Wikipedia:Village pump (proposals)/Archive 79

Why not post a wikipage that all topics related for the goodness of our nature will be there. So people could be exposed what are the bad effects that could kill our home planet. With this simple message hope you find the idea what I want to say.. Thanks guys and More power to your team — Preceding unsigned comment added by 202.57.35.170 (talk • contribs) 09:51, 14 October 2011‎ (UTC)

That might fall afoul of Wikipedia is not here to tell the world about your noble cause, but we do have articles on environmentalism and related topics. Dcoetzee 10:08, 14 October 2011 (UTC)

Maybe you're looking for

Wha?

11:51, 15 October 2011 (UTC)

We have to remember that Wikipedia is not a soap box, much as you might have some noble cause you wish to tell the world (see Wikipedia: What Wikipedia is not), but as has been pointed out above, we do have an article on environmentalism. If you feel strongly that Wikipedia needs more of this type of article, why not set up a WikiProject on environmentalism if there is not one already? ACEOREVIVED (talk) 15:08, 20 October 2011 (UTC)

There is actually Wikipedia: WikiProject Environment. ACEOREVIVED (talk) 15:21, 20 October 2011 (UTC)

Hi all,

I've just checked some of the old messages about this topic, which has been put in light time to time, most recently there: here and there (with the response "We do have the Simple English Wikipedia.")

I am a French wikipedian, and would like to tell you about this pretty old project: a wiki encyclopedia designed and partly maintained by children.

The idea of a equivalent of Wikipedia for children was discussed in particular in 2005-2006 on this page : meta:Wikikids

Although it had no continuation as a Wikimedia project, wikis with such a feature were launched first in Dutch: WikiKids; in french a few month later: Vikidia ([1]) and then in Spanish (Vikidia in spanish) for 8-13 years old children. I opened Vikidia.

WikiKids and Vikidia in French are doing well and are quite alike in size and activity, whereas Vikidia in Spanish doesn't make it so well.

On Vikidia in French, we currently have a guest-book opened, and the comments left on it are quite encouraging (Google translation). Children say they appreciate the articles to be more readable for them than on Wikipedia, their main reserve being that some article are not developed enough, or that there isn't articles on every subject they would like to know about... They clearly expect (and claim) some substantial content, though it has to be easier than the wikipedia's content.

We have yet a bit more than 10,000 articles in Vikidia in French, and about 220,000 unique visitors a month.

This "Wikikids" question was mentioned again in the Wikimedia scope one year ago there: meta:2010 Wikimedia Study of Controversial Content: Part Two#Recommendations: Controversial Text:

Recommendations: Controversial Text

Because of the considerations outlined above, our recommendations surrounding text in WMF projects are the following:

It is recommended:

(...)

3. That, however, the Foundation investigate the creation of a “WikiJunior” version of the Wikipedias, aimed at children under the age of 12, either as a stand-alone project or in partnership with existing and appropriate educational institutions.

Explanations:

(...)

Recommendations 2 and 3

(...) Much more successful, in our opinion, is a project specifically targeted to children, and to the quite different needs of children in different age groups. Some projects of this nature have already been begun in the WikiJunior section of WikiBooks,^[1] but it is our feeling that the scope of such a venture might necessitate the formation of partnerships with institutions who have experience and resources already devoted to this area.

...to which I wrote this response on meta:User talk:Sj.

Another feature of WikiKids/Vikidia is of course that it let children be involved in building the content, and does not "only" aim to produce and offer content for children, for the same benefits that you can find in Wikipedia-editing workshops for students. There is both some school projects and volunteer (and often enthusiastic) children editing on Vikidia along with teenagers, students and adult writers.

I do have in mind that there is Simple English Wikipedia, whom no other equivalent have been created in any other language. I guess that it would make less easy to create a wikikids in english, as it would be partly similar to simple. I would also say that the fact that this wiki is not so big and hasn't been created in other language may show that its aim would not be so clear, mobilizing and justifying. By the way, on Vikidia, we do hope to be usefull not only to children but to people of any age that would like a simple approach of a subject, or a shorter one than on Wikipedia. We beleive that if its content were not relevant to older people, it wouldn't be good either for children.

My question is of course: why is there no Wikikids in English yet ? ;-)

Greetings, Astirmays (talk) 18:04, 20 October 2011 (UTC)

Well, as your quote mentions, there is wikibooks:Wikibooks:What is Wikijunior?. But with regard to your propopsal, I wonder if this is really the right forum for discussing this? It seems like you are proposing an English-language version of Vikidia, but Vikidia isn't a WMF project... --Philosopher ^{Let us reason together.} 22:04, 20 October 2011 (UTC)

On second thought, if you are proposing a new WMF project, the correct place for the proposal is at Meta's m:Proposals for new projects. --Philosopher ^{Let us reason together.} 22:05, 20 October 2011 (UTC)

Actually, this project is not the same as Wikijunior. It is a Wikipedia-like project, which Wikijunior isn't. Secondly you are right. The right place is meta:Wikikids. However, would it be a WMF project or not, I think that it is worthwhile to sound out wikipedians about it. Just as other wikis or other enterprise, its realization depends in the first place on a few people to undertake it. That's a call to them.

By the way, here are (among others) two other discussion on that subject: in June 2008 and July 2010.

Another thought, one may even link this projet with one articles of the convention on the Rights of the Child which is the following:

Article 13

1. The child shall have the right to freedom of expression; this right shall include freedom to seek, receive and impart information and ideas of all kinds, regardless of frontiers, either orally, in writing or in print, in the form of art, or through any other media of the child's choice.

2. The exercise of this right may be subject to certain restrictions, but these shall only be such as are provided by law and are necessary:

(a) For respect of the rights or reputations of others; or

(b) For the protection of national security or of public order (ordre public), or of public health or morals.

Last thing, the point of this message was also to address the "I don't think it can work well" objection, associated with issue such as vandalism, censorship, defining the appropriate content, children motivation and ability to edit articles... since it does work well at least in two language, where there is children, teenagers and adults working together to better the wiki. And thousands of readers every days.

Greetings, Astirmays (talk) 00:59, 21 October 2011 (UTC)

If you wanted a wikipedia in simple french, yoou should of voted in the request for comment. This pretty much a prennial proposal. ~~Ebe123~~ (+) ^talk
_Contribs • 10:52, 21 October 2011 (UTC)

I don't want a Wikipedia in simple french, because: - I beleive that a wiki designed for children is a more judicious feature than aiming to be written in a simple language - because their scope would intersect quite a bit - and at last because Vikidia allready exists.

The point is to suggest to launch an encyclopedic Wiki for children in english (and possibly other languages). Astirmays (talk) 15:17, 21 October 2011 (UTC)

There was just a discussion of this in Jimbo's talk page. It was pointed out that we already have a Kid's English Wikipedia:

Wikipedia CD Selection. It is on CD, not online, and it is not editable. All that would be required would be to put this online. This would not be too hard, since all the work has already been done and it already exists. If we just wanted to put this on-line and make it official Wikimedia offering, that would be trivial. It is already a bit out of date and would become more so, but that is not so terrible. If we wanted to make the basis of an editable, updateable project that would take a bit more work, but much less than starting from scratch. I would probably favor this. I'm not sure where to begin to make this happen -- possibly at Wikimedia proposals for new projects as mentioned, or maybe just lobbying Jimbo and Sue and on the mailing lists? Somebody has to be willing to do this political work. While I'm supportive, I'm not that interested and so it won't be me. But it should be doable. Herostratus (talk

) 15:29, 21 October 2011 (UTC)

Thanks for this info. I found this discussion here. Actually, the 2008/9 Wikipedia Selection for schools is online there, thought it has no search engine.

A wikikids project has to be editable (and maintained with the same methods as Wikipedia) See above: "The child shall have the right to (...) seek, receive and impart information..." I think that the Wikipedia Selection doesn't fit to provide a base content because articles have been checked but not rewrited to be easily readable by children. A possibility is rather to use and import content both from b:Wikijunior and Simple: when it fits to this project. That would make it possible to start much faster that from scratch. Astirmays (talk) 20:30, 21 October 2011 (UTC)

In view of the fact that yet again Henrik's Page View Statistics' is not doing its job: for example, for the Yeung Kui-wan article: from 1 Oct, to 4th Oct, 2011, then fixed and now [14 Oct] it is 4 days adrift), it is sad (and a great pity) that such a great tool doesn't seem to get official Wikipedia support. For me, as a fledgling contributor, it is essential to be able to follow the evolution of hits on 'my' articles' on a daily basis. It also seems to be very popular with many other contributors.

