Уикиновини:Разговори/Архив 2019

Content review #3: Wikipedia as a "source"?

I see that the article "Facebook премахна 500 страници, разпространяващи фалшиви новини (17 януари 2019 г.)" is using "w:en:Internet Research Agency" as a "source". I mean, is this acceptable? If so, why? Most wikis have never used other wikis as a source. --George Ho (беседа) 11:52, 21 януари 2019 (UTC)

I confess that this question is strange to me. Why Wikipedia shouldn't be an acceptable source? I see absolutely no reason for this.

Also, I see that in the different-language Wikinews it is an ordinary and accepted practice to translate an article from another-language Wikinews. This appears to me to be a bigger reliance to another wiki than simply referring to it as a source. So, the notion that other wikis - especially WMF projects - should not be used as sources appears to me not very logical. -- Григор Гачев (беседа) 14:25, 21 януари 2019 (UTC)

So, WP:NOTSOURCE must be wrong then. As for the translated articles, they are supposed to come with reliable sources from the original. If they don't, KABOOM! Gosh, can't believe I need to explain things like these...
— Luchesar • Б/П 19:00, 21 януари 2019 (UTC)

WP:NOTSOURCE is a Wikipedia rule that governs the work of Wikipedia. Here is another, different project, named Wikinews. It doesn't contain a rule banning the usage of information from Wikipedia, and therefore does not have to follow such a rule.

In addition, like with Wikinews, the sources for a Wikipedia article also have to be reliable. Due to this, I don't see any reason to believe that its information is less reliable than an information in another-language Wikinews. If a Wikipedia article is used as a source in a Wikinews article, everyone can go to it and decide if its sources are good enough, exactly like they could do it with the other sources of the Wikinews article. -- Григор Гачев (беседа) 19:22, 22 януари 2019 (UTC)

“ If a Wikipedia article is used as a source in a Wikinews article, everyone can go to it and decide if its sources are good enough ”

Which would be just the same if a Wikipedia article is used as a source in another Wikipedia article. Apparently, you haven't bothered to read even the very first sentences of WP:NOTSOURCE.
— Luchesar • Б/П 08:11, 23 януари 2019 (UTC)

It is a fine rule in Wikipedia, but this is not Wikipedia. Therefore, it is up to the editor (eg. me in this case) to decide whether to consider it appropriate and follow it, or not.

In my opinion, this rule makes a lot of sense in some other WMF projects, but far less in Wikinews, and none in this particular case. For this reason, I have chosen to not follow it in the case we are discussing now.

I believe that there is a profound difference between a Wikinews article using as a source a Wikipedia article, or a Wikipedia article doing it. I admit that Wikipedia is unreliable to some small degree, but so are all possible sources, so this reason to not use it as a source does not appear all-important to me. Practically all other reasons to not self-source in Wikipedia - possibility for circular sourcing, etc - also appear to be either impossible or unrealistic when Wikinews uses Wikipedia as a source. Finally, Wikinews also can be edited by anyone like Wikipedia, so the comparison between them is very similar to that between a paper-printed encyclopaedia and a traditional media, eg. a newspaper. A newspaper will be only diligent to use an encyclopaedia as a source - so, for Wikinews it should be only diligent to use Wikipedia as a source.

Strictly personally, could it be that our arguments eventually stem from having two different priorities, quality vs. quantity? Neither is any good without the other, but still, people tend to emphasize on one or the other. (Like with so many other subjects - eg. the Wikipedia inclusionists and deletionists.) If so, it looks like neither of us will be happy with the results and compromises we will eventually achieve, but as a whole they will make for a better project than any of us would nourish separately. -- Григор Гачев (беседа) 18:58, 23 януари 2019 (UTC)

No, the root problem is that disseminating news and information in general requires responsibility.
— Luchesar • Б/П 12:19, 24 януари 2019 (UTC)

