jQuery 1.9 RC1 and Migrate RC1 Released

Posted on by

Drop everything and start testing, jQuery 1.9 Release Candidate 1 is now available! Today we’re also releasing RC1 of the jQuery Migrate plugin, to help in migrating and upgrading your older jQuery code and plugins. We really need your help to make sure this code is free of bugs and ready to release.

The jQuery 1.9RC1 files are available on jQuery’s CDN. We recommend that you start by including the jQuery Migrate plugin. Be sure to check the browser’s console for warnings, which all start with JQMIGRATE and are described in the plugin’s documentation. Just replace your current jQuery include with this:

<script src="http://code.jquery.com/jquery-1.9.0rc1.js"></script>
<script src="http://code.jquery.com/jquery-migrate-1.0.0rc1.js"></script>

For diagnostic purposes, you can also include the jQuery Migrate plugin with versions of jQuery all the way back to 1.6.4 to see what changes may cause issues with your code when you do upgrade. At minimum, please test your site with your existing jQuery version and the jQuery Migrate plugin to see what it reports. Our goal is to make the jQuery Migrate plugin into a tool that can smooth out any bumps in your upgrade road.

If you haven’t looked at the previous 1.9 beta, check out the jQuery 1.9 Upgrade Guide for information about what has changed in this go-round. There’s been quite a bit of API cleanup in 1.9, and the jQuery Migrate plugin is your friend when it comes to finding issues.

In the last beta we introduced a powerful CSS array getter that simplifies and optimizes getting multiple properties. But then we looked, and oh, I dunno, it seemed like this release just needed to “pop” more. So we added more features to make your development life easier.

New! Sizzle Selectors

Sizzle now supports the following additional CSS3 selectors: :nth-last-child, :nth-of-type, :nth-last-of-type, :first-of-type, :last-of-type, :only-of-type, :target, :root, and :lang. That means all those selectors are usable in every browser jQuery supports all the way back to IE6, regardless of native browser support for the selector. There are only three remaining selectors unsupported by Sizzle/jQuery (:link, :visited, and :hover) which we won’t implement due to the overhead of events that would be required to track active elements.

New! .finish() Method

jQuery animations have gotten even better, and our API improved in the process. It all started with a suggestion to eliminate the Boolean trap posed by the .stop(Boolean, Boolean) signature. During that discussion it became clear that we didn’t have a way to handle one of the most useful cases: running all queued animations to their final value. So we added one, .finish(). The best way to show its benefits is with an example.

As you can see from the example, most uses of the .stop() API requiring Boolean arguments can be replaced with some combination of .stop() (with no arguments), .clearQueue() and/or .finish() and be less cryptic. We’re not removing the old behavior though, so any existing code should work fine. We’d prefer that you use .finish() when appropriate, since it reads better and avoids confusing Boolean parameters. And, when you need to go to the end of a whole series of animations, .finish() will do that in 1.9 when nothing would before.

New! Source Maps

jQuery 1.9RC1 has added the ability to use source maps in browsers that support them. At the moment that’s just Google Chrome, but Mozilla Firefox is planning support as well.

What’s all this about? Well imagine that you are using compressed versions of your files on your production site, including a compressed version of jQuery. You get a report that an important customer is running into a problem. You could debug it a lot easier if you had the uncompressed source, but using that on your high-traffic production site isn’t an option. With source maps, you can let the browser’s debugger “map” the lines in the compressed file into the uncompressed source. That makes it so much easier to set breakpoints, inspect or change values, and see meaningful variable names. This video gives you a taste of what it can do.

The jQuery CDN has the map files for this beta version; the jQuery, Google, and Microsoft CDN versions of jQuery final releases will also get matching source maps from now on. If you are using jQuery 1.9 or 2.0 from the CDN, just check the “Use source maps” option in the debugger and you will be set. Open the uncompressed version of jQuery in the Chrome debugger and you can set breakpoints or inspect variable values there, even though the compressed version is being used!