So, I plead that someone 'up there' has a look at this tool and makes it an official, Wikipedia function, which will be maintained on a regular basis. Please advise what are Wikipedia's policy and planning for this tool.

It would be perhaps interesting to set-up a poll to sound out just what is the level of interest in Henrik's 'Page View Statistics'.

I hope this forum is the right place in the panoply of Wikipedia forums for this plea. [Note I am advised that this is the case: "The correct place would probably be Wikipedia:Village pump (proposals)".-- Obsidi♠n Soul 07:30, 4 October 2011 (UTC)"] Duncan.france (talk) 01:05, 14 October 2011 (UTC)

This great thing that you want us to either support or fix (it's not clear to me yet): what is it and where would we find out more about it? Who is Henrik? And what of Esmerelda? — JohnFromPinckney (talk) 02:01, 14 October 2011 (UTC)

See

Help:Pageview stats; the tool is maintained by Henrik (talk · contribs). -- John of Reading (talk

) 07:32, 14 October 2011 (UTC)

Just to be clear, this refers to stats.grok.se, which apparently relies on data from dammit.lt, a site which appears to be down at the moment (perhaps explaining the problems with Henrik's tool). According to [2] [3], the man to talk to is Domas Mituzas, aka midom (talk · contribs).

Henrik's tool is well-used, especially by

WP:DYKSTATS, and it is linked from every history page. It is fairly important to have such a tool in good working order. — This, that, and the other (talk)

09:21, 14 October 2011 (UTC)

There was once a superstition against making page-view statistics available (I think there was a fear that people might hit pages deliberately to increase the statistics, or something). But now we have an external tool anyway, people find it useful, and it doesn't seem to lead to abuse, perhaps the developers would be persuaded to make such information available on-wiki (so it's not dependent on an outside site being up).--Kotniski (talk) 10:14, 14 October 2011 (UTC)

Page view stats is already available from the link on all page histories. It works perfectly, is very fast, and is indispensable for some work. Kudpung กุดผึ้ง (talk) 10:18, 14 October 2011 (UTC)

Trouble is, it's a bit intermittent. — This, that, and the other (talk) 11:03, 14 October 2011 (UTC)

Henrik's page view tool is great but it does go down from time to time. If Henrik is busy or out of town the tool may be down for a week. This tool should be supported by more than a single volunteer. -- SWTPC6800 (talk) 02:21, 15 October 2011 (UTC)

If that's the case, then yes. Kudpung กุดผึ้ง (talk) 06:04, 15 October 2011 (UTC)

It's a bit depressing to think that all this data gathering is thanks to one Lithuanian database engineer (Domas Mituzas). None of these tools (except it seems http://en.wiki-watch.de/ ) will work when the server in Lithuania is down. And let's face it, it doesn't look like the foundation is going to help any time soon. Maybe they see it as potentially opening a can of worms, "official" view statistics that can be manipulated from outside would make them an easy target for critics. Any solution will have to come from volunteers I think. Split up the data from the squid server logs, send to multiple clients so each receives all data on their share of the articles, and set up a p2p network to distribute the data to anyone who wants it, providing a robust backup system, redundancy, load balancing... DS Belgium ٩(͡๏̯͡๏)۶ 13:04, 15 October 2011 (UTC)

The raw stats are now being made available through WMF servers, so I suppose that Domas maybe no longer needs to maintain his site. I guess Grok hasn't migrated. Way back when, we didn't have page view stats, not because of philosphical objections, but because the servers couldn't handle the load required by logging. This was resolved by embedding a "bug", if I remember correctly, which picks up a hit on another server that basically did nothing but log. Rich Farmbrough, 17:45, 16 October 2011 (UTC).

In response to the 'optimistic' observation of Kudpung กุดผึ้ง (talk), I recommend that he look at : http://stats.grok.se/en/201110/Yeung%20Kui-wan (obtained using the quoted Henriks' website: http://stats.grok.se/en/latest/Help:Pageview_stats. As of now (17/10/2011), this shows, no data after 7/10/2011... for the 'Yeung Kui-wan' article!! Now, how is that "working perfectly"??? Please, please, will someone 'up there' support this very useful (and popular) 'service'??Duncan.france (talk) 10:37, 17 October 2011 (UTC)

It now shows data till 16/10/2011. Domas server was up for a while today, (well, yesterday, 5 hours ago) but stopped responding 2:40 hours ago. Long enough for Henriks' site to download the data it seems. DS Belgium (talk) 01:57, 18 October 2011 (UTC)

I looked at http://www.wikiroll.com/ he sais "Using Amazon AWS services I discovered that "Wikipedia Traffic Statistics" data in freely available". I found this http://aws.amazon.com/search?_encoding=UTF8&searchQuery=Wikipedia but still can't find the source!! Goddamn! Probably missing something obvious.. 4:08 am, time to sleep DS Belgium (talk) 02:10, 18 October 2011 (UTC)

Just out of interest, how do people know this is not doing its job? I thought that if one went to the article history, one could click on "Page View Statistics". I am willing to bet that sometimes this might be a little slow, but could there not be template to indicate this, as there is when one clicks on one's "Edit Count"? ACEOREVIVED (talk) 15:58, 20 October 2011 (UTC)

well, it's still working perfectly for me at Malvern,_Worcestershire and I use it a lot because it's a GA I wrote. Why can't it be hosted on the ToolServer?

Just want to 'plus one' this call. The page view data is very valuable for research and for explaining to people how important Wikipedia is. Its an important data point for recruiting and keeping editors as it gives them some great positive feedback that their work matters. Whether its Henrik's tool or another, please support access to page view data. Benjamin Good (talk) 22:57, 20 October 2011 (UTC)

In answer to ACEOREVIVE's query as to how one can tell whether Henrik's page-view statistics are doing their job or not, I keep an on going record of the hits on 'my' articles 'Yeung Kui-wan' and its redirect 'Yang Quyun'. I have observed the following anomalies:

1) Periodically, the data for a number of days 'jump' to a different set of days. This occurs at times when 'the server is down'(?). When the 'server comes back on-line' (say), the data-set jumps to the correct days.

2) There are days when zero hits are shown in both articles. Surprisingly these 'zero-hits' days are off-set by one day.

3) The last days of data for 'my' two articles are always off-set by one day [the redirect being one day behind]. I am not sure whether this is an error or not.

Hope that is clear... Suffice it to say, that the effects I am trying to describe are VERY frustrating...

Please can the powers that be up there in the Wikipedia hierarchy take this service under their wing. This would hopefully prevent such anomalies.

Having said that, Henrik's service, frustrating as it is at times, is preferable to having no hits stats at all.Duncan.france (talk) 03:59, 23 October 2011 (UTC)

I would like to have a "supplementary information" page to another page, but I am not sure on what the guideline is. Can I do the "main_page/extra_page" trick like on user-pages or do I have to make a self-standing page? Basically, I have a page with large cladograms ("family trees" but for species) that I would like to hide unless the user wants to see them (table collapsable collapsed messes up the formatting). Thanks --Squidonius (talk) 05:03, 22 October 2011 (UTC)

• Oppose technical implementaion for now. I definitely support using it for taxonomic purposes, but it would cause too much confusion. However, I weakly support using it non-technically (like Software/feature without an actual link back to Software from Software/feature) provided that they are supplemented with appropriate redirects, as it can definitely be useful for organizational purposes. There are still, however, many other things like page moves that need to be resolved.Jasper Deng (talk) 05:09, 22 October 2011 (UTC)

It seems you (Squidonius) are not so much proposing this as asking for guidance. This isn't really the right place to ask I think (but I don't know (other than the article talk page) where the right place would be; Maybe a helpdesk?). However a possibility to solve your issue would be to contain the cladograms (must look that word up later) in scrolling divs. That way each could take up just a few lines but be fully viewable without leaving the page. Further to this each could (since they don't spread far horizontally) could reside next to others thus shortening the page as a whole. -- fg^TC 06:21, 22 October 2011 (UTC)

I put an example here -- fg^TC 06:33, 22 October 2011 (UTC)

Thanks for checking! That is a good idea, however, the problem is not only of space but also that there is a limit on number of templates per page and Bacterial phyla has gone over that. So given than the name/sub-name option may cause technical problems, I suppose that making a normal article for each tree is the only option... --Squidonius (talk) 07:25, 22 October 2011 (UTC)

Ah I see. Good luck. -- fg^TC 18:42, 22 October 2011 (UTC)

Subpages are disabled in mainspace, so if you create

wp:Chemistry has been using supplementary information pages for years :Methanol (data page), Arsine (data page), Tryptophan (data page). Yoenit (talk

) 08:04, 22 October 2011 (UTC)

We (User:Pleiotrope, User:i9606) are thinking of expanding the use of the {{SWL}} template in article space and would like to have some community input. Basically, the {{SWL}} template allows editors to encode semantic information within the article space. The semantic information winds up as a machine-readable representation of a relationship hidden in the HTML that could later be extracted and used in a number of different ways.

