Now that we can't use: If it ain't broken don't fix it.
That's precisely the answer, and I don't see why we shouldn't be able to use it. The
event.layerX deprecation warning scare crow has been running around the internet for a while now, but so far, these warnings haven't appeared yet. Note that the question you link to is about the canary build; if we started considering that a supported browser, we'd have no time to answer Meta questions anymore, being knee-deep in Chrome bugs 24/7.
That said, the reasons we upgrade the jQuery version are usually one of these:
- Plain ol' bug-fixes. Something is broken in the jQuery version we use, and fixed in a later one. "Broken" is defined as "actually impacts us". An issue that causes IE6 to puke when appending a flash object to the DOM if your Windows ME machine has less then 128MB of RAM does not make jQuery broken for us.
- Awesome features. The reason to upgrade to jQuery 1.5 were Deferred objects. Pretty much the only reason. They rock, are useful all over the place, etc. Did I mention that I love them? (And yes, I'm looking forward to
.always(), which was only added in 1.6, but that alone is not a reason to upgrade.)
- Somewhere in between those two: jQuery fixes an issue we're having (work around a browser bug etc.). Now one of these usually isn't enough to make us upgrade, but they can stack.
Well, if we suddenly find ourselves in 2028 and still running jQuery 1.5, that may be a hint as well…
Anyway, what we definitely don't do is upgrading for the sake of upgrading. An upgrade always comes with issues; often very subtle ones.
Example: Until a few months ago, chat was still running jQuery 1.4.2 (I know, shudder, right?); however, when we made some changes to the list of "people in this room", something broke with the animations. This was a bug in jQuery 1.4 causing animation of the
width of a
display:inline-block element to fail. That bug was fixed in 1.5, so I upgraded chat to use the same version as the main site.
It took a long time until someone noticed that this silently broke the "replied-to message" highlighting in the transcript – because between these versions,
$().live("hover", ...) was changed from raising
mouseover events to
Things like that are the reason why we don't upgrade just so. In particular, we very much avoid upgrading to 1.x.0 versions; experience has shown that rule number 1 (above) will make us upgrade to 1.x.1 very soon anyway.
Now, when/if the deprecation warnings start appearing all over the place, that will probably be a good reason to upgrade. It's not like we don't like using the shiny new stuff; it's just that such an upgrade always comes with maintenance costs.
Update: Meta is now running on jQuery 1.7.1, we'll be monitoring it for issues for a while and then roll it out to the rest of the sites.