When hosting your own copy of jQuery, copy the map to your server and use it from there. For simplicity we assume that the compressed and uncompressed copy of jQuery is in the same folder as the map file; this is the case for CDN versions and you should follow the same rule if you make local copies. The map file name is the same as the compressed version, with .map replacing .js. Do not rename the files when you copy them. So, if you were to use a local copy of jquery-1.9.0.min.js the corresponding map file would be jquery-1.9.0.min.map and the uncompressed file would be jquery-1.9.0.js in the same folder.

Congrats, Contributors!

Many thanks to the people who have contributed to the 1.9 release: Akintayo Akinwunmi, Alexander Farkas, Allen J Schmidt Jr, Ben Truyman, Bennett Sorbo, Callum Macrae, Carl Danley, Corey Frang, Daniel Gálvez, Dan Morgan, David Bonner, David Fox, Devin Cooper, Elijah Manor, Erick Ruiz de Chavez, Greg Lavallee, Ismail Khair, James Huston, Jay Merrifield, Jonathan Sampson, Julian Aubourg, Marcel Greter, Matt Farmer, Matthias Jäggli, Mike Petrovich, Mike Sherov, Oleg Gaidarenko, Paul Ramos, Richard Gibson, Rick Waldron, Rod Vagg, Roland Eckl, Sai Wong, Scott González, Sebi Burkhard, Timmy Willison, Timo Tijhof, Tom Fuertes, Toyama Nao, and Yi Ming He.

jQuery 1.9.0RC1 Changelog

Ajax

  • #12004: Rename ajax.type to ajax.method
  • #12550: jQuery Ajax cache=false doesn't always work

Attributes

  • #9905: .removeAttr( "id" ) sometimes crashes IE 7
  • #10299: hrefNormalized === false also needs a propHook
  • #12048: [IE6/7/8] xml set attribute
  • #12584: jQuery wrongly serializes &lt;select&gt; with one disabled &lt;option&gt;
  • #12600: jQuery('select').is('[value="value"]') works inconsistently depending on number of elements returned
  • #12945: attr throws exception in IE9 on Flash &lt;object&gt;s
  • #13011: Setting type attribute on an input doesn't work as intended

Build

  • #12254: Reflected XSS
  • #12490: Move submodule update process to grunt
  • #12725: Avoid localized UTF-8 characters in intro.js @DATE
  • #12741: inconsistent line endings in official jquery-1.8.2.js download
  • #12886: Add source map support for build
  • #13044: Execute all QUnit modules in TestSwarm
  • #13064: Improve test suite fixture cleanup

Core

  • #9469: Remove semi-functional .selector calculation from .pushStack()
  • #9904: Move deprecated functionality to compatibility plugin
  • #10417: jQuery.later
  • #11290: selector interpreted as HTML
  • #11737: Remove jQuery.sub
  • #12107: Change proxy to allow arguments currying without overwriting context
  • #12134: implement HTML5 compilant form data construction into $.fn.serialzeArray
  • #12191: jQuery.type() should return "error" for native ECMAScript Error objects
  • #12519: Public API methods should not have private arguments
  • #12840: Remove (private) parameter "pass" from jQuery.attr and jQuery.access
  • #13021: each() cannot work well with a literal object who has a length member
  • #13075: Performance optimization for $.type
  • #13076: Performance optimization (10-30%) for $("some-selector")

Css

  • #11938: jQuery.css should accept an array to get multiple properties
  • #12990: 'px' automatically added to column-count css property
  • #13088: under IE8, $(selector).attr('style') always return lowercase string

Data

  • #10544: Remove ALL special meanings of "." in keys for $.fn.data

Deferred

  • #11405: deferred.notify() invokes progressCallbacks with deferred as context
  • #13160: Deferred.then doesn't propagete custom context

Effects

  • #12803: Smarter hook point for jQuery.timer

Event

  • #3827: Checkbox state inconsistent in click event handler
  • #12061: $(window).beforeunload() clobbers previous handler and return values
  • #12518: Don't use offsetWidth in jQuery.event.trigger()
  • #12610: jQuery.event.dispatch should remove window.event
  • #12736: Move pseudo:hover to jquery-compat / deprecated.js
  • #12739: Name: Passing in an Event to trigger strips namespace
  • #12827: Remove "exclusive" events
  • #12828: Remove event properties: attrChange attrName relatedNode srcElement

