong starting point [09:31:58] oh haha [09:32:08] but it still looks like you're right [09:32:12] I can make the change [09:33:11] ok thanks [09:33:34] it probably deserves a Sizzle ticket [09:33:46] yeah [09:33:54] on to https://github.com/jquery/jquery/issues/1967 [09:34:34] in various places we document the support.cors flag to be set manually by users [09:34:37] on one hand I agree we should just kill jQuery.support once we figure out how to do that and still keep tests [09:34:45] OTH, what you said ;) [09:34:58] this was also advised in a couple of tickets [09:35:04] yeah, it's not "official" but that's how you get some unusual xhr impls to work properly [09:35:09] as a way of hacking around our detection [09:35:20] i believe it happens in phonegap [09:35:35] and also on ie xhr when used locally for getting remote files [09:35:42] strange cases like that [09:35:56] basically where we can't know it's legal to do a cross-origin request without help [09:36:20] and the transport finder in $.ajax would get confused otherwise [09:36:41] so i dunno how to get out of this trap very easily [09:37:13] perhaps we just document it [09:37:46] if we document anything out of jQuery.support, we need to document jQuery.support as wekk [09:37:48] * as well [09:38:00] we don't document jquery.fx [09:38:06] not that i'm proud of that one either :) [09:38:09] ;) [09:38:21] it's just a mystery incantation i guess [09:38:35] IMO if we're going to document anything out of jQuery.support we should move everything we can out of it [09:38:36] "If you have a strange setup, set jQuery.support.cors to true [09:38:56] yeah we don't need to dump any more stuff in it [09:39:15] and don't need to maintain stuff there for sure [09:39:18] the only problem is that we need jQuery.support props to test results in supported browsers [09:39:24] so we need a way to get to it in tests [09:39:43] maybe we just should run unit/support.js tests in AMD mode only or sth [09:39:51] oh right [09:40:14] well perhaps we maintain the status quo [09:40:14] none of our internal feature detects should be documented there [09:40:15] that's another option ;) [09:40:30] and we just specify the one exception regarding jquery.support.cors [09:40:38] though I feel telling people that setting jQuery.support.cors is OK is inviting them to touch other support props [09:40:48] they'll learn quickly enough :) [09:40:54] ;) [09:41:01] if we could move it that woudl be great [09:41:05] but it's been around a long time [09:41:10] so we probably can't move it [09:41:20] and we'd just create a new way to get to it that nobody would use [09:41:38] ok, then we should test the result of setting that prop [09:41:58] but it only applies in rare cases, not sure we test any of those [09:42:05] like phonegap [09:42:33] really it seems like we got into this trouble because of the transport selector [09:42:44] if we were just doing xhr we wouldn't need to test cors i think [09:43:00] if you just did a request and there was no cors, it would give you an error [09:43:40] hey timmywil [09:43:47] hi there [09:44:43] just wondering if there is a way to replace the "no transport" with a "screw it, just try xhr anyway" [09:45:03] we can discuss on that ticket i guess [09:45:19] if that can be done, that'd be best [09:45:30] starbucks wifi [09:45:37] we wouldn't have to document private props and external libs wouldn't even need to modify anything [09:46:39] I agree [09:47:02] i didn't wrap up https://github.com/jquery/jquery/issues/1967 from last week [09:47:43] seems like we should just tell him to get as .html() and set as .text() which i think works [09:48:05] that's the support.cors issue [09:48:22] you mean https://github.com/jquery/jquery/issues/1956? [09:48:25] https://github.com/jquery/jquery/issues/1956 [09:48:27] sorry [09:48:55] gotta improve my copy/pasta game [09:49:13] he said: ".text() converts all the tags to html entities, so I can't." [09:49:35] oh both ways [09:51:30] setting as .text() seems to work? [09:52:04] http://jsfiddle.net/ku26kaq3/19/ [09:52:12] it did just now when I tried it [09:52:35] hmm, right [09:52:51] but so does `.innerHTML = value`, which means it's something we're (not) doing [09:53:08] we're trying to convert to a DOM first tho, right? [09:53:30] setting as text avoids that [09:53:44] i'll ask him about setting as text [09:53:50] works for me [09:54:25] so, my basic Node smoke tests have landed [09:54:47] I hope jsdom weird compilation requirements didn't break anyone's build ;) [09:55:04] gibson042 how is that .then() patch coming? [09:55:13] I was just working on it today [09:55:25] i should have time this week to work on several things [09:55:32] should have something to discuss by next meeting [09:55:34] hope to get a lot of tickets and pulls cleaned up [09:55:43] sounds good, i'd like to get that one landed asap [09:55:56] right [09:56:05] any other critical items? [09:56:22] we've gotten a few new contribs lately which is awesome [09:56:33] always nice to have others doing some work :) [09:57:13] okay, you'll probalby see me summoning various people as i go through the tickets, hope I don't get too annoying! [09:57:21] thanks guys [09:57:29] peace [09:57:33] happy new year! [09:57:42] bye guys! happy 2015!