I couldn't agree more to that. However, I think that it should not climb to drastic levels. Otherwise, things can reach a degree where it is very hard to do (or permit the others to do) anything, out of fear that if it is not good enough, releasing it would be irresponsible. Even the best medicine can start harming, if the dose is exceeded. And it is with the best medicines where it is the easiest to overexceed the dose, exactly because they are so good... Just my take on it. -- Григор Гачев (беседа) 00:38, 25 януари 2019 (UTC)

You still miss—or are trying to evade—the point: it isn't about news being “not good enough”. It is about news that are good enough, but misleading. There is no “right dose” for them.
— Luchesar • Б/П 06:19, 25 януари 2019 (UTC)

That might depend on the definition for "misleading": I am not sure about yours. Could you please provide it, for my reference? -- Григор Гачев (беседа) 11:43, 26 януари 2019 (UTC)

We've been talking about fake news from the very beginning and you still want me to explain that for you.
— Luchesar • Б/П 19:54, 29 януари 2019 (UTC)

So far I saw you having problems with a lot of things here, and none of these was fake news - at least, not in the established definition of that term. So I am coming to the conclusion that "misleading" means to you other, or more things. That is why I am asking about your criteria for "misleading". (Or for "fake news", if you prefer the term.) Having those spelled out, we can achieve understanding much more easily. -- Григор Гачев (беседа) 22:03, 31 януари 2019 (UTC)

Grigor, the problem with what you've written is that it makes no sense given the long discussion on these very topics in m:Proposals for closing projects/Deletion of Bulgarian Wikinews.
— Luchesar • Б/П 09:36, 2 февруари 2019 (UTC)

There is no doubt that I am not perfect. That is why I try to understand your position. The discussion about the deletion of the Bulgarian Wikinews doesn't give me the answers to some important questions. Your criteria about what is misleading news, or fake news, might give me these answers, and allow me to understand your position. (As I have noted before, I am not very far from supporting that deletion myself, after seeing how few of the loudly complaining journalists actually use it as an opportunity to publish what they say is censored at their media outlets.) Understanding your position might or might not change mine, I have no way to tell the end result beforehand, but even if it doesn't, it will make me understand you better - and we all need to understand better one another in such arguments. -- Григор Гачев (беседа) 18:16, 2 февруари 2019 (UTC)

My best advice for you, Grigor, is using sources from Wikipedia to replace Wikipedia itself as the source. Have you thoroughly inspected the sources from the Wikipedia article yet? --George Ho (беседа) 08:38, 30 януари 2019 (UTC)

Sometimes this can indeed be a very good idea. In this particular case however, I decided that the article itself is more valuable than the list of the sources it uses (and it also lists them). Due to this, I decided that it would be better to point to the article as a source. -- Григор Гачев (беседа) 22:03, 31 януари 2019 (UTC)

Am I the only one thinking that your argument has some kind of fallacy here, like probably begging the question, i.e. re-stating your premise for the conclusion? Well... Anyway, I think you should also read w:en:Wikipedia:Researching with Wikipedia and w:en:Wikipedia:Citing Wikipedia. By the way, I also thought about discussing this matter at English Wikipedia or English Wikinews. Thoughts or any other suggestions? --George Ho (беседа) 00:16, 2 февруари 2019 (UTC)

I see that I haven't been clear:

Non-primary sources that assemble information from multiple other sources - eg. a Wikipedia article - often tend to emerge patterns that fewer readers will notice from reading the sources they cite: the whole there is more than the parts. To notice these patters from the cited sources only, one would have to effectively re-create most of the Wikipedia article in their head, doing almost the amount of intellectual work that was needed to create it. This often is a lot of work, and not all readers will do it. Those who will not will miss the patterns that the Wikipedia article makes easy to understand.

