Wikinews:Water cooler/proposals/archives/2016/March

From Wikinews, the free news source you can write!
Jump to navigation Jump to search

dialog-based submit button

{{flag}} I'd appreciate feedback from others concerning a replacement I now have working for the current dialog button, using my dialog tools.

  • Demo page: Test 34 — the button at the very top is the new one; there's also a conventional submit button on a {{develop}} template just below.
  • Kinds of feedback I'm hoping for:
  • Does it work on various platforms; I only know from personal experience that it works in Firefox under Linux. Especially, I haven't tested it (in recent times) under Windoze, and I haven't tested it on any mobile device.
  • Does it work acceptably to replace the button now on the {{develop}} template? More broadly, does anyone have any objections to replacing the current button with the new one on {{develop}}?
  • Advantages of the new button:
  • The new button will only allow an IP to submit for review if the article was created by an IP. We've talked about that as a likely upgrade over, iirc, the past several years.
  • The old button is driven by extra code added to our common.js; with the new button, we can remove that clutter from common.js.
  • Disadvantages of the new button:
  • It's slower. That's unsurprising when replacing a one-off customized button with a button implemented through a generic tool.
  • If we end up with all our automation based on the dialog tools, and then they break, we're screwed. In fairness, though, right now afaik all the major stuff depends on User:Bawolff/mwapilib2.js.

--Pi zero (talk) 22:44, 16 January 2016 (UTC)

Okay, Let me test it. But can you please explain The new button will only allow an IP to submit for review if the article was created by an IP. We've talked about that as a likely upgrade over, iirc, the past several years. (talk) 07:11, 17 January 2016 (UTC)
  • When I clicked Submit for review, I got a message. Error: (edit request declined by Template:Dialog/submit/form): sorry the article was created by a registered user, so this button will only work for a registered user. And that, I guess will not be good.
    NASA announces water on Mars. This article was started by Sethmtraut (t · c · b) and it did not pass while reviewed by Pi zero. Then, Mindhunter77 (t · c · b) made changes and submitted again. But the style was not good. tried to correct some typos, links, credits... It did not pass for the second time as well. (t · c · b) tried to distance from the article and submitted it. It did not pass for the third time. .195 re wrote the article and then with some copyedits by and later by the reviewer, it passed.
    That is not a good idea. Else, the article could not have been submitted for review. (talk) 07:56, 17 January 2016 (UTC)
I'm willing to disable the IP/creation condition in the button; the idea has been discussed on various occasions, but I don't recall it was ever really more than speculation. The declining-to-submit message produced in that situation could also be changed.

I should point out, though, that you're an experienced Wikinewsie who ought to be using a registered account. It's not really right to be deliberately using an IP instead of a registered account; Wikinews runs on accumulated reputation, and this is undermining the basic means of authenticating identity that the Wikinews community relies on. I've got pretty mixed feelings about having agreed to this "voluntary block" thing; I'm not at all sure I'd agree to do it again. Except for this one peculiar situation with an experienced Wikinewsie systematically not using their account, I don't recall a single case in my years on Wikinews where an article submitted by a registered user was legitimately submitted by an IP. True enough we don't want to come across as unfriendly to newcomers and therefore, perhaps, oughtn't prevent them from using the button, but I don't think this particular case really says anything much about it one way or another. --Pi zero (talk) 14:48, 18 January 2016 (UTC)

  • Working fine for Mozilla Firefox on Windows. For Chrome, it did. Safari for Windows, yes. But do note that I hat trouble for submitting it from the button not in ambox in Chrome, and did not work in Safari. Same problem with Opera. I don't have Netscape or konqueror. Did not try with Internet Explorer.
    AGastya (talk) 08:14, 17 January 2016 (UTC)
When you say you "had trouble" in Chrome, what form did that trouble take? What exactly does "not work" mean in Safari, in Opera? --Pi zero (talk) 14:04, 18 January 2016 (UTC)
Did the button appear correctly on the page? (I'm obviously going to have to get myself an installation of one or more of these to tinker with, and there are a slew of questions that might be worth asking depending on accumulated evidence, but whether the button appears correctly must be the first question, because the gadget is written so that the button won't even display until the gadget runs.) --Pi zero (talk) 15:18, 18 January 2016 (UTC)
@Pi zero: (proxy settings changed, else IP .195)In Chrome, nothing happened in the first try. In the second try, it was submitted for review, but consumed more time than in Mozilla Firefox. (There are editors who still use Chrome over Firefox, though not me) Yes, the buttons appeared correctly. Does not work in Safari and Opera meant that the button did not complete any task. (talk) 15:34, 18 January 2016 (UTC)
Do the scroll up/down buttons on categories, such as at Category:Science and technology, work in those browsers? --Pi zero (talk) 19:19, 18 January 2016 (UTC)

(Safari for Windows): up/down buttons are working, but I was unable to submit test 34 still.
AGastya (talk) 14:20, 20 January 2016 (UTC)
And same goes for Chrome. A point to add, this time, nothing triggered when I tried to submit in Chrome. For once, the hourglass cursor came, but it did not work.
AGastya (talk) 14:24, 20 January 2016 (UTC)
Add Opera as well.
AGastya (talk) 14:26, 20 January 2016 (UTC)