For example, the SWL template could be used to encode the attributes of animals. Using the

It is a strictly [[nocturnal]] [[owl]], which feeds on small mammals and birds.

into the following:

It is a strictly {{SWL|type=behavior|target=nocturnal}} [[owl]], which feeds on small mammals and birds.

The two render identically, except that the {{

We’ve also developed a userscript that demonstrates the usefulness of these relationships. The script parses the SWL templates in the article and, if prompted, highlights them and displays an infobox describing each relationship and target. This allows for fast and concise subject comprehension: for example, on this Spotted Owl sandbox copy (with SWL templates added), the infobox would clearly inform the user that the Spotted Owl is nocturnal, mates monogamously, and lives in North America. The userscript is available here if you'd like to test it out, and a screenshot of the userscript is below if you would rather not (the infobox is above the owl infobox):

We would appreciate any feedback or thoughts the community has regarding this concept; if it's something worth discussing more formally, we could make it a RfC... but that seems a bit involved at this point. Thoughts? -- Pleiotrope (talk) 23:38, 18 October 2011 (UTC)

Is there a reason this isn't incorporated into the infobox? --Philosopher ^{Let us reason together.} 19:39, 19 October 2011 (UTC)

From the reader's perspective, this userscript gives context to each asserted fact. It also allows the addition of new fields that do not exist in the relevant infobox (perhaps because they are too specific). In addition, it allows users to clarify the relationship if they find it to be imprecise; for instance, "behavior = nocturnal" may be better represented as "active_period = nocturnal" or something like that, which would be quickly fixable by casual editors. In addition, infoboxes tend to be "least-common-denominators" of their subjects. By continuously adding conditional fields to the infobox for attributes that may, for example, only apply to owls, the maintainability of the infobox decreases for only marginal benefit. Do you agree? --Pleiotrope (talk) 21:09, 19 October 2011 (UTC)

The casual editor is likely to strip out the template entirely, because they can't make heads or tails of it and just want to update the article to correctly link to nocturnality. --Carnildo (talk) 02:03, 20 October 2011 (UTC)

The promise of semantic coding to replace the current infobox method is the much greater flexibility and reusability; I think the immediate applications would be the ability to rewrite and recombine articles in different formats (for example, solving the perennial AfD disputes over whether we can justify small articles or need to combine them, by permitted the display in either manner), and to improve findability by providing a basis for intersection of categories; in the longer term I foresee a fill in the box approach to routine articles. Perhaps the casual editor is not ready for this, but neither is the casual editor able to properly organize articles. Doing either right takes discipline and organization. DGG ( talk ) 04:32, 20 October 2011 (UTC)

I'm far from expert (in fact I was introduced to

this idea by seeing/reading this proposal) but am sure there is a great value in embedding data in such a way that scripts can use the page more effectively. I would like to see a standard devised/constructed to shape the use of this template. Without a standard form this template might be used in so many different ways that scripts would be no better off reading the embedded data than reading the page without it. In the OP's example we see 3 types with each their targets. type=behaviour is all well and good but only if behaviour is a recognised term a script might want to use. If all of Category:Birds had semantic-behaviour as an expected type the embedded data would be easily used and thus (more) useful. I'm probably not using the right terms but hope my meaning is clear enough. If nothing else it gave me some typing practice . I guess in essence what I'm saying is that random collections of data are no more organized than human-readable pages so if this data is to be used by machines it would be best to get it as organised as possible before it spreads too far in a disorganised form. -- fg^TC

05:26, 20 October 2011 (UTC)

If I understand you (FG) correctly, it sounds like you are suggesting the development of an ontology of relationship types that would be used to standardize the use of the SWL template. I think this is a laudable goal that would indeed increase the value of the semantic links, the main question is how we achieve it. It sounds like you are advocating a pro-active attempt at building this ontology and some kind of subsequent enforcement of its application. (Correct me if I’m wrong!) In my opinion the relationship types should emerge organically from their use in much the same way that articles do. Someone boldly creates one, others disagree and make changes, and in most cases a stable consensus eventually emerges. Also, in the worst-case scenario that you allude to - of prolific SWLs without consensus and consistency - I still feel that the data would be more useful for scripts than plain text. If you look at Pleiotrope’s example, I would personally would prefer to use a different property than “behavior” to link the owl to “nocturnal”, yet the presence of this loose semantic link is still valuable. The script extracted and displayed it correctly. I got the basic idea and, because I saw it, I am substantially more likely to get in there and alter the SWL to something that makes more sense to me. Benjamin Good (talk) 18:09, 20 October 2011 (UTC)

Yes "ontology" is a very good word. I would have used it if I'd known it. "Organic ontology" makes for an interesting Google search. It seems we are paddling along with the swell of a wave. Do we try to stand or let it roll beneath us? I grabbed a few links to share. I visited each and have seen nothing suggesting ill will at any of the addresses (I worry about viruses etc).

• http://bricobase.com/organic.shtml
• http://fractalontology.wordpress.com/category/parasite
• http://www.duo.uio.no/sok/work.html?WORKID=58292&lang=en

If the semantic data were held in something akin to a JavaScript or JSON "object" we would have the basic set of rules steering the way for standardization with the freedom that our object is effectively unlimited. This concept would allow as a tree allows twigs that all semantic data would be born of a solid trunk (root would sound better but not so good for the analogy). Although we Humans are extraordinarily adaptable, machines/scripts are not (yet). Since the best use of semantic data is as machine read data we need to consider the needs of a script. It wants predictable structure. A JSON parser deals with any JSON object thrown at it because it follows specific rules. The object it reads could contain anything but it won't upset the script. There we have Organic (unlimited and adaptable) Ontology (formal, explicit specification of a shared conceptualisation). I think we should be bold. Take this bull by the horns and tame it. -- fg^TC 22:08, 20 October 2011 (UTC)

So, just to make sure I've got your point here - you are saying that the ontology we are talking about can and should emerge and evolve organically as the examples you provide in the links enable, yes? Also, regarding the needs of scripts. The SWL template does currently emit machine readable HTML and there are a couple prototype scripts that already process it. It has a consistent, parseable structure (i.e. 'syntax'). The only challenge is clarifying the meaning or semantics of the relationship types it is used to encode. Right now the relationship types are represented as categories (sub pages of the upper SWL category). This is where the community can work to evolve a useful ontology of relationship types. Benjamin Good (talk) 22:50, 20 October 2011 (UTC)

Yup it's the style of the embedded data that I think needs to be standardized not the data itself. The JSON object example is a close to a practical example I can think of. Allowing author freedom while a script sees only order. That the script can read the template is not in question it's more what it will get from it. On page a of cat:a the script gets perfectly reasonable data and on page b of cat:a it also gets perfectly reasonable data. If it then tries to compose a table using the data from all pages in cat:a it runs into a fatal problem. The data types have no obvious correlation. As Humans we are used to interpreting what we find and filling in the gaps. We can edit and alter things as we see fit. Scripts can't (unless extremely complex and massive) and thus the perfectly reasonable data from all cat:a pages becomes almost valueless. If however all cat:a pages had a specified object structure a script could draw data from any page in that category and use it any way it was written to without hiccough. If the idea of categorical structure was carried outward to encompass all categories e.g. ((A(a)(b)(c))(B(a)(b)(c))(C(a)(b)(c))) we have a secure working environment for scripts across all categories but each cat can have it's own quirks and even each page. I think I may be explaining this badly. It is quite possible that I understand the subject so poorly that my opinion is pretty much worthless too . I am just thinking that while the chance exists it would be good to at least attempt to lay down some guides that make the templates both user-friendly and script-friendly with an eye on the forseeable future. Too much of our modern systems are reliant on old ideas so deeply embedded that we are trapped into their continued use. It seems there is great value in semantic data on the web and it would be nice if it was organized...My thinks are coming out upside-down. Owl:type=behaviour|target=nocturnal;Sparrow:type=color|target=brown;result=messed up table of unrelated/able data. More options and nested options. If we are gonna have a tool lets make damn sure it's as good as it can be so it doesn't snap and sharp bit stabby stabby! -- fg^TC 23:37, 20 October 2011 (UTC)