The Wikipedia article Internet Research Agency is a good example: by connecting facts across several sources, it allows the reader to notice links that are hard to see even after reading all of these sources separately, despite that between them they contain all the facts used. Also, it packs these links into a far smaller volume than that of all sources cited, thus making easier to read them (or possible at all for readers that are short on time). Some of these links are useful for understanding the initiative of Facebook to remove disinformation pages, I believe: they make much easier understanding the truthfulness of the Facebook blog post by Nathaniel Gleicher. Thus, citing the article as a source brings value that citing the sources for this article would not bring (or rather would, but to far fewer readers). That is why I decided to cite it instead of its sources.

This is by far not the only Wikipedia article that connects facts across sources and thus makes understanding their links much easier, available to far greater percentage of the readers. In some cases, that improved understanding may be important - the article will be far more valuable source than the list of the sources in it. For this reason, my personal opinion is that citing Wikipedia articles should be permitted in Wikinews, and possibly in other sister projects: there are cases when it brings more value than citing the list of the sources for that article. -- Григор Гачев (беседа) 18:16, 2 февруари 2019 (UTC)

"Several" isn't that plenty; they would be better sources than Wikipedia itself. Probably that would lead to needs of rewriting, but I think that's better than status quo, i.e. using Wikipedia as a "source". --George Ho (беседа) 21:15, 3 февруари 2019 (UTC)

The emphasizing in my explanation is both on the need for the readers to read sources with bigger volume and on some of the facts not being gathered together and connected in a convenient way in these sources. Not using the article as a source would mean that the readers would need to invest more time and effort on reading sources, and despite that will get a picture that lacks some fact linking that is presented in the Wikipedia article. To match the benefit of reading a cited the Wikipedia article, the readers should do intellectual work, comparable with writing that article - and, as you know, a good article may require hours of work. I would dare to guess that very few readers will be willing to do that just to understand better a news article. As a result, the usefulness of that article will be seriously decreased. It is beyond me why an editor here should strive for this - probably I am misunderstanding you? Please help. -- Григор Гачев (беседа) 23:21, 3 февруари 2019 (UTC)

I'll rephrase: Compromising harder work by doing something simple just to make the article "useful", like merely citing Wikipedia instead of a few or several sources that Wikipedia uses, would compromise the overall quality of the news article. Because anyone can edit, academics have often, if not always, questioned the reliability of Wikipedia. Also, because a common person without expertise of relevant subjects can edit Wikipedia, many academics wouldn't take Wikipedia more seriously. Off-topic (maybe?), but Croatian Wikipedia has received academic criticism for its poor sources and heavy favoring especially toward historical revisionism and nationalism. --George Ho (беседа) 01:13, 4 февруари 2019 (UTC)

Umm... that appears to me a kind of implying a reductio ad absurdum.

Following this logic, the existence of the Wikipedia articles will not be needed beyond merely listing the sources. Anyone could read the sources and learn all the facts. The article itself only can compromise the truth, as anyone can edit it, and many editors don't have an expertise of the subject - so permitting writing an actual article beyond the list of the sources should be forbidden, as a source of potential misinformation.

This argument can be easily used to reject the usability of any encyclopaedia. If you don't assume a Godlike infallibility of the encyclopaedia writers (that assumption would be obviously wrong), then any article introduces danger from mistakes, and it would be better if the encyclopaedias just listed their sources. Unhappily, they would not be able to do that, as the sources itself would be subject to the same rule: why write a scientific review article when you can just include citations from the primary works, and in turn, why write them when you can just include the raw unprocessed data? Eliminating each of these stages removes an important point of introducing errors...

This thinking is what I call for myself "the triangle wheel fallacy". ("A square wheel will mostly do the job, but will cause inconvenient bumps. Therefore, it must be made triangle - that will reduce the bumps by one per rotation." It is essentially a demo that by ignoring some facts in the picture and emphasizing others one can make the perfectly wrong solution to a problem appear the only logical one.)

In reality, every form of derivative work is a compromise between the gain of usability it creates and the risk of introducing mistakes. Good derivative works give a big gain of usability while maintaining the amount of the mistakes low. (Outside of truly exact sciences - mathematics and formal logic - removing the mistakes completely is impossible. Seeking it is like trying to build a machine: if one tries to achieve the highest possible efficiency, that one is a good engineer, but if one tries to achieve exactly 100% efficiency, that one is trying to build a perpetual motion machine, and they are detached from the reality, to say it politely.)

