Attending: DaveMethvin, markelog, gibson042, m_gol, jaubourg (12:20)

link Note: Dave on vacation Aug 21-25

link Trac

• Volunteers are making progress on new setup
• Let's give it a few weeks
• Github issues is still an option IMO if Trac stays this bad

link $.xhr

• Scott's proposed API

• https://gist.github.com/scottgonzalez/cc85abedbca0ae4e9f47

• Uses chaining Promise for setting args
• Downside: requires Promise always
• will be supported in 1.x branch if possible
  • significantly reduced functionality (no XHR2 or CORS)
• Promise lib?
• $.xhr() returns Promise so we MUST have shim in the bundle
• Insulate via jQuery.Promise = window.Promise || jQuery.PromiseShim

link markelog to look at a super basic thenable we could build using Deferred

• perhaps build Deferred on thenable?
• jauborg returns! Work with markelog on Deferred

link Pull Requests

• Please review and comment

link Tickets

• 1.12/2.2 tickets: http://goo.gl/NsAfgV
• http://bugs.jquery.com/ticket/15205
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: DaveMethvin, markelog, gibson042, m_gol, timmywil

link Trac

• not sure it will ever be fixed…
• Dave to check with Corey if this will be fixed
• could we switch to GH issues?

• volume is low nowadays
• No serious objections
• Need to port over open tickets
• Dave to talk to Scott

link $.xhr

• discussed w/ Scott at dev leads mtg
• staying with options object
• will be supported in 1.x branch if possible
• not sure if XHR events are reliable there

• don't want to use polling

link Promise lib?

• $.xhr() returns Promise
• User should build with a Promise lib?
• jQuery.Promise = window.Promise || shim

link jQuery data cleanup: http://bugs.jquery.com/ticket/15205

• Should go away once we attach data to elements

link Pull Requests

• Please review and comment

link Tickets

• 1.12/2.2 tickets: http://goo.gl/NsAfgV
• http://bugs.jquery.com/ticket/15205
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: DaveMethvin, markelog, gibson042, m_gol

link Trac

• status?

• gnarf is coordinating another meeting to move it along

link Attaching data to DOM elements

• Igor Minar is concerned about leaks from attached data
• PR: https://github.com/jquery/jquery/pull/1428
• Ticket: http://bugs.jquery.com/ticket/11570

link $.xhr

• markelog did https://gist.github.com/markelog/5a5aec471886a863a136
• General feedback was that people preferred an options object
• See replies on email thread from mikesherov and scott_gonzalez
• Want to keep our $.xhr very small and simple to start
• Which browsers to support? 1.x if possible
• Can we leverage this: http://fetch.spec.whatwg.org/

link hide and show mess - http://bugs.jquery.com/ticket/15037

• Following up on last week's discussion: Is this feasible really?
• Initially CSS-hidden elements need to work with .show()
• Simple display:block show will break tables without some special case
• Probably several other breakages
• In any case let's avoid making it more complicated, but what can we simplify?
• gibson042 to do a PR for a simplification
• .hide() always results in display: none; .show() results in display: "" unless there was a pre-existing non-none value, in which case it will use that

link Pull Requests

• https://github.com/jquery/sizzle/pull/278

link Tickets

• iOS7 tests failing: https://github.com/jquery/sizzle/issues/277
• 1.12/2.2 tickets: http://goo.gl/NsAfgV
• tickets needing an owner: http://goo.gl/v8YQdp
• Android 2.3 two main failures:

• http://bugs.jquery.com/ticket/15202 (appending checked radios)
• http://bugs.jquery.com/ticket/15203 (exceptions in async scripts)

Attending: DaveMethvin, gibson042, markelog, timmywil

link $.xhr

• markelog did https://gist.github.com/markelog/5a5aec471886a863a136
• please review and comment

link Trac

• gnarf coordinating work to update our setup

link Object event handlers

• http://bugs.jquery.com/ticket/12031
• implementation is likely to change internal data structures
• shouldn't affect anyone using documented API

link hide and show mess

• http://bugs.jquery.com/ticket/15037
• Can we simplify to just display:none for hide and display:block for show?
• ...or even absence of display:none for show?