It did nothing for Safari on Mac. But yes, {{develop}} worked for submitting it.
AGastya (talk) 05:49, 30 January 2016 (UTC)
It worked when I was logged in on Chrome browser for Mac. It did no mess like that on Windows, but nothing happened when I used IP.
AGastya (talk) 05:52, 30 January 2016 (UTC)
  • The issue with an IP submitting an article for review, where the article was started by a registered account, relates to drive-by curiosity/stupidity. Someone wholly unfamiliar with the project stumbles on a page with the {{develop}} template on it, and can't resist the temptation to click the button and see what it does.
With an experienced contributor working through an IP address, there's always the option to edit and manually switch {{develop}} to {{review}}. --Brian McNeil / talk 13:08, 30 January 2016 (UTC)
I now have Opera installed on my laptop. Sure enough, the submit button doesn't work. Interestingly, not only do category scroll buttons work for me, but also the null-edit button I've got on a diagnostic test page; so, it's not something that happens for all edit operations, rather something that happens for some of them. (I expect to tackle it in a serious debug session when I'm wide-awake, which I'm not atm.) --Pi zero (talk) 21:59, 3 February 2016 (UTC)
Did some preliminary testing with Opera. All my diagnostic test pages work fine; the only problem I've observed is that the submit button doesn't work. --Pi zero (talk) 02:01, 10 February 2016 (UTC)
  1. Firefox 43.0, running on Linux
    Works for registered user using button
    Fails, with error, as-intended for IP
  2. Chromium 47.0.xxxx running on Linux
    Fails for both tests

I keep Chromium around as it is stripped of much of the Google "Do some evil", and I want a second browser hanging around with Flash support for BBC iPlayer. I mainly use Firefox, but they've not been particularly good at keeping the British English dictionaries available. --Brian McNeil / talk 13:21, 30 January 2016 (UTC)

One bug down; more?

Retests, anyone? I've implemented a bug fix that I know works for my Opera 35.0. The same problem seems likely to occur on some other browsers, and the fix seems likely to work for a variety of browsers too, but that's guesswork. --Pi zero (talk) 21:23, 12 February 2016 (UTC)

  1. Mozilla Firefox (Windows)
  2. Firefox Developer Edition (Windows)
  3. Chrome (Windows)
  4. Safari (Windows)
  5. Opera (Windows)
  6. Opera developer Edition (Windows)
  7. Internet Explorer (Windows)
  8. Safari (Macintosh)
  9. Chrome (Macintosh)

Working perfectly with IP and account. Not tested:

  1. Microsoft Edge
  2. Chromium
  3. Konquorer
  4. Netscape

It works! (talk) 18:55, 14 February 2016 (UTC)

Cewl. --Pi zero (talk) 21:58, 14 February 2016 (UTC)

@Brian McNeil: Hopefully working with Chromium now. --Pi zero (talk) 13:57, 15 February 2016 (UTC)


  • I have modified the button so that, when used by an IP on an article created by a registered user, instead of outright refusing to submit, it points out the oddity and asks the user to confirm that they really want to do that.
  • As far as I currently know, the tools now work on all major browsers.
  • If nobody objects, I'll probably deploy the new button sometime in the moderately near future, and remove that bit of code from our common.js. Our common.js is a mess, and I do feel it needs to be distributed into gadgets — ultimately, the one line that enables the dialog tools may be the only thing that needs to be in common.js — but since I'm pouring all my software-development time into other things, I've no idea when I might be able to devote the unknown quantum of attention needed for gadgetization.

--Pi zero (talk) 14:18, 2 March 2016 (UTC) Tested and worked fine on:

  • Waterfox
  • Firefox
  • Chrome
  • Chromium
  • Opera Developer Edition
  • Opera
  • Firefox Developer Edition
  • Internet Explorer

I will try on Safari (Mac). I don't have Safari for Windows now. All the above browsers were for Windows. But, why is the message being displayed for an IP? Why this confirmation is not for any user, that are they sure to submit it? I could not understand the reason, @Pi zero:. (talk) 16:37, 2 March 2016 (UTC)