The academics can and should criticize poor sources where they find that, but they cannot refute the fact that almost all articles in Wikipedia still give more gain from usability than harm from mistakes (incl. poor sources). And, given the vast volume of Wikipedia, the gain it brings to its users is much bigger than the gain that brings, say, Encyclopaedia Britannica, who might be better sourced, but is also much smaller. That is the other main point of the existence of Wikipedia, along the freedom of the information in it, I believe. (A bit off-topic: in some of my wikis I have introduced a "put up or shut up" policy: editing to improve is welcome, criticism without contributing is noise to be ignored. Once the people around see that this policy is to stay, things start going much better.)

It is also important to note that the mechanism of achieving correctness in Wikipedia is very different from that in a classic encyclopaedia. In the latter, an article is written by a person or a small group who is implied to be an absolute and infallible authority on the topic. (Having the articles reviewed by other scientists and proofread by editors just increases a bit the size of the group, while still keeping it very small.) In Wikipedia, correctness is achieved by a classic open-source principle: enough of eyes make all bugs shallow, and some of the eyes turn out expert at the topic; initially there may be some mayhem, but eventually things smooth out... The Wikipedia principle is somewhat counter-intuitive to people who are too used to the classic encyclopaedia principle, so it is only normal that they will not trust it and will reject it, due to that bias. It is our role to see that this is just a bias and point it. (Another comparison with the open-source methodology: people without software security expertise often tend to not trust software produced through it, but those wiitth plenty of expertise are almost unanimous that it tends to produce more secure software.)

Wikipedia has another problem too: no article in it is ever finished completely. This is also its greatest strength and should never be removed under any conditions: as a minimum, it is what attracts volunteer editors to it. However, it also means that there will always be articles who are still unpolished, and they will bring the overall quality a bit down. What is important to understand here is that all other encyclopaedias are in a hidden way the same. The percentage of articles still in the works in them is far lower, but also they are static copies that do not develop and get more complete. Today that is the greatest strength of Wikipedia: it is mostly current in a time when an encyclopaedia that is even only one year old is already partially obsolete. As the development of our knowledge will speed up in the foreseeable future, this strength will become even more important. Having articles that are still unpolished is an unavoidable price for that - and is a price worth paying to have it.

To sum all that long explanation: the notion that one should not rely on sources like Wikipedia appears to me to be based on a fallacy that rejects the need for the existence of Wikipedia. Of course, I am open to discussion about it. -- Григор Гачев (беседа) 09:28, 4 февруари 2019 (UTC)

You are open to discussion, Grigor? This begins to look like stonewalling: there is a well-established consensus (WP:CITEWIKI, WP:NOTSOURCE), which you try to bludgeon with walls of text—and not for the first time.
— Luchesar • Б/П 11:40, 4 февруари 2019 (UTC)

I humbly note again that all these are Wikipedia rules, that are valid within Wikipedia and make sense within Wikipedia. They are not Wikinews rules, because they do not make the same sense within Wikinews, due to reasons explained in what you call "walls of text". I admit that it is verbose, but apparently some extra clarity and detailedness is needed: someone might decide to read it and either agree or point deficiencies in it. Some people call this "discussion on merit", and say that it is the way to decide whether something is a good idea or not.

