proposed upgrades to jQuery, jQueryUI, and QUnit

[Colin Clark]

Hi Justin,

I agree that, unfortunately, we'll want to continue to support IE8 for a little while longer. It's the latest version of IE that is supported on XP, which is still quite widely used in some environment. From all the browser marketshare stats I've seen, IE8 is trending downward significantly, so I think it's likely we'll be able to drop it for the Infusion 2.0 release.

If it helps any, I have been using jQuery 2.0 with a framework-only build of Infusion for several months now in Flocking and it is working great. I think this is a fine choice for our implementers who don't need to worry about IE8.

Colin

---
Colin Clark
http://fluidproject.org

On 2013-09-13, at 8:34 AM, Justin Obara <obara.justin at gmail.com> wrote:

> We are in the midst of planning an upcoming Infusion 1.5 release. One of the major tasks will be to upgrade our versions of jQuery, jQueryUI and QUnit.
> 
> For jQueryUI and QUnit, we should just take the latest available at the time we do the upgrade. Currently that is jQueryUI 1.10.3 and QUnit 1.12.0. jQuery, however, is now supporting two versions. A 1.x line and a 2.x line. While both have the same API and general feature set, the 2.x line removes IE support below IE9. This blog post has more information about the different versions of jQuery, http://blog.jquery.com/2013/01/15/jquery-1-9-final-jquery-2-0-beta-migrate-final-released/ . For the time being we'll probably still want to support IE8 in Infusion. This means we should pick up the latest 1.x release, currently 1.10.2. 
> 
> The major benefit to taking the 2.x line from jQuery will be that it is smaller and faster. For our users who do not need to support older versions of IE this could be a real benefit. Fortunately our build system already supports us excluding dependencies, meaning that an integrator could create a package of infusion without jQuery and slot in their own copy from the 2.x line. Since the API's are supposed to be the same, this should work. Another alternative would be to allow integrators to build an Infusion package with either version of jQuery. Although this would require changes to our current build system and possibly carrying around two copies of jQuery in our repo.