OK, I think its important to separate authoring/viewing tools from representation here. I think it would be very useful if we had a tool that would help editors author SWLs in a consistent way by suggesting relationship types to use as the semantic link is being created. This is totally achievable given the current representation and I would be happy to help work on it.Benjamin Good (talk) 00:02, 21 October 2011 (UTC)

Now that is an extremely good idea! -- fg^TC 00:00, 21 October 2011 (UTC)

I agree with Benjamin, the ontology should emerge from the community itself, suggestions in order to facilitate things to users should be provided as well. However, will this ontology be empty at the beginning? I think some common relations can be added at the beginning and then let it evolve. What about using relations coming from other ontologies? skos:broader and skos:narrower for instance? -- ljgarciac 17:43, 23 October 2011 (GMT)

There are several problems with this idea. First of all, the hard part is not creating a script that makes a "fake" infobox from some data. The hard part is coming up with a reasonable data structure. And in this case it hasn't been done. You have a pair "type=behavior|target=nocturnal". Now you are proposing "type=active_period|target=nocturnal", someone else will mark it as "type=is_nocturnal|target=yes", "type=is_nocturnal|target=1" etc. Also, someone will take something like a sentence in lead of article "Pony" "Ponies are generally considered intelligent and friendly, though sometimes they also are described as stubborn or cunning." and make pairs "type=behavior|target=intelligent", "type=behavior|target=friendly", "type=behavior|target=stubborn", "type=behavior|target=cunning". The result will be a complete mess - any program that can make sense of such data will probably make more sense of the article itself. Furthermore, that way we lose information: in the article the equivalents of "type=behavior|target=intelligent", "type=behavior|target=friendly" were qualified by "are generally considered", while "type=behavior|target=stubborn" and "type=behavior|target=cunning" - with "sometimes they also are described". We lose those qualifications in the "fake" infobox.

There is another problem concerning confusing syntax, but others have already explained it, thus I'll just add that almost every possible way to make this system more useful to computers will make the syntax even more confusing.

Thus I'll put it as harshly, as I can: nothing similar must ever be implemented in Wikipedia. Each Wikimedia project must do one thing and one thing only - but do it well. Wikipedia is meant to be an encyclopedia - and nothing else.

But while nothing like that can be allowed on Wikipedia, I don't see why we shouldn't create a different Wikimedia project, that would be dedicated to just that. "Wikiontology", perhaps, referring to Ontology (information science)..? So, go to m:Proposals for new projects and propose it! Yes, seriously. Looks like there are enough users who would like to work on such project. --Martynas Patasius (talk) 18:29, 20 October 2011 (UTC)

Dear Martynas, a quick note regarding your pony example, I assume that we would trust editors to use only add the template to links where it would make sense. Currently none of the words you highlight ("intelligent", "friendly", "stubborn", "cunning") have been wikilinked, so personally I don't think anyone would add the SWL template to them. On the other hand, I could see someone clarifying the wikilink to "horse" using SWL ({{SWL|type=type_of|target=horse}}. That would allow a user to easily generate a list of animals that are types of horses, differentiating them from animals that eat horses and plants that are eaten by horses, for example. Cheers,

talk

) 19:26, 20 October 2011 (UTC)

Er... So, someone who wants to put some fact into "fake infobox" must find a way to insert some internal link (using this template) into the article text? And that is supposed to be easier than the alternative..? --Martynas Patasius (talk) 20:00, 21 October 2011 (UTC)

Hi Martynas, I've addressed some of your concerns below. I would encourage you to give the userscript a try and see what happens when you actually write one of these templates. Pleiotrope (talk) 01:40, 22 October 2011 (UTC)

I'm a prolific writer and editor and I think this is a really, really bad idea. It's going to make it much more difficult for casual editors and experienced editors to edit. That's a lot of parsing to do in the edit screen. Then there's the issue of what you should put in this template. Some editors will add it to almost everything so that the article is full of little orange boxes. Some will spell behavior as "behaviour", making it harder for any script to parse data. We need to devise ways to make editing easier, and while this may make reporting easier (I'm not convinced), it will make editing more difficult. Karanacs (talk) 18:05, 20 October 2011 (UTC)

Agree with Karanacs, just above. I add that the concept is very - very! - interesting, but the implementation seems very, very bad. editing is already hard enough, we do not need to ad this. No good having some interesting tools, if there is not enough people to use them (properly) - Nabla (talk) 18:37, 20 October 2011 (UTC)

Hi Karanacs and Nabla, thanks for your input. You're absolutely right- this adds a degree of complexity to an already-difficult editing process. Our current implementation is very much a proof-of-concept, and we were hoping to get suggestions as to how to make it both unobtrusive and useful. One way is to alter the template- the fields and syntax are not set in stone by any means! I also would imagine that guidelines could be agreed on that established the use of these templates to avoid being disruptive. As precedent, the <ref></ref> system adds large blocks of complicated markup into the article, but we as editors work with it because it's necessary and useful. While our template isn't on the same level of importance, it's also not nearly as complicated (and could be made much less so). Finally, I think there's a not-unreasonable tradeoff in short-term complexity for long-term benefit: while it does make the source more obtuse, imagine the benefits it could bring to even more tedious tasks like manual list maintenance and categorization. Would you be able to think of possible changes that would make this something you could use? Best, Pleiotrope (talk) 19:17, 20 October 2011 (UTC)

I think this concept works much better in a format such as the persondata template - we have a defined topic (biographies) with defined fields about which we'd like to have more information. This is then placed at the bottom of the page, so it does not clutter up the editing screen, and it can be used in scripts. I don't see any good coming from a system that doesn't have a hard-and-fast definition of what types of keywords should be used. Karanacs (talk) 20:29, 20 October 2011 (UTC)

Agreed. Without a pre-defined ontology I can't see this being of any use at all.

Fatuorum

20:45, 20 October 2011 (UTC)

I understand where you're both coming from, though I would posit that Wikipedia thrives because of the lack of hard-and-fast definitions around a lot of its content. Editors are currently already under the onus of making sure that, for example, articles they create don't already exist, and words they make wikilinks have articles backing them (or, if not, they're okay with having them be redlinks). In a hypothetical future Wikipedia where people use this template, I think there'd be the same type of practice around the template parameters. If an editor makes a template that uses some arbitrary relationship, it's easy enough to change it to whatever may be already established. If an editor was to add them everywhere, another editor could remove the ones that would be inappropriate (as per the bold, revert, discuss policy). Additionally, if the use of this template becomes fairly common, there's nothing preventing the establishment of a preferred ontology, especially within specific WikiProjects. This, like most other parts of Wikipedia, would be a matter of consensus, don't you think? --Pleiotrope (talk) 21:11, 20 October 2011 (UTC)

I'm not a great fan of consensus in areas such as this (and many others as it happens, but that's another story). Karanacs example of PERSONDATA is a good one.

Fatuorum

21:15, 20 October 2011 (UTC)

Would you care to elaborate on why you are opposed to enabling consensus formation on this issue and how this is different from, for example, creating and applying categories or naming articles in the existing system? Benjamin Good (talk) 21:31, 20 October 2011 (UTC)

Because metadata is useless unless its meaning is defined. (We already have guidelines for naming articles, and categories are just a useless distraction that few bother about anyway.) What's the point of embedding metadata if every editor is inventing their own flavour? It's like everybody inventing their own language.

Fatuorum

21:50, 20 October 2011 (UTC)

As long as it is created in good faith, individually-authored metadata can be used to enhance single articles as shown in Pleiotrope's example. As you suggest, when making use of metadata across many articles, for example to compose a list based on a query (e.g., nocturnal animals in north america), it is important that standard forms are used consistently across articles. There are many ways to approach the creation of a standard. One is for some power to dictate the standard. Another is to let it emerge organically (see above discussion with user FG). Within Wikipedia I think the latter makes the most sense. Benjamin Good (talk) 23:12, 20 October 2011 (UTC)

Good faith has nothing to do with anything in this context, except for leading to the construction of a pointless Tower of Babel.