As a start, why not stop trying to just apply here rules from elsewhere, and try to discuss the actual merit of the ideas? George Ho does it very successfully - obviously it can be done. For example, do you think that my impression of his logic being an implication of a reductio ad absurdum is true? I have written my reasons to think so - can you eg. point to mistakes or omissions in my logic? If you can find mistakes that cannot be corrected, or omissions that break my logic, I will have to change my view. -- Григор Гачев (беседа) 17:01, 4 февруари 2019 (UTC)

After long and tedious discussion, it seems that we need a deletion discussion on this article: "Facebook премахна 500 страници, разпространяващи фалшиви новини (17 януари 2019 г.)". We should invite Bulgarian Wikipedians to vote on the fate of the article. Shall you (Iliev) or I start the deletion discussion and then ask the stewards to intervene on this? --George Ho (беседа) 19:37, 4 февруари 2019 (UTC)

What makes Wikipedians eligible to vote in Wikinews? -- Григор Гачев (беседа) 19:59, 4 февруари 2019 (UTC)

Is there a rule against allowing Wikipedians to come to Wikinews to vote here? Many Wikipedians voted for Iliev's adminship; did you forget? --George Ho (беседа) 20:55, 4 февруари 2019 (UTC)

In general, a WMF project only allows voting by those who have some minimum of contributions to the main namespace of the project. I agreed to have Wikipedians with no contributions here voting on Iliev's adminship, because the voting could add a participant to the project, and I believe that approving an admin with a single vote (mine) would set a bad precedent. Had I instead objected before the stewards to having votes by people without a minimum of Wikinews contributions, I an pretty sure that they would agree with me: the established WMF project governing practice is a meritocracy, which requires contributions for getting the right to vote. Even projects who don't have specific rules on voting require contributions for that, except in special cases.

The deletion of a news article does not look a special case to me - certainly much less than the election of an administrator. Also, there already are editors who contribute to the main namespace here, so the rationale for permitting to vote people who don't contribute has decreased. In the light of these, I believe that only editors who have contributions to the main namespace here should be permitted to vote. -- Григор Гачев (беседа) 22:04, 4 февруари 2019 (UTC)

Please cite one rule regarding what you are against. BTW, there's the meta:Consensus, an intriguing read IMO. --George Ho (беседа) 01:55, 5 февруари 2019 (UTC)

The Bulgarian Wikinews mostly don't have any official rules, which I hope to change once (and if) editors come to the project. What I am saying above is dictated by common sense. (In this particular aspect, meta:Consensus is indeed an interesting read.) -- Григор Гачев (беседа) 11:07, 5 февруари 2019 (UTC)

Here's the page about meta:Common sense. --George Ho (беседа) 11:25, 5 февруари 2019 (UTC)

I started a discussion at English Wikinews about the similar matter (but in different fashion). Seems that the community there nearly matches Iliev's view about this. --George Ho (беседа) 02:55, 5 февруари 2019 (UTC)

That could explain something I wonder about - why it should be expected to generate hundreds, maybe thousands of news per day, but in reality can't reach even 10 per day. (The expectation is based on the number of news some other-language Wikinews generate, as a ratio to the editors of the corresponding-language Wikipedia.) It may also explain some of the motivation behind the Wikitribune project. Anyway, it might be a good place to ask some questions - thank you for the idea! -- Григор Гачев (беседа) 11:07, 5 февруари 2019 (UTC)

An article not using sources

JMilusheva wrote "Информационен поток", but there are currently no sources so far. --George Ho (беседа) 11:54, 21 януари 2019 (UTC)

Yes, the article is still being edited and improved. Thank you for your vigilance! -- Григор Гачев (беседа) 14:26, 21 януари 2019 (UTC)

You renamed it to Хоров концерт „Новото начало – Хорове на Ангели“. I mean... Really, citing "Нели Трошева, диригент" as the source? --George Ho (беседа) 19:53, 21 януари 2019 (UTC)

Yes, I renamed it to a more appropriate name (the original one means "Information stream" - not very appropriate for its text). And I can hardly think of a more reliable source about a future choir performance than the conductor of that choir. -- Григор Гачев (беседа) 19:25, 22 януари 2019 (UTC)

Content review #5: More articles not currently using sources

Two more articles "Отворена среща за представяне на ННП „Нисковъглеродна енергия за транспорта и бита (Е+)” на МОН" and "Учени от ИЕЕС-БАН – гости на официален държавен прием в Китай" are not currently using sources to cite essential info. I don't find it acceptable, but I would like to see how admins can handle this. --George Ho (беседа) 00:17, 15 февруари 2019 (UTC)

