Web ui breaks with CoffeeScript 1.6

[neumino]

CoffeeScript just released 1.6.

It fixes the previous issue we had with the version 1.5, but we get a new problem about super.

The error in the JavaScript console looks like that:

Uncaught TypeError: Cannot read property 'models' of undefined cluster-min.js:1319
AbstractList.AbstractList.reset_element_views cluster-min.js:1319
AbstractList.reset_element_views cluster-min.js:1218
AbstractList.AbstractList.initialize cluster-min.js:1261
AbstractList.initialize cluster-min.js:1233
Backbone.View backbone.js:1162
AbstractList cluster-min.js:1235
IssueList cluster-min.js:11814
Container.Container.initialize cluster-min.js:11793
Backbone.View backbone.js:1162
Container cluster-min.js:11783
Container.Container.initialize cluster-min.js:11418
Container.initialize cluster-min.js:11392
Backbone.View backbone.js:1162
Container cluster-min.js:11394
BackboneCluster.initialize cluster-min.js:15318
Backbone.Router backbone.js:882
BackboneCluster cluster-min.js:15289
collections_ready cluster-min.js:15697
$.ajax.success cluster-min.js:15727
o jquery-1.7.2.min.js:2
p.fireWith jquery-1.7.2.min.js:2
w jquery-1.7.2.min.js:4

For more detail about the problem, see jashkenas/coffeescript#2773
It's fixed on their master branch (at least it seems to be), but it has not been pushed on npm yet.

I'll try to recompile again once I get the new version of coffeescript (for now 1.6.1 doesn't work).

[neumino]

I confirm that using the version of CoffeeScript in their master branch works.
I'll close this issue once it makes it in npm.

[coffeemug]

I'm closing this issue because there is nothing that remains to be done for us (and it makes my issue categorization job difficult). Feel free to reopen if I'm mistaken.

[ghost]

Well, it's strange to close an issue while it's still an issue. For example I was using the existence of this issue in Open state as a signal it is still not safe to use 1.6. Put there a label "not our bug" and keep it open.
The same for closed Ruby bug without actually linking to merged pullrequest/commit that actually fixed it.
But it's only my view; after all it's your process.

[coffeemug]

That makes a lot of sense, sorry (reopening). I sometimes forget that shuffling issues around in github affects people other than the dev team (whom I can just talk to since we're in the same room). GitHub really blurs the line between internal and external dev processes (which is a great thing, IMO, but I find myself still getting used to it and blundering from time to time). Reopening, will let @neumino close when it makes sense.

As for the Ruby bug, do you mean #473? (I'll comment on it in a second, but if you mean a different one, please let me know).

[ghost]

Yes, I meant that one. I am used to see which commit fixed an issue (just appending "Fixes #nnn" in the commit message closes the issue automatically while showing the link there).

coffeemug wrote:

As for the Ruby bug, do you mean #473
#473? (I'll comment on
it in a second, but if you mean a different one, please let me know).

[coffeemug]

Sorry -- I commented on it. We typically try to put internal code review number and git commit number on the issue before closing it, but working around the clock to ship 1.4 made us slip up a little bit. You're absolutely right and we'll be more careful.

[coffeemug]

@neumino -- what's the state of affairs here?

[neumino]

I completely forgot about this issue.

I just checked, I have coffeescript 1.6.2 and it works perfectly fine.

If someone tries to compile with 1.5, 1.6 or 1.6.1, they just need to update CoffeeScript

sudo npm -g update

Closing this issue.

[mglukhovsky]

CS 1.6.2 is out, the past two point releases fixed some regressions, and
the related issue on their repo has been closed. Can we test and close?

On Tue, Apr 2, 2013 at 10:57 PM, coffeemug notifications@github.com wrote:

@neumino https://github.com/neumino -- what's the state of affairs here?

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/408#issuecomment-15819789
.