Manipulation

  • #4087: insertAfter, insertBefore, etc do not work when destination is original element
  • #9646: IE7: Cloning of form-elements and changing their names also changes the name of the elements that are cloned.
  • #10470: wrap() evaluates scripts
  • #11226: .after and .before returns incorrect data for $('not_existing_element')
  • #11230: .appendTo .prependTo .insertAfter .insertBefore returns incorrect data for $('not_existing_element')
  • #11280: appending elements to an object element fails in IE &lt; 9
  • #11795: Resolve script manipulation/execution inconsistencies
  • #12120: Inconsistency of .end() with respect to .after()
  • #12336: Calling $('#select').empty() should set options length to 0
  • #12392: Elements created from HTML strings have a parentNode
  • #12449: replaceWith() doesn't clone elements where required
  • #12503: before/after will choke if collection has not first disconnected node
  • #12777: Applets don't work when appended on IE
  • #12863: behavior:url(#default#savehistory) causes event error on oldIE
  • #12957: Improve wrapMap
  • #13019: New pre-1.9 .replaceWith() behavior leaks data and events
  • #13094: If to jQuery#before passed function argument it should receive index of the current element in the set

Misc

  • #12758: Make sure Summit new authors are credited

Offset

  • #6446: Mobile Safari 4.0.4: $.offset.top() reports wrong position after scroll

Selector

  • #11115: ".is()" and ".filter()" disagree on attribute selector "[checked]"
  • #12856: Syntax error, unrecognized expression in jquery 1.8+
  • #13070: filter()/is() does not work correctly with attribute equals selector which contains special characters

Support

  • #12569: Improve feature detect for oldIE event bubbling
  • #12869: Support tests affect page layout in IE8/9/10 running in IE7 mode

Traversing

  • #12009: jQueryObject.find(element) corrupts the stack
  • #12816: .find can return elements in the wrong order

16 thoughts on “jQuery 1.9 RC1 and Migrate RC1 Released

  1. Sorry, I should have changed the description on that ticket. We didn’t need jQuery.later() because window.setTimeout() without a time is basically the same. That’s what the patch associated with the ticket says and does.

  2. jquser on said:

    Is it possible for the migrate plugin to show *where* your code is using invalid features? Right now, when the console generates the JQMIGRATE entries you see *what* you called that’s invalid, but not *where*.

    If you are running with several dependencies this would be a large issue to track them down… any thoughts or am I missing something?

  3. @jquser, great question! Use the debugger on any recent browser and set a breakpoint at the top of the migrateWarn() function in jquery-migrate.js. Then look at the stack trace. Here is how it looks in Chrome (http://i.imgur.com/ahn5x.png) and IE9 (http://i.imgur.com/pjAZX.png), just double-click the entries in the stack trace to have it take you to those locations. From there you can probably figure out which code is causing the issue.

  4. jquser on said:

    @dmethvin, thanks, I’ll give it a shot!

    Just out of curiosity, are you able to access the stacktrace within the plugin and print it out, or does this have to be a manual step?

  5. Unfortunately, stack traces aren’t standard across browsers and would also take a LOT more code to decode inside the plugin. Since you have to be in the debugger anyway to see the warnings you might as well set the breakpoint yourself and view it interactively.

  6. @dmethvin
    Well, you could always use console.error, though it’s weird to use it for warnings. But maybe under some flag, jQuery migrate switch etc.? console.log & console.warn don’t provide a stack but console.error does, at least in Chrome Dev Tools & Firebug – with an easy, clickable interface. I suppose there could be a problem with IE but their webdev tools are just terrible anyway.

  7. @m_gol, We could add some optional flags to use console.error instead of console.warn, but to me it’s much easier to avoid editing the code and use the debugger.

  8. @dmethvin
    Well, it’s not that big an issue after all. If sb really prefers to see stacks in messages, the Migrate plugin source code is available. :) And it’s simple find & replace.

  9. Mplungjan on said:

    Where would I request that toggle-event would be renamed to toggler rather than deleted?