Both articles of this day are written by a participant in the events who witnessed all described firsthand (in the case of the second, got all the information from one of the two scientists mentioned). Maybe I should think of marking such articles in some way. Thank you for noting this - and if you have ideas about marking properly such articles, I will be grateful! -- Григор Гачев (беседа) 21:01, 16 февруари 2019 (UTC)

Accuracy of statistics

I am increasingly concerned with the accuracy of statistics, saying that there are (as of date) 66 content pages remaining. However, the other page shows 200+ remaining mainspace pages otherwise. I thought about taking this to Phabricator, but I'm unsure. I decide to raise the matter here instead. --George Ho (беседа) 09:37, 25 март 2019 (UTC)

That is definitely wrong, thank you for raising the subject! Do you have a suggestion how I can help for fixing it, other than taking it to Phabricator? -- Григор Гачев (беседа) 18:20, 26 март 2019 (UTC)

Unfortunately, I can't think a good idea that would produce a good result as of date. The error seems too complex to fix for us ordinary users. I read meta:Statistics and mw:Analytics/Wikistats; I can't tell whether that's related. You or I can take this matter to either meta:Tech or Phabricator. --George Ho (беседа) 13:21, 30 март 2019 (UTC)

Phabricator appears to me the slightly better option. Would you be OK if I raise the subject there? -- Григор Гачев (беседа) 17:15, 31 март 2019 (UTC)

Sure, I don't mind. No problem at all. :) --George Ho (беседа) 21:36, 2 април 2019 (UTC)

I tried to count the pages shown by Special:Allpages, and they appear to be the right count. Looks like it is the Special:Statistics page that is wrong. Checked Phabricator carefully: I am not sure that it is the right place for asking for help about that. Decided to ask User:StevenJ81 for a counsel about this. -- Григор Гачев (беседа) 19:05, 4 април 2019 (UTC)

I responded on my enwiki talk page. StevenJ81 (беседа) 14:49, 5 април 2019 (UTC)

Here's the link to create a task, and here's the link to know how to make markups. --George Ho (беседа) 07:38, 6 април 2019 (UTC)

You need to set $wgArticleCountMethod to 'any'. After that statistics works. Any is a good value for Wikinews. --Any-value (беседа) 19:58, 8 април 2019 (UTC)

Thank you very much - I didn't knew about this setting. -- Григор Гачев (беседа) 21:19, 10 април 2019 (UTC)

Промяна в конфигурацията на сайта

На 25 март George Ho забеляза важен проблем с българоезичните Уикиновини - страницата със статистиките на сайта не отразява правилно броя статии (новини) в него.

В крайна сметка беше изяснено, че проблемът идва от настройка на конфигурацията на софтуера (MediaWiki), който задвижва сайта. Към момента той е настроен да брои само тези новини, в които има уикивръзки към други новини. В сайт като Уикипедия, където статии без връзки към други статии обикновено са недоправени микромъничета без стойност, това може да бъде по-оптимална стратегия от простото броене на всички статии. В сайт като Уикиновини обаче, където сравнително малко новини се очаква да имат уикивръзки към други новини, това ми изглежда по-скоро неправилна стратегия.

За технически любопитните: въпросната настройка е променливата $wgArticleCountMethod. Тук тя е настроена като link, което обуславя използването на сегашния алгоритъм за броене на статии. Ако стойността ѝ бъде променена на any, това ще обуслови просто броене на всички статии в сайта. (Статии са тези страници, които са в основното именно пространство.)

Всички потребители с приноси към Уикиновини към началото на гласуването са поканени да гласуват на страницата Уикиновини:Гласуване#Промяна на начина за броене на статии (новини) в сайта.

--Предният неподписан коментар е направен от Григор Гачев (приноси • беседа) 22:01, 12 април 2019‎ (UTC)