Well, here are some thoughts on this design point.
  • If you always ask a certain question in a common situation, it loses most of its meaning; people just answer it automatically, and so it may fail to serve its purpose in some infrequent situation where it does matter more than usual. It's also annoying to be asked a mostly-meaningless question like that; unless the stakes are especially high, it's probably better to streamline the process in most situations and reserve the double-check for special ones.
  • It seems to me, just as an informal impression, that driveby submissions for review are usually an IP submitting somebody else's article. This would check for that, and even (if I've coded it correctly) contains an exception for pages that were created by an IP in the first place. There are even more sophisticated checks one could do, although the one practical limitation is that it would be too expensive to check the whole of a long page history, so we only get to examine either the oldest 50 revisions, or the newest 50 revisions. What do we do with that information? Well, one thing we could do, which it's set up to do now, is check that if the submitter is an IP, then the page creator was an IP too; that requires the 50 oldest revisions. We could also check to see whether the submitter has also made some other edit to the page; one might want the 50 newest revisions for that. Another one would be to raise a warning flag if the article is submitted for re-review without having changed anything since the previous review; that definitely needs the newest revisions rather than the oldest.
  • What do you think would be most useful? Under what circumstances to ask for confirmation, and how to word the question when doing so?
--Pi zero (talk) 16:56, 2 March 2016 (UTC)
Okay, now it it is clear. Fine then. But can we have a Test 35 made by an IP to see how it is working on the other hand?
—The preceding unsigned comment was added by (talkcontribs) 17:09, 2 March 2016‎
You're particularly well placed to create such a Test 35 page yourself, since you're using an IP that's already spread around. Just copy the content of Test 34 to the new page, and give it an edit summary with a link to the other page, such as test page, content copied from [[Test 34]]. --Pi zero (talk) 17:32, 2 March 2016 (UTC)
Well, it did not ask that an IP created this article, do you want to submit for review. It should ...
Agastya Chandrakant ⚽️ 🏆 🎾 🎬 🎤 📰 17:54, 2 March 2016 (UTC)
I meant check-for-confirmation to occur only if the article creator was registered and the submitter is an IP. I wasn't concerned with the reverse situation, in which the creator was an IP and the submitter is registered, because I figured this is either a more experienced user helping out with a less experienced user's article, in which case the more experienced user probably knows what they're doing when they click the submit button; or a newcomer creating the article as an IP, belatedly registering, and submitting through their newly created account. --Pi zero (talk) 18:41, 2 March 2016 (UTC)

@Pi zero:Works fine with Safari on Macintosh. (talk) 12:41, 8 March 2016 (UTC)

Cat people

I have in mind to rename [[:Category:People]] to Category:People by occupation, and move all Wikinewsie categories to a separate category. I've found the usefulness of that part of the category hierarchy has been hampered by the unobviousness-of-purpose of [[:Category:People]], and based on recent archeological research I understand this to be the rationale for the category. I want to straighten this out because (prompted by remarks from Blood Red Sandman) I'm designing a much-upgraded splash page for our archives, and had trouble figuring out which category ought to be the "top" of the people-oriented branch of the hierarchy.

This follows the same logic as the several categories of the form "News articles by X", each of which is an internal category meant to contain all article-categories that identify the article by X. For example, Category:News articles by country contains (or should contain) all categories that identify articles by particular country: Category:Spain, Category:India, and so on. Category:News articles by person should contain all categories that identify articles by particular person: Category:Narendra Modi, Category:Pope Francis, and so on. Similarly, [[:Category:People]] should contain all categories that identify people by occupation: Category:Actors, Category:Royalty, and so on. (Archeological evidence: here.) --Pi zero (talk) 20:08, 30 March 2016 (UTC)

Agreed That makes it seem like some news articles aren't about humans, which is really not true: even weather disasters still focus on how these phenomena effect humans. —Justin (koavf)TCM 03:01, 31 March 2016 (UTC)
Support Seems to be a good idea!
Agastya Chandrakant ⚽️ 🏆 🎾 🎬 🎤 📰 04:59, 31 March 2016 (UTC)

Done --Pi zero (talk)

Modifying the welcome template

I was thinking what if we modify the welcome template by asking the users to let us know about the topics of their interest, and encouraging them to write news articles on that topic. It can make them feel that this project is interactive and thus people would not just create an account but start contributing? Truth to be told, there are many accounts created, but only a fraction of them contribute. Comments?
PS: If approved, let's make it interactive using CSS and jquery.
Agastya Chandrakant ⚽️ 🏆 🎾 🎬 🎤 📰 16:18, 31 March 2016 (UTC)

I've imagined we might want to do something with dialog, down the road. There's also been the possibility of switching back to the old version of the welcome template, which we changed a few years ago hoping to get through to people what they're supposed to be doing. The trouble is, nobody reads the welcome messages; I actually had someone on Wikibooks once complain aggressively that we'd been so inconsiderate as to put a welcome message on their user talk page — they said it was really, really annoying, they'd contributed to a bunch of different wikimedia projects and each one had insisted on putting a welcome message on their user talk page. One gathers that they had in fact never read any of these messages, since if they had read them they'd have been aware that every project is different and goes to a lot of trouble to design a welcome message that explains what's different about that project.

Something to keep in mind about en.wn's Howdy template: It's not subst'd. There are (last time I checked) more than a million pages on en.wn that transclude that template, so if we modify it (even if the modification is just fixing a spelling error), it puts a much larger load on the servers than modifying some template that's (for example) merely used on all of our 20,000 published articles.

If we were changing the welcome template I'd also rather like to set it up so that the different between the one for registered users and the one for IPs is a parameter somewhere; right now they're two separate templates, and iirc they're not even in sync. --Pi zero (talk) 17:23, 31 March 2016 (UTC)