Fatuorum

23:53, 20 October 2011 (UTC)

I think its also important to keep in mind that this is in no way intended to be a replacement for infoboxes like persondata. When there is an established desire for a particular set of attributes to be displayed on a large number of pages, an infobox template is an excellent mechanism and could replace the corresponding SWLs on relevant articles. Benjamin Good (talk) 21:45, 20 October 2011 (UTC)

Pleiotrope, adding more specialised templates is bad, the current situation is already way overboard, I think, and editing the wiki is already very much un-wiki (not fast, not friendly, not 'for all'), much because of the proliferation of specialised templates, de facto enlarging the wiki-markup to thousands(?) of keywords. If ever, this would better be kept out of the main text - like the already mentioned persondata (it does not look that great too, but at least a casual editor may ignore easily) or categories (aren't categories somewhat similar? how about ideas to improve - eventually replace - categories instead of adding a feature?) - Nabla (talk) 22:01, 20 October 2011 (UTC)

Nabla, I'm not sure I agree that adding more templates is bad. Templates are an accepted method of controlling the complexity of Wikipedia articles. I have seen new templates be created and used on pages without ill effect- I hypothesize that inexperienced editors simply ignore them.

One thing that I want to point out is that we've been discussing mostly predictions: that these templates will be useless without an established ontology, that they will add too much clutter to source text and discourage editors. While your (and Malleus', and Karanacs') reservations are totally valid, don't you think we should actually test our respective predictions? It's perfectly possible that these concerns won't come to pass. If this works, we've started building a valuable tool to categorize and annotate the huge amount of information we've created here. If it doesn't, the templates are gradually removed and no harm is done. This encyclopedia is not made out of delicate parchment, we can and should innovate and experiment.

Would you all be totally opposed to a small, limited test of these templates on a selected group of pages? We could then actually get some data for our predictions by seeing how they're received on live articles by editors. --Pleiotrope (talk) 23:46, 20 October 2011 (UTC)

Without an established ontology it's more of a foregone conclusion than a prediction.

Fatuorum

23:56, 20 October 2011 (UTC)

Yes, I'm opposed to a small, limited test of these templates. There's no consensus that these would be useful and no methods to neutralize any of the concerns. I'm a computer programmer; a good portion of my job is enabling people to enter and report on data. The number one lesson that I teach all of my new employees is that one of the most important requirements to define for any project is "what type of reporting will be necessary?" Without specifics on that, whatever you design will ultimately fail, because if the planning doesn't take place up front it will end up being impossible to get the reports you want. If we don't know up front the type of metadata that we want to collect, then how on earth can the metadata be useful? If we don't know what it is being used for, why would we want to put it in an article? If we don't know whether it is useful to anyone, why clutter the editing screen and make editing harder than it already is? There must be a way to measure effectiveness for something like this, and I don't see one yet. Karanacs (talk) 14:40, 21 October 2011 (UTC)

How would you measure the effectiveness of hyperlinks? Benjamin Good (talk) 17:10, 21 October 2011 (UTC)

I don't see any clear purpose to these templates. We've had PERSONDATA for years; does any program actually use that data? I don't think I've ever noticed it being used anywhere. Does any program use the geo coordinate data on location articles? The idea of a generic metadata class to be used anywhere without a clear schedule is destined to fail (and to deter a lot of editors along the way). I'd rather start with specific examples of use cases before starting any "limited test". Otherwise we have no idea what's being tested.

The pony case seems worthless to me, for example. What does it mean that a certain relationship is behavior-based? What kind of a program would "mine" that information? How would it be better than just having a database of animal characteristics on some other site, rather than mixing it with natural language? And as others pointed out, you're losing all kinds of qualifiers. Presumably people would add these tags en masse with little regard for subtlety. Bad information is worse than no information. The other example was that one article is the capital of a different article; wouldn't both articles already link to each other in the infobox for that reason? And, again, wouldn't someone looking for this information turn to public databases rather than English-language articles? —Designate (talk) 20:15, 21 October 2011 (UTC)

Hi Designate, yeah, there are actually a number of really useful tools that use metadata in Wikipedia- check out freebase.com and dbpedia.org. But these are limited in their parsing because they are limited to preexisting structured content. For context, I can attest that the use of this generic metadata creates "noisy", not bad, information. Noisy information can be filtered and cleaned easily. As far as the use cases for these template, I was attempting to show how an editor could create a rapid summary of the useful information in the article, which could then be used to categorize the pages, provide candidates for additional fields in the article's infobox, or simply highlight key points at minimum. My example may not have been the best, but if you use your imagination, you can imagine better ones. I've addressed the issue of people adding these templates en masse above; my point was that people should show the same restraint as they do in any other task on WP and appropriate "tagging" policies could be agreed upon. In regards to an often-brought-up point that it would deter casual editors, I haven't seen any evidence of this and think that may be an unfair implication of casual editor's intelligences. Pleiotrope (talk) 01:40, 22 October 2011 (UTC)

Re: Karanacs: Hello fellow programmer! I'm currently working in bioinformatics as a research programmer, and I agree that project planning is very important. However, I would point out that nobody was planning the structure or organization of Wikipedia when it began, but it organically developed tremendous amounts of organization and information regardless, and people are able to use it today (both casual readers and groups like the links I posted above). The same idea is behind our resistance to pre-defining an ontology: we could never design one that could be as appropriate as one that grew naturally through trial and consensus. Pleiotrope (talk) 01:40, 22 October 2011 (UTC)

Do we really need a test to find out if "{{SWL|type=behavior|target=nocturnal}}" is going to be more confusing than "[[nocturnal]]"..?

And the problem with "I would posit that Wikipedia thrives because of the lack of hard-and-fast definitions around a lot of its content" is that in this case we do not get an equivalent of Wikipedia - we get an equivalent of Wikipedia with no policies, no guidelines, no essays and no talk pages. And we know it wouldn't work like that.

For example, what if I want to find out what "type=behavior" is supposed to mean? How would I find some documentation? How would that documentation look like? That is what you have to show, and not just a fun toy that does what we can already do in another way (that might also be easier).

Also, I'd like to repeat my proposal to stop trying to implement semantic links (and the like) in Wikipedia and to start a sister project dedicated just for that. Then, for example, you might be able to make a namespace for those "types", implement redirects (for example, between "behaviour" and "behavior"), add documentation there, make "articles" that consist only of that "link metadata" (reducing the confusion)... Wouldn't that be a much better option for everyone? --Martynas Patasius (talk) 20:55, 21 October 2011 (UTC)

There have been many attempts to build other projects outside of Wikipedia with the goal of creating open, structured knowledge bases. See for example, Freebase. The trouble with all such examples so far is that none of them have managed to come even vaguely close to the level of editing activity that Wikipedia has. It is worth attempting to bring this kind of functionality (some level of structured data) directly into Wikipedia because this is where the people that really care about sharing the worlds knowledge openly are actually working. If the Wikipedia community had a better knowledge authoring platform to work with we simply wouldn't be having this discussion. The ability to clarify the meaning of the relationships between concepts in a way that could be used within the system to enable queries, improve navigation, improve search, and automate mundane tasks such as list creation would simply be built in. But, for various unknown reasons, we don't. We can wait until another better platform magically appears (I don't see it coming soon), we can move such efforts like this away from this wonderful community (and likely towards corporate concerns) or we can work right here with what we have. Benjamin Good (talk) 21:50, 21 October 2011 (UTC)

So, in other words, you want to make Wikipedia into a knowledge base, because that would be good for that field of study..? And you hope that if this field of study will develop significantly, Wikipedia will also get some benefit (but you do not really know or imagine how exactly that will happen)? --Martynas Patasius (talk) 23:30, 21 October 2011 (UTC)

What? Could you clarify this comment, please? Pleiotrope (talk) 01:40, 22 October 2011 (UTC)

Well, I guess I could, but what exactly should I clarify? Anyway, I'll try again: do I understand correctly that the supporters of this change want to implement semantic links on Wikipedia because that would help to develop the field of study that concerns those semantic links? And that they hope that those developments, in turn, would be beneficial to Wikipedia (in some way)? --Martynas Patasius (talk) 16:25, 22 October 2011 (UTC)

