TU #58 states "…please mention this fact in the report". What fact it refers to? To the fact that one has reported the bug to the extension developers? Then IMO TU #58 should start with "If you also decide…". But I am not a native English speaker.
BTW. Also we do not report bugs on Wikipedia.
I admit I don't understand that sentence either. Maybe just fine to remove it? It was added in https://www.mediawiki.org/w/index.php?diff=3615728 by @קיפודנחש
I believe that "this fact" refers to "If you find that the problem is indeed related to a browser extension". If you know that the browser extension is involved, but you want to report the problem on wiki, then please tell the other editors/volunteers that this browser extension is involved.
Ah thanks, that makes sense! I've made an edit in https://www.mediawiki.org/w/index.php?title=Help%3ALocating_broken_scripts&type=revision&diff=5235056&oldid=5234865
Hello, I've just copied a script from Meta (where it was working) to w:fr:MediaWiki:Gadget-Smart_patrol.js and now it's only working with "?debug=true". Otherwise it says " ReferenceError: markAllAsPatrolled is not defined", but I don't understand why because it is defined in the script and I've purge the page and tested from another navigator and account (with no other gadget). JackPotte (talk) 21:36, 24 August 2018 (UTC)
@AKlapper (WMF), my first step for diagnosing is asking user to test while logged out, or in incognito window to reduce the hassle. (How) is that different from the safe mode thing? If it is, it is worth adding somewhere?
@Elitre (WMF), if I understand https://phabricator.wikimedia.org/T152169 correctly, safemode also bypasses your browser cache. Might be a question for the developers involved in that task?
safemode doesn't run any of personal user scripts, gadgets, or sitewide JavaScript in the MediaWiki namespace, e.g. MediaWiki:Common.js. Logging out will still run default gadgets and sitewide JavaScript in the MediaWiki namespace. I think this difference should be mentioned. If safemode stops a problem but logging out does not then it may be a waste to test your personal JavaScript and opt-in gadgets.
Another thing worth mentioning: If you preview a personal JavaScript page, e.g. your common.js, then the previewed version runs instead of the saved version. If the issue can be seen on the previewed page then you can make tests without saving all the time.
Those screenshots are not up-to-date, the developer tools are different in Firefox 58.
I have time, but then the text of this page should be changed too.
For me Firefox console looks like that: https://snag.gy/SpFGsA.jpg
So there is no own section for "net", "css", "js", "security", "logging" and "server", they all seem to be in the same.
Well then if you have time to update anything, please help as well! :) I'll mark your edits for translation when they will be done. Thank you for your help!
I don't remember if I already put this in a more relevant place, but https://commons.wikimedia.org/wiki/User_talk:Elitre#Your_VFC_installation_method_is_deprecated is imho a good example of communication about a broken script. It redirects to more info about the issue, but also says what it is that needs to be done /and/ provides a direct way to enable the replacement gadget. (I am not sure if that code is actually a good idea or anything, but it looks effective.)
A user here says that emptying his cache prevents a lot of his problems. So I wonder if we should recommend these easy fixes before anything else?
Yes, that's a good first move. I think we have created this page based on the assumption that users have already purged the cache.
Changed!
If you want to explain how to track broken CSS, please create a page. :)
The same methods apply to debugging CSS (at least using URL options). I think it should be mentioned here.
