Wikinews talk:Alternative Review Process Proposal

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

I think that this process is a good outline of a workable Wikinews model. -- IlyaHaykinson 02:06, 1 Dec 2004 (UTC)

I think the assumption of the Review Process Poll is that articles be "reviewed" before being "published". This method doesn't seem very applicable to that particular poll? --119 03:02, 1 Dec 2004 (UTC)

This method isn't a direct response to that particular poll. I mentioned it in my comments as a reason for my vote in that poll. -- IlyaHaykinson 07:47, 1 Dec 2004 (UTC)
Illya, I think you're on to something here. I've just posted a comment on this page, but I guess we're thinking on a similar track. The speed is everything in this, and for that we need to reach critical mass. And I believe that the number of participants is not important, their ethnographic diversity is. What is your opinion? Fadereu 14:30, 2 Dec 2004 (UTC)

The review process described at Wikinews:Article stages is now substantially revised. What do you think about it, Ilya?--Eloquence

I think that the new stages are better. However, I see a fundumental flaw there — we are putting articles into buckets of in-development, reviewed, unreviewed, etc, but nobody will ever take them out of these buckets. If Wikipedia functioned like this, then 95% of the articles would be in some random state other than "everyone reviewed it and said it was OK".
I strongly believe that while articles have inherent stages, it is impossible to force codification of those stages by tags. Take the following scenario. I write an article and it's of local interest. I believe it to be correct, and I publish it to, say, my geographic location's news page. I believe the article to be correct (NPOV, complete, verifiable, etc) -- yet I am forced to either submit it for review or put the ominous noreview tag on it. Since nobody else will be interested in this article, it will remain in this mode forever.
Instead, I think we need to assume that each article is completely fine, and mark it with specific problems. Much like on Wikipedia we will mark things with npov or candidate for speedy deletion, we would mark problem articles here.
Don't read me wrong: I think the new process is better. But I'd like to hear your thoughts on what I've said above. -- IlyaHaykinson 09:16, 3 Dec 2004 (UTC)
Well, I very much disagree with that. We should not blindly assume that just because you think your articles are NPOV, complete, verifiable, etc., that they actually are. Otherwise we can have lots of cranks working alone on niche subjects whose articles are treated the same as those on major issues which have undergone intense review by the community - that just doesn't make sense. The reader should immediately know whether or not the facts in an article have been verified by other editors. If the article is on such a niche issue that no realistic verification can take place, then not only is the noreview tag appropriate, but the very existence of the article is questionable: Verifiability is one of the key criteria for Wikipedia articles, and to some extent at least, we will have to transfer it to Wikinews.
Please also understand that the proposed revised process is a compromise between two very sharply different positions: Those who would like to see all articles screened and reviewed systematically before publication, and those who want no review process, or a very limited one. As such, it is necessarily not completely satisfactory to either party in the current debate.--Eloquence 09:21, 3 Dec 2004 (UTC)
I think that the new article stages process deserves to be field-tested. I understand that working in a group requires these compromises; it's only fair. We'll see over the next few days / weeks how well the compomise process works to get Wikinews built up to be some sort of a useful data source. Till then I guess I will try to make it work despite my reservations :-) --IlyaHaykinson 10:30, 3 Dec 2004 (UTC)

--Shana 22:00, 5 Dec 2004 (UTC) I think the "bucket" problem mentioned above stems from the fact that the policy that we should follow to publish articles that have been reviewed is not clear. When reading about what happens when the 8 hours expire and no one has objections, I get the sense that the article has therefore been accepted by the community and should be published.

That raises a logistics problem, because the author might not have realized that he has to move the news out of the workspace and possibly publish in the main page, he/she might not have the time, and the story will end up stuck forever in review, as I've seen several yesterday and today.

I've published a couple of stories so far, reviewed a few others, and mainly went about moving headlines around because they were in the wrong places, and I'm feeling that this whole process is still terribly slow. I worked with implementing a CMS system for a major news site here in Portugal, and I can tell you that they are more concerned about getting the news out than with following a fixed reviewing policy. We all know how that can turn out sometimes, but still, if an article has just been written, I think our target readers are sufficiently intelligent to be aware that it might not be in the ideal form, but at least they were able to read the news, even if in incomplete form. Before we reach critical mass there won't probably be many problems with incorrect news being published in the main page, and after we reach critical mass, we'll have enough editors to quickly handle that situation if it arises.

Anyway, I think we should have the main page as updated as possible, even if risking some more controversial news. That'll definitely draw more folks in, and that's really what we need at the moment.