We are not trying to "develop a field of study", we are honestly trying to make Wikipedia a more useful place to capture and share knowledge. I assume the 'field of study' you speak of is the

semantic web and, more specifically, semantic wikis. These fields are actually pretty well established and the point of this exercise is not to extend that field of research. We simply hope to make use of their progress in the context currently offered by Wikipedia. To see a long list of examples where semantic relationships are freely created by wiki editors and used to great effect, have a look at this list of semantic media wiki deployments. By the way, thanks for all the hard work and thought you've put into this discussion. Though we clearly don't agree, I appreciate your enthusiasm. Benjamin Good (talk

) 17:21, 24 October 2011 (UTC)

The userscript is not actually a toy. If you had installed it and given it a try, which you weren't obligated to do, of course, you would have seen that clicking on the relationship in the "infobox" would have directed you to a page regarding the meaning of that relationship. If you had actually tried writing an SWL template in the sandbox article, you would also have seen that your previous examples regarding the pony and behavior fields would have resulted in malformed wikitext and would have been inappropriate. But since you didn't, and are playing the part of a user without the userscript installed and curious as to the use of the SWL template, I would imagine that you could look on the template's documentation (at Template:SWL/doc, like every other template), and discuss it on the article's talk page. This is the established procedure for everything else on Wikipedia. Pleiotrope (talk) 01:40, 22 October 2011 (UTC)

Sorry, but it is a toy. A nice toy that is fun to play with and must have been fun to create, but still just a toy and not a serious tool, nor a prototype of one. For if we really wanted to create custom infoboxes, we would do what we do with succession boxes (you know, a template for the start of the box, a template for the end, templates for the records - Template:s-start, Template:s-end etc.). I don't see in what respect that would be worse than this "fake infobox" (unless, of course, you count using semantic links as a good thing and not a bad one).

Looks like another proposed use of semantic links was generating lists automatically. Now that has actually been tried in Lithuanian Wikipedia: parts of articles about days and years that concerned births and death were generated from biographic articles (lt:Vikiprojektas:Gimtadieniai ir mirtys, lt:Vikiprojektas:Neaprašytos asmenybės). The result was a complete mess. I guess that only the initiator himself (and his sockpuppets) could actually use the whole system of biographic articles, day articles, year articles - and lists of birth and death records for persons without articles about them...

Now about "previous examples regarding the pony and behavior fields". Well, now I know that you expect that internal links will not be created just so that semantic links could be added. But still, there should be some draft of a guideline that would actually say so - and I think that you should start from it and not from some template or user script.

Also, concerning one more "If you had actually tried [...]"... I see that the type "behavior" should be described and documented in Category:SWL/behavior, right? I guess that in such case its description could be discussed on the category's talk page... But, unfortunately, no description has been given. Furthermore, the category hasn't been added to Category:SWL like some others. And also, "if any of us had actually looked at it", we would have seen several mainspace articles in that category... In short, the trial has been started some time ago. And the users who are the most enthusiastic and knowledgeable about semantic links participated in it. And - yes, the system is a mess already... So, do we need another trial, or can we accept the results of this one..? Oh, and I think we should actually end it and remove the template from mainspace articles. --Martynas Patasius (talk) 17:27, 22 October 2011 (UTC)

I don't like the overall cost/benefit ratio. There is some benefit for readers (though I'm not convinced it's that big a deal) and there's a lot of potential for benefits for algorithms trying to make sense of the data which in turn may help readers. But the costs are way too high. The biggest problem is the added markup. Moving from

It is a strictly [[nocturnal]] [[owl]], which feeds on small mammals and birds.

to

It is a strictly {{SWL|type=behavior|target=nocturnal}} [[owl]], which feeds on small mammals and birds.

is a significant problem for new editors. True, the ref tags are also a big problem but we absolutely need mechanisms for referencing whereas this proposed experiment is not essential. I'm also pessimistic about an orderly introduction that doesn't involve megabytes of discussion about the preferred ontologies and widespread confusion. Editing time is a limited resource and I believe it would be better spent elsewhere. Pichpich (talk) 16:07, 25 October 2011 (UTC)

At this stage, especially with in-line refs already being a huge nuisance, I would also agree that the cost/benefit ratio is too high. And, as small potatoes as this sounds, I also dislike the visual clutter of the orange underlines. --Hobbes Goodyear (talk) 12:07, 26 October 2011 (UTC)

In the interest of drumming up some comments, there's a protected edit request at Template talk:Db-meta#First sentence should be italicized completely. – Luna Santin (talk) 20:35, 26 October 2011 (UTC)

De-archived for more feedback.

books

} 07:31, 26 October 2011 (UTC)

I'm browsing the wiki and it seems all references I see to

A couple I just checked in the LinkSearch list did not seem helpful. Johnuniq (talk) 06:53, 14 October 2011 (UTC)

Ditto for answers.com, answers.yahoo.com, wolframalpha.com, metafilter.com, vark.com, nndb.com and similar sites. Talk and other pages are OK but no way are these reliable sources in an article. Do need to be able to include the main website link in the article about the site. ---— Gadget850 (Ed) ^talk 11:51, 14 October 2011 (UTC)

Speaking technically, is there a user right that allows its holder to add a link that's on a blacklist? Of course, an admin could remove the link from the blacklist, add it to the page, and restore it to the blacklist, but that definitely doesn't seem ideal. If so, it would seem to me that we could simply blacklist these pages, add comments to the articles about these sites noting that their links are blacklisted, and add a note to the relevant warning (i.e. the one you get if you try to add a link that's blacklisted) instructing users to post at WP:AN or whatever the relevant page would be if they have good reasons for including their links. Nyttend (talk) 21:32, 14 October 2011 (UTC)

De-archived for more feedback.

Pardon the potentially unwanted metacommentary, but is there a reason that the blacklist still exists as a separate entity to the edit filter at this time? It seems as though the edit filter has all the features needed for it to handle blacklist content (i.e. match a regular expression and prevent an edit containing it). FWIW I'm in full agreement that the UGC sites in question (yours and Gadget's) should be filtered from mainspace. Chris Cunningham (user:thumperward) - talk 12:34, 26 October 2011 (UTC)

One problem with moving the blacklist entries to the edit filter is that the edit filter already hits the condition limit regularly, which lets things slip through that shouldn't. The more stuff that's added to the filter, the more often this limit is hit. 28bytes (talk) 15:40, 26 October 2011 (UTC)

I'd prefer blacklisting it, but the edit filter would be better than nothing.

talk

) 13:47, 28 October 2011 (UTC)

• Neutral I can imagine a "google knol" article being a courtesy copy of a real published paper, so the black list may be inappropriate. However, I quite agree that almost all uses are either of "articles" completely unusable, or being suggested in violation of an ArbCom rulling (CH), I don't have a definite opinion. — Arthur Rubin (talk) 22:14, 29 October 2011 (UTC)
• Support as an edit filter. If some slip through so be it, LinkSearch will allow patrollers to catch usages. Don't blacklist it, as there are some legitimate uses (
  WP:SELFPUB). — Train2104 (talk • contribs • count
  ) 13:47, 31 October 2011 (UTC)
• Neutral The externals to it I've seen have been rubbish or spam for a blocked user and it could never be counted as a reliable source but I'm unsure about this idea of actually blacklisting it.
  Dmcq (talk
  ) 17:31, 31 October 2011 (UTC)
• Wary of blacklisting anything that isn't actually a) being actively spammed by it's owner or b) demonstrably useless or c) both. Rich Farmbrough, 00:25, 1 November 2011 (UTC).
  

Recently (very recently) i was having a discussion which was archived way too quickly. I noticed WP:ARCHIVE doesn't give any examples of misuse of archiving. I propose we add a certain time span before any manual archives are done and to allow reverting certain archives.