• probably not because initially-CSS-hidden stuff wouldn't be shown

• Eliminates defaultDisplay mess and saving states

link Tickets

• 1.12/2.2 tickets: http://goo.gl/NsAfgV
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: mikesherov, DaveMethvin, m_gol

link Test fails

• m_gol trying to get all the Browserstack issues resolved
• timeout does not seem to be working
• one job ran 6 days
• we have plugins (neither of them works :():
  • a Jenkins timeout plugin
  • TestSwarm timeout setting: https://github.com/jzaefferer/node-testswarm/issues/14
• disabled Android 4.0 temporarily
• Spawning jobs on Kindle Fire HD that hang
• error messages are not very specific, making diagnosis hard
• mobile devices are failing:
  • Android 4.1 gets stuck on a login screen: https://www.browserstack.com/automate/builds/9eac6d2af8b21a4b4d6581029011d85bbab3f74e/sessions/42504cd22f37b3fa31fc91016adb01aab87273eb#automate_button
  • Android 2.3 gets stuck in the queue, doing nothing and then timeouts: https://www.browserstack.com/automate/builds/9eac6d2af8b21a4b4d6581029011d85bbab3f74e/sessions/41cf9b3bc366f0f5e1a7dccc14ebda8c1c3b3ba7#automate_button
• iOS 7.0 fails for mysterious reasons
• BrowserStack emailed, they’re on it

link Migrate: button.attr("value", val) failing, jQuery 1.8 or older

• http://swarm.jquery.org/result/1914335
• I think the test may need to be removed
• Behavior is correct for IE<9 and jQuery<1.9
• Resolution: Move existing test to the 1.9+ branch in the unit tests
  • gibson042 has a pull

link Support matrix

• iOS 6.1+ since Browserstack doesn't provide older versions
• Android 2.3, 4.0+ (not just 4.0+)
• we have a couple of manipulation failures on 2.3 but no one reported them so let’s claim support

link Dave will be on vacation July 17-26

link $.xhr strawman

• https://gist.github.com/dmethvin/43ffd1c743554e5c50ae
• Updated last week with feedback from comments
• Return value of $.xhr?
  • thenable requires user to translate, and also means we always have a dependency on some sort of thenable shim (e.g., $.Deferred)
  • Promise means we need either a shim or native Promise
  • maybe https://github.com/getify/native-promise-only
• Oleg on a plane right now, will have some thoughts/comments tomorrow
• dave to ping domenic for input on how to add .xhr to Promise instance
• dave to ping snover on collab with Dojo
• mikesherov to ping getify on using native-promise-only for our shim

link Download builder in 1.12/2.2?

• Owner?

link Trac sucks

• Nacin volunteered to help, connected him with gnarf
• Got a second person (Christopher Schramm) interested as well

link Pulls

• markelog landed several things this weekend

link Tickets

• 1.12/2.2 tickets: http://goo.gl/NsAfgV
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: markelog, m_gol, DaveMethvin, mikesherov

link $.xhr strawman

• https://gist.github.com/dmethvin/43ffd1c743554e5c50ae

• Feedback?
• Always have $.xhr return a Promise? Deferred? thenable?
• Others can promisify using the node-like callback
• No consensus here
• markelog to prototype some things for the next meeting

• Continue to enhance $.ajax?

• e.g. https://github.com/jquery/jquery/pull/1541
• (no progress events in Promise so how would we support in $.xhr?)

link Download builder in 1.12/2.2?

• Yes, or No?
• Need to be able to test individual modules
• Use jQuery UI builder as a guide

link Trac sucks

• But moving to GH issues would be really hard
• https://github.com/trustmaster/trac2github might be worth a look
• Need to maintain Trac and move to 1.0
• Dave to put out blog post call for Trac help

link Test fails

• TestSwarm, no browsers are scheduled
  • m_gol restarted
• Migrate: button.attr("value", val) failing, jQuery 1.8 or older
  • http://swarm.jquery.org/result/1914335
  • dave to take a look

link Pulls

• Support responseURL
  • https://github.com/jquery/jquery/pull/1615
  • Why do this unless we can cross-browser?
  • Pull looks nice tho .. maybe get their help on $.xhr?

link Tickets

• 1.12/2.1 tickets: http://goo.gl/NsAfgV
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: gibson042, markelog, m_gol, timmywil, DaveMethvin

link Sizzle on npm; ender support; browserify support

• https://github.com/jquery/sizzle/pull/266

link $.xhr strawman

• https://gist.github.com/dmethvin/43ffd1c743554e5c50ae

• please provide feedback in the gist
• Promise support - thenables only?

link Download builder in 1.12/2.2?

