idea: issues bot that points out live examples whose HTML doesn't validate

[cvrebert]

[Edit: See https://github.com/cvrebert/lmvtfy !]

[zlatanvasovic]

Can you describe idea a bit more? 😄

[cvrebert]

A bot that scans for newly posted comments with JS Fiddle or JS Bin links, extracts the HTML portion of the Fiddle/Bin, runs it thru the HTML5 validator, and posts a follow-up comment listing any validation errors.

We get a decent number of "bug" reports where the problem is actually due to invalid HTML.

[zlatanvasovic]

Ok, I understand now.

[mdo]

That'd be interesting, but we have to be careful with whatever bot we come up with. It was too easy the last time for someone to accidentally comment on a huge set of issues with the wrong comment.

[cvrebert]

Unlike the old haunt script, I definitely wouldn't make the bot a Grunt task or have it in the main Bootstrap repo.

[mdo]

👍

[cvrebert]

Other useful bot: One that points out PRs against the gh-pages branch.

[zlatanvasovic]

Yea. +1

2014/1/8 Chris Rebert notifications@github.com

Other useful bot: One that points out PRs against the gh-pages branch.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/11984#issuecomment-31804477
.

Zlatan Vasović - ZDroid

[cvrebert]

• Also, flag pull requests that edit only dist CSS files and not any of the less source code.

[zlatanvasovic]

And pull requests that edit only dist JS files and not any of the JS source
code.

2014-04-27 10:03 GMT+02:00 Chris Rebert notifications@github.com:

• Also, flag pull requests that edit only dist CSS files and not any
  of the less source code.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/11984#issuecomment-41490654
.

Zlatan Vasović - ZDroid

[hnrch02]

+:100: Maybe try and resurrect fat/haunt?
(What happened to that anyway? 2.1.1 had it and 2.2.0 didn't.)

[hnrch02]

/cc @fat

[cvrebert]

@hnrch02 It got removed in 92e7f16. AFAIK, Haunt stopped being used because (a) it was too trigger-happy and incorrectly auto-closed a number of legitimate issues (b) since the code for it was in the main repo, some folks would run it themselves by mistake or out of curiosity, resulting in some comment spam.

Also, it wasn't a real-time always-on service, so there was the inconvenience of having to run it manually on a periodic basis.