After following the December 3, 2004 article-stages process, I'm convinced (like Shana above) that the current process is not optimized for a news site. The concept of article stages can only work efficiently in an organization with clear article owners (authors), who will care to push the article they've written through the review process. On Wikinews, however, an article moves through stages only by accident — when discovered by an enterprising editor who decides that it's been long enough, or that consensus has been reached or whatnot. I thus return to promoting this process as a more workable solution. -- IlyaHaykinson 19:20, 10 Dec 2004 (UTC)

This review process has my support. The current review process should be switched out immediately by blanking the current stage templates, and adding a couple of trmplates like NPOV and disputed and such. I'd like to see a vote on this. Or, the necessary templates can be created and a "Alternative Process" template can be created and used for articles that use this review process. --Regebro 20:04, 10 Dec 2004 (UTC)

Guys, I'm thinking I disagree with you, for a few reasons. One, the bucket issue. Above, Shana mentions that, and worries about the logistics in transferring articles to and from stages, espcially when the author is not around. The answer is simple: Anyone can do that. The stages are just like anything in the wiki- editable by anyone. That's the benefit of a wiki, not a failure. And essentially, that seems to be what you want. The ability for articles to be linked at anytime, etc. The issue becomes that there truely are two sides to this, as Eloquence said, one that wants a strict review policy, and one that does not. By allowing a free for all, we come to the issue of accuracy, which I will address below.
Two, regarding Shana's issues with the speed which things are coming out, versus the chances of mistakes/POV. because we are a sister project of Wikipedia, and WP has always been under scrunity for accuracy, we stand to be under greater scrunity. Add to that being a news site, and it would be very easy for us to publish a very highly inaccurate pages and get terrible reviews/opinions from all over the place. Would that matter, nesseciarly? No, but it's never a good thing for momentum/drawing folks in. I agree, speed has been slow, but I think we are fighting a chicken/egg arguement here.
I'd rather have accuracy first, and work as a "weekly magazine", then have piles of badly done articles scrutinized by the public as "news". We already have piles of news sources making mistakes or just plain making things up. We shouldn't be another one of those. Yes, because we are online and a wiki, we have the ability to allow anyone to read any article, but accuracy has to be our goal here.
Let's actually encourage editors to act like they were the first two days of the demo. Articles were written, tagged, and then, lo and behold, people would comment on them, and review them. Result? A few solid articles a day. Now? Instead of that, we ahve people ignoring reviewing, and randomly writing articles. Articles languish with bad information and bad wording, not BECAUSE we have the review tags in place, but because we are basically acting like your proposed policy is doing, save that articles are only added to the workspace/editor tasks, instead of to any random page.
Lyellin 21:13, 10 Dec 2004 (UTC)
The weekly magazine idea is very hard to implement in an open Wiki, and requires an immense workload, much higher than the type of news site that it is already shaping up to be.
When it comes to accuracy, look for example at the article Researchers discover high temperature enables more efficient hydrogen generation. Can I say if the statement about the efficiency is correct or not? Nope. But I can spot a bunch of science mumbo-jumbo supposedly said by a "Jeremy Desterhoft". A quick google and a search through the references turns up no Jeremy Desterhoft at all. He might not exist. In the current review process that means I make a comment, and if no consensus is reached, the article fails. That's very strange. It seems to imply some kind of ownership to an article, an ownership that does not exist in a wiki. Normal wikipedia review process means I review the offending part. That makes the article fine, and everybody are happy. How, instead, we have a failed article that still contains bogus information. What is the point of that? How does this help accuracy? Not at all, of course.
This is now about pitting accuracy against speed. It is about using the wiki instead of fighting it. This is not a weekly magazine, or a daily paper, or a TV channel. Nor should it try to be. Embrace the wiki! (Read my user page on this topic). --Regebro 00:08, 11 Dec 2004 (UTC)


I think that aiming for a weekly magazine is nearly impossible, and can't be a result of Wikinews — if we strive for top-notch accuracy over everything else then unless we focus on a very narrow niche we are unlikely to get enough articles to be useful as any kind of a news source.
The problem with the bucket system is that if nobody is in charge, nobody will do it. Note how this bucket system differs from Wikipedia: there, the only explicit bucket that is part of the core process is the VfD one, and even there with lots of editors the process is complicated and requires a lot of work to maintain. In our current system we have many explicit buckets that have to all be maintained, but unlike votes for deletion the review process is much more complicated and subjective.
The value of a news site is that when people visit it, they see the news that is current and relevant. If we have 10 really really really top-notch articles since yesterday, we will be of little use. If we have 100 so-so articles we'll be closer to useful: and useful means that people will visit and fix problems in articles that are bad.
The current system ecourages authorship — reviewers make comments like "this article seems incomplete" instead of fixing an article, probably expecting whoever posted the article originally to care enough to flesh it out. This is not the wiki way, and has to be discouraged. The Wikipedia model of publish first, fix second, ask questions later, would work much better, and is the goal behind this current proposal. --IlyaHaykinson 02:40, 11 Dec 2004 (UTC)


Hi I´m from the german Wikinews. And we just have the same discussion about article stages. Currently we have split up the newsarticles in two groups: articles in development and completed articles (both are still fully editable of course). I totally agree with your Alternative Review Process and I am trying to convince the german "Wikinewsler" for this solution. As Regebro said it´s all about fighting the Wiki or embracing it. --MilesTeg 02:15, 11 Dec 2004 (UTC)