• already a third-party one
• might encourage more subset use and reduce bloat whining
• need to warn people about third-party code needing everything

link Test fails

• Core: Android 2.x and window.onerror not firing
  • https://github.com/jquery/jquery/pull/1573#issuecomment-46168028
  • m_gol to look
• IE8 fails on effects
  • http://swarm.jquery.org/result/1937999
  • markelog to look
• Migrate: button.attr("value", val) failing, jQuery 1.8 or older

• http://swarm.jquery.org/result/1914335
• dave to take a look

link Pulls

• https://github.com/jquery/jquery/pull/1606

• I suggest we don't land for API consistency -- done

link Tickets

• Should .animate() work for non-DOM properties? I think it's supposed to :)
  • http://bugs.jquery.com/ticket/15169
• 1.12/2.1 tickets: http://goo.gl/NsAfgV
• tickets needing an owner: http://goo.gl/v8YQdp

Attending: timmywil

link Test fails

• Core: Android 2.x and window.onerror not firing
  • https://github.com/jquery/jquery/pull/1573#issuecomment-46168028
  • m_gol to look
• Migrate: button.attr("value", val) failing, jQuery 1.8 or older
  • http://swarm.jquery.org/result/1914335
  • dave to take a look

link Pulls

• https://github.com/jquery/sizzle/pull/267
  • merged
• https://github.com/jquery/jquery/pull/1601
  • closed
• https://github.com/jquery/jquery/pull/1593 (external)
  • merged
• https://github.com/jquery/sizzle/pull/270 (external for sizzle)
  • merged

link Tickets

• 1.12/2.1 tickets: http://goo.gl/NsAfgV
  • tickets needing an owner: http://goo.gl/v8YQdp

Attending: DaveMethvin, m_gol, gibson042, markelog

link Test fails

• Core: Android 2.x and window.onerror not firing
  • https://github.com/jquery/jquery/pull/1573#issuecomment-46168028
  • m_gol to look
• Migrate: button.attr("value", val) failing, jQuery 1.8 or older
  • http://swarm.jquery.org/result/1914335
  • dave to take a look

link Pulls

• https://github.com/jquery/sizzle/pull/267
  • good to go
• https://github.com/jquery/jquery/pull/1601
  • better way to do those dependencies?
  • volunteers welcome for a better way

link Tickets

• http://bugs.jquery.com/ticket/15148
  • plugin

Attending: DaveMethvin, markelog, m_gol, gibson042, scott_gonzalez, timmywil

link 1.12/2.2 - Let's get going

link Pulls

• QUnit eliminating globals

• https://github.com/jquery/jquery/pull/1586
• Cannot create globals for everything on our side since e.g. assert is passed as a function param
• let’s alias globals where we can & change only what remains

• Increase the set of tested browsers in Sizzle

• https://github.com/jquery/sizzle/pull/263
• FF 3.6 seems to work
• prob not worth removing old browsers, so let's leave it
  • Still need to support all selectors for custom selectors and sane scoping behavior

link Tickets

• http://bugs.jquery.com/ticket/15131

• Decode by type regardless of status code?
• Compat impacts? Nobody should be depending on errors not decoding at this point we could assume?
• Let's leave $.ajax as is and push folks towards $.xhr

• http://bugs.jquery.com/ticket/15127

• We aren't testing in this env (or several others like Win8)
• Ways to do this?
• No obvious solutions here that are easy

• http://bugs.jquery.com/ticket/10406 http://bugs.jquery.com/ticket/13132

• :visible, :hidden behavior change
• empty elements (with 0 height) will be hidden instead of visible
• Although, they were only visible in some browsers
• Will now be consistent for empty elements and