Another fresh example: how can we verify this change with current diff tool which marks the whole paragraph as changed: see https://ru.wikisource.org/w/index.php?title=%D0%9C%D1%91%D1%80%D1%82%D0%B2%D1%8B%D0%B5_%D0%B4%D1%83%D1%88%D0%B8_(%D0%93%D0%BE%D0%B3%D0%BE%D0%BB%D1%8C)/%D0%A2%D0%BE%D0%BC_I/%D0%93%D0%BB%D0%B0%D0%B2%D0%B0_X&curid=30427&diff=2259376&oldid=607212

Thanks. I'll be logged in on #wikimedia-tech as well...

Let's schedule it for "Thursday, March 24 23:59 UTC+1".
I'll be online for several hours, so you'll be able to contact me at ru.WS (https://ru.wikisource.org/wiki/User:Hinote) in case of urgency.

Hi there, so what's the current planned deployment date? We (admins @ ru.ws) need to be prepared if something goes wrong with our templates after this change...

As expected, P2752 demonstrates a quadratic dependency on word count:

Same for the Russian Wikisource. Please fix it ASAP!

In T128697#2089766, @MaxSem wrote:

Yes, examples of pages where this creates problems will be highly helpful.

Nemo_bis, thanks for stats, I've got similar results (of course) for ruwikisource. We have more than ten thousands of pages with long lines...

Btw, it would be very nice to add the wgDiffSomethingLimit setting passed to Diff() or wikidiff2, so that we could request a bigger value. Our high limits may be a problem for busy projects like Wikipedias while we at not so crowded Wikisource could request a bugger one to be set in the project confuguration which would fullfill our needs...
(This probably should be moved into a seperate request)

Sure, we have alot. Here comes the change made by a new user today: see it. I don't know how to patrol it without placing both variants into an external tool in order to view the real diffs made... As you see, the page represents an encyclopedic article comprising of a single paragraph (and the content was uploaded by bot), so we have thousands of similar pages containing long paragraphes. Patrolling of changes in such pages is very annoying in the meantime...

Hi there from the Russian Wikisource. The situation is even much worse for Wikisource projects. Unlike Wikipedian projects, where the users could adjust the long paragraphes in order to make the text more readable, we at Wikisources in most cases are simply not allowed to insert arbitrary empty lines in order to break up long paragraphes simply because these are the paragraphes in their original from the books and SHALL NOT be shorter... Plus, there are some issues in the current implementation of the proofread extension that prevent us from inserting single linefeeds inside a paragraph in order to split a long paragraph into several lines... In many, many cases, very often the result of the current wikidiff is completely unusable for our patrolling users since it marks the whole paragraph as changed...

Hmm, Should I remove the Community-consensus-needed tag by myself in order to make this issue active again?

Please remove the Community-consensus-needed tag if it fits your sence of community consensus,

Our discussion is closed. See results here.

The new wide local discussion (with all users invited via Sitenotice message) was started on 26.01.2016. Planned duration: one month (till 26.02.2016). The discussion is located here.

Okay, no problem. We're going to perform new wide local discussion. Not a problem at all.

We generally require community discussion for changing a specific wiki's site configuration.

Well, you are not quite correct here to my opinion. I just would like to note that, from the formal point of view, the rules on placing such requests for changing configuration (as described in here on meta) include "local on-wiki consensus" as a requirement (and only for non low-activity communities). The community consensus and the community discussion, in general, are a bit different procedures, since local community rules may define different local definitions on how the local consensus is reached, right?

In T124615#1960696, @Krenair wrote:

There would be a big problem if only sysops were able to participate

In T124615#1960564, @Base wrote:

why was the discussion just amongst the sysops?

Oops, 'kay

In T107664#1756882, @waldyrious wrote:

Yes, the version without the text layer was uploaded over it to allow the page images to be generated.

"No thumbnail" -- hehe, we do not really need that fu..n thumbnail ;-) The problem for us at Wikisource is that the pagelist tag produces error in the Index page for such file, the Proofread pages also do not work properly, so we obtain complete mess at Wikisource for such files... Especially, if we try to upload (for some local reasons, working on the books) a new version of a book, for which we have already got many pages in work. Please set the priority for this issue a bit higher (requesting from the Russian Wikisource).