It would be hard to establish a minimum time before archiving that would work across all talk pages. An appropriate archiving delay on an obscure article talk page that can go months without comments may be 3 months, while the talk page of an article about a controversial news event, getting 50 comments a day on a variety of topics may justify very fast archiving with a delay of a week or less (sometimes 2 days). (the problem is when the topic cools down, the fast archiving may not be slowed down) Other then to call on editors to use good judgment, I don't know how one would word specific guidelines. Monty₈₄₅ 18:31, 28 October 2011 (UTC)

wp:CREEP says no. Looking at the actual event it just seems to be a new user not understanding how our archiving works (and thinking the situation was resolved) and a somewhat unlucky revert from User:C.Fred. Just deal with the situation (look, I did!), no need to make up new rules for it. Or does this happen to you daily? Yoenit (talk

) 19:46, 28 October 2011 (UTC)

Not exactly what i'm trying to say. Archiving can "easily" become vandalism. For articles that have 50 post a day, i can definitely understand as it gets so big it'll be difficult to handle. Anyways.....i still believe manual archiving should have more control. I was also thinking this would help control more situations where manual archiving isn't done properly. Anyways relying it on judgement alone is a bit difficult.

talk

) 19:04, 31 October 2011 (UTC)

This is not a proposal so much as I would like to get the village pump opinion, and ensure compliance with Wikipedia:Bot_policy#Mass_page_creation which indicates the pump should be notified. Dr. Blofeld is mass creating literally hundreds of pages (5-7 per minute) of a stub of every river in germany, with no content except for a link to the german article saying it could be translated over. I am not super invested in the result, but it seems like a violation of policy, and almost spam to be creating hundreds of articles. If the consensus says its fine, I will go away to my corner :) Gaijin42 (talk) 20:29, 28 October 2011 (UTC)

stop spamming this everywhere, it is disruptive. Yoenit (talk) 20:31, 28 October 2011 (UTC)

Given a discussion at Wikipedia:Administrators' noticeboard/Incidents#Clarification needed, I am about to embark on posting a number of proposed tasks for Δ. There is no attempt on my part to overwhelm this board with such requests. I understand some of you may feel this way. Whatever the case, I would appreciate it if meta discussion about the appropriateness of these requests is kept in this thread, rather than in each specific task request thread. Please restrict your comments on those threads to those tasks only. Thank you, --Hammersoft (talk) 17:28, 24 October 2011 (UTC)

I support them all most of them, but I am not clear where do you want us to say so :) --_{Piotr Konieczny aka Prokonsul Piotrus| talk to me} 18:31, 24 October 2011 (UTC)

• You don't need to. The restrictions are to query if there is opposition, and if so to allow for consensus to develop before undertaking further edits of the kind. --Hammersoft (talk) 18:32, 24 October 2011 (UTC)

• Oppose all - there is no evidence of any change in the approach that gave rise to Beta's automated editing restrictions. The last thing we need is an end run around those restrictions via a flock of new proposals for automated editing. It would be wonderful if beta would overcome all of the hang-ups that landed us here, but for now he apparently has not. - Wikidemon (talk) 08:07, 25 October 2011 (UTC)
• Oppose all for now. Delta needs to be answering questions about what he's planning and the proposals shouldn't be open-ended but rather a specific set of edits he's planning on making. As such this should _really_ be Delta making the proposals. Having a 3rd party do it is a recipe for disaster as the "telephone" game just gets worse. What does _Delta_ mean by being able to prod articles as a "pattern?" Does he think that gives permission to prod (or for a different proposal) redirect 100s of articles at a time? I think this whole section misunderstands the point of the editing restrictions. I agree with JohnFromPinckney's comments and he says it a lot better than I could. Hobit (talk) 10:38, 25 October 2011 (UTC)
  • In any case, I think any "mass" permissions would have to come with the restriction of actually using meaningful edit summaries. Hobit (talk) 10:43, 25 October 2011 (UTC)
• opppose all per my comment on #10, and honestly this entire process seems rather to be rather bad faith, pointy and almost a little
  Crossmr (talk
  ) 11:38, 25 October 2011 (UTC)

• Ah, so he's required to make these requests but if they're made in accordance with requirements, it's bad faith, pointing, and gamey? If the requests aren't made, you slam him. If the requests are made, you slam him. I've known for a long time that you would oppose anything having to do with Δ editing here, but you're really jumped the shark now. --Hammersoft (talk) 13:04, 25 October 2011 (UTC)

• Actually I'm slamming you for being disruptive at this point. 20 proposals? Are you serious? This is beyond
  Crossmr (talk
  ) 15:28, 25 October 2011 (UTC)

• This is exactly what all the past restrictions and ANI threads have reduced this process down to - getting the community to approve every single action that Delta is allowed to take, one action at a time. And that's because of editors that aren't giving any good faith on Delta's ongoing attempts to improve his communication and stick to beneficial edits and instead focus on numbers and editing rates and anything that may end in a permaban from the project. Editors want Delta to be operating under a microscope, so to prevent him from doing anything undocumented, we better be getting a writeoff on the community for every single step in the process.
• This is what all the wikilawyering to try to bust Delta has come to. --MASEM (t) 13:39, 26 October 2011 (UTC)

• Because good faith is not a blind shield or an everfull well that one can go to again and again. Delta spent his good faith a long time and needs to earn it back. That's why he's under such heavy restrictions, and why people won't let him out from under the microscope. Because it never ends. It was only 4 months ago or so we had the whole NFCC drama due to his edit warring (technical or non-technical), poor communication with a little uncivil behaviour on the side. That is a problem that is still present despite being let back in and given that nth last chance a couple years ago.--
  Crossmr (talk
  ) 23:03, 26 October 2011 (UTC)

• That itself, I can't disagree with. The problem is the issue of complaining about what this specific process (Hammersoft proposing what tasks Delta is doing) is trying to accomplish. It's defining the stage that Delta should be operating on. And yet you're complaining that this process is "pointy" despite the fact its trying to establish bounds. The process being used is almost exactly the same as for many bots that operate indefinitely, fixing as corrections are needed, as opposed to a "task". There's nothing wrong with that. Yes, I understand that many don't have good faith in Delta anymore, but its another thing to throw out any responsible presumption of someone trying to work within those bounds within good faith. If you can't support that, then, as Hammersoft has often said, you might as well start writing the proposal to permaban Delta from the project. If instead you think Delta can improve - even if there's a very long and hard road to reach that improvement, and that ban isn't needed, then let some good faith work here to let the process as already outlined by his restrictions work through instead of trying to stymie it. --MASEM (t) 23:20, 26 October 2011 (UTC)

• As I've clarified multiple times, the issue is with how he's done it and it's not remotely in the spirit of Delta's current restrictions. Hence why I've taken issue with it and outlined the various guidelines and policies I think this entire process violates, and I'm not alone in that assessment. The point of this restriction was for specific focused tasks, not unlimited exemptions to his restrictions. The issue was further pushed when it was clear there was not really any support for the vast majority of his initial proposals (the only that gained it was a task bots already do) and then going back and putting in a whack more. That reaks of
  Crossmr (talk
  ) 10:16, 27 October 2011 (UTC)

• There is no difference between a task that affects a finite number of pages and an ongoing task that potentially can affect all pages on a recurring basis; the issue is that both are "patterns of edits affecting more than 25 pages". There was nothing about having focused specific tasks; the restrictions are there to prevent Delta from using rapid-fire bots or user scripts like AWB to perform changes that he wasn't correcting.
• Now, I will say that there is fair concern that you're worried that its an omnibus of tasks that are being proposed inside of one at a time. But that's why there is a push to get a proper edit summary and a page to outline what exactly has been approved for Delta to do, as well as what I suggested, a trial period with deep review to verify he's doing the tasks he's approved for. And just because a bot can do a task doesn't mean Delta's prevented from doing that; there's no restriction on his or any editor from doing bot work on their own - if you feel that needs to be in place for Delta, then you need to go ahead and start proposing yet another restriction on his editing. Instead, to me, I'm taking the fact that the tasks approved for Delta below are ones that since a bot can do it, that Delta's changes have little chance of harm even while throttled or the like - in other words, the community has no problem if Delta's wikignoming are simple to correct and impossible to screw-up tasks even though he's still required to manually check those edits.
• Remember, that the community restrictions are indefinite - that means that unless you want to ban Delta now from the project, you have to give him the opportunity to work within them to demonstrate that he's capable of doing that. This process is part of that. --MASEM (t) 13:01, 27 October 2011 (UTC)

• They're indefinite but they're not permanent. If Delta can ever get it together for a long period of time he could probably start to have them removed. There is a difference between edits bots do and edits bots tend to do very quickly, like dating maintenance tags. These days I find it pretty uncommon to come across undated maintenance tags. As for the specific focus, we've heard from the person who drafted that restriction and he stated that that was indeed the purpose of that restriction, and some of those opposing seem to feel the same. And again this entire process was born of bad faith, VPP was brought up and discussed last may/june over Delta because he undertook a day and halfs worth of edits before that whole blow up and him going ahead and proposing them at VPP. it wasn't in his block log, but it has been out there at least 4 months and no one has been doing the things hammersoft and dirk claim were going to happen. [4]--
  Crossmr (talk
  ) 22:59, 27 October 2011 (UTC)

• Before someone actually noted that this appeared to be a pattern of edits affecting more than 25 articles, Delta was going along and did 8000 edits like this with minimal drama; people did point out problems that Delta responded to to either stop doing or alter things. And no, this is not to wave away some of the problems that editors like Fram have identified in spot-checking these edits but at the same time, the end result is not a complete failure in the article presentation unlike the edits that led to the original restrictions being placed. The issue here was that this wasn't positively identified as a series of edits until now, and thus led to the current drama. Though if Delta wasn't blocked and he continued these types of edits I'm sure someone would have eventually done so, but at the same time while not on anyone's radar, they aren't tipping the boat or raising any issue that haven't be dealt with easily through communication. Exactly the aim of the original restrictions. The current problem is due to lack of clarity, and certainly doesn't seem like an intent for Delta to bypass them. --MASEM (t) 17:10, 28 October 2011 (UTC)

• The problem is you're trying to sidestep the issue again. My main point is hammersoft violating:
  Crossmr (talk
  ) 23:17, 28 October 2011 (UTC)

• From the ANI thread that led to this, it is clear that while the "26 edits in 2 years" was an exaggeration, it is clear that editors want Delta to not do more than 25-like edits in some time frame without prior approval, believing that was a pattern of edits the 1st restriction warned about. Whether it was or wasn't is immaterial here: the reason Hammersoft started this was to assume the answer would be yes (since it was far from a small minority that felt it was a pattern), and thus establish the VPR process laid out by that restriction. Basically, looking two steps ahead of where the ANI discussion was going to establish these tasks that Delta would work under. I cannot see how you're taking that in bad faith; yes, I realize Hammersoft is not soft-spoken, but I see nothing but trying to help Delta gain the good will of the community - even that community that is making sure Delta follows to the exact letter of the law - by making sure Delta's following the set process. It is far far far from being pointy or in bad faith.
• As for the second point, considering these as blanket exemptions, that's why I introduced the idea of trial periods, but I would have no problem if that further was extended into trial blocks. Maybe a first block of 100 articles, a second of 250, and so on, allowing the community to judge if the edits were all appropriate before asking for a larger block and/or adding a new task to his cleanup aspects. What you're trying to ask for, however, "This will affect X articles" is impossible to define for cleanup tasks since how many articles are actually affected is the size of the entire en.wiki. But again, if we restrict him to small blocks and give him more allowance as he shows the competency that the community is expecting of him, he can still be the productive wikignome he wants to be but within the limitations of the restrictions. --MASEM (t) 23:44, 28 October 2011 (UTC)

• Yes, it is an exaggeration, but let's not mince words, it was repeated several times by the both of them and it's clearly a bad faith assumption which has no basis in reality. The VPR restriction as clarified is supposed to be targeting specific tasks. Not open ended exemptions as these indicate. If Hammersoft wants to help delta he should be helping him work within the restrictions laid out and not try and find a way to wriggle out of them. Disruptive side of this became obvious when despite there being a dozen or so proposals here already that didn't have any support, he went and added several more with an ominous message of "more to come". It doesn't need to target X articles. It could target articles in a specific project, category, etc. It could be limited to run for X edits, or X days. However, Delta needs to come up with that himself and actually show that he understands what is going on, because it is becoming increasingly clear that he doesn't seem to. I've got more, but gotta head out. His insistence below that Delta is not even allowed to edit is beyond absurd.--
  Crossmr (talk
  ) 06:12, 29 October 2011 (UTC)

• Even though you're already involved in that discussion, I've raised this entire proposal business at the AN/I thread on Delta.--
  Crossmr (talk
  ) 15:43, 25 October 2011 (UTC)
• Catch 22 on Wikipedia? Who would've thought... sigh. --Piotr Konieczny aka Prokonsul Piotrus| talk to me 17:55, 25 October 2011 (UTC)

• I asked for and received permission from Δ to operate on his behalf in making these requests. He has helped in me conducting these requests. He's aware of them, and supports them. I am not acting independently. Given that, splitting hairs about who is making the proposals isn't going to gain traction. He is required to make these proposals. They are being made. If you have an alternate proposal (other than him being banned from the site), I'm all ears. --Hammersoft (talk) 16:18, 25 October 2011 (UTC)

• The problem is that you've created a blanket set of generic exemptions when it seems very clear that the intention of the provision in the sanctions is so that Delta can make requests on specific tasks. The provision was not intended to indefinitely reduce the scope and effect of the sanctions, which is fundamentally what these proposals are about, and this is what makes it appear that you're trying to game the system.
  talk
  ) 22:48, 25 October 2011 (UTC)

• So what's the solution then, if not to approve highly specific task descriptions short of getting approval for every single specific edit? The reason we're here now is because the idea of "pattern of edits" in his editing restrictions has been read very broadly, and leads to a grey area of what are patterns and what are not. Because some are taking his most recent edits, regardless if beneficial or not, as a pattern, Hammersoft's gotten Delta to enumerate them all the propose each one per the editing restriction. Ergo, it is not carving out exceptions, it is identifying tasks he is allowed to do within the confines of the editing restrictions. This is exactly what the community was pushing towards. The fact that as I read the input back from the it now that some tasks are acceptable, some not, means this is working within the intent of the editing restrictions. --MASEM (t) 13:45, 26 October 2011 (UTC)

• My reading of the restrictions was that when Delta has a specific task he wants to complete (eg. correcting link order on a specific set of 100 articles), he proposes that action here where it's reviewed and then approved or declined. It's not my reading that one can simply propose 'Delta can change the link order in an infinite number of articles', since that's a blanket reduction in his restrictions - he can now do exactly the things that caused his restrictions in the first place, without being in violation of those restrictions. I simply don't see that as the intent at all. These are strict restrictions, they're not meant to be reduced indefinitely as each of these proposals requests, they're meant to be waived temporarily for the completion of specific tasks.
  talk
  ) 22:01, 26 October 2011 (UTC)

• I don't read it that way, but taking that point into consideration and a point I made below, it may be a very good idea for Delta (or Hammersoft) to define how the articles that Delta plans to work on will be selected so that we know what will be impacted. That said, that gives me another idea to add below in regards to a "test block" for community review so that we can have a limited task be extended to an ongoing task, just like we sometimes do with bots. --MASEM (t) 23:27, 26 October 2011 (UTC)

• Oppose all (overriding my individual votes below) if these exemptions can in any way be interpreted to allow Delta to perform automated or semi-automated edits, or edits that appear automated, in contravention of his current restrictions. Some of the comments below, including by Delta himself, imply that he intends to use scripts to perform these tasks. My reading of the sanctions is that these exemption requests only allow Delta to exceed his 25 article limit. His restriction on automated edits is not within the scope of these requests to release, and any interpretation of these exemptions must strictly acknowledge that Delta may not perform such edits.
  talk
  ) 23:26, 25 October 2011 (UTC)
  • Don't kid yourself. Him and his supporters plainly admit he uses scripts, but that he is not barred from doing that as long as he manually reviews the edits. The difference between fully automated use of scripts and semi-automated use thereof but with occasional lapses is a matter of
    talk
    ) 06:29, 26 October 2011 (UTC)
    • I'm not overly concerned with the verifiability of the restriction so much as the essence of it. I don't think it should be lifted, and people are often quite creative at finding ways to verify something I didn't think could be verified.
      talk
      ) 21:56, 26 October 2011 (UTC)
    • AFAIK, the community has said it is fine for Δ to use supervised scripts - This is in no form a request overriding the speed limit, the civility restriction, or use full automation limit or whichever other restriction. If it is clear that a script runs without him (e.g. that someone asks him to stop on the talkpage, and the edits continue without response) then yes, that is fully blockworthy, as are incivil remarks and too fast editing or other violations of restrictions. It also does not allow other patterns to emerge which are not granted here, it does not allow to continue with patterns which are not granted here. It strictly is, that any pattern of >25 edits is disallowed, except those for which here is consensus that they can proceed.
    • I still think the word 'pattern' is too ambiguous and unclearly defined at the moment - but there clearly is no support in trying to define it better, support is more that it is left to administrator discretion. I am sure we will be here soon (and/or AN/I and/or at Δ's blocklog) because someone will find a new pattern soon enough, but maybe that will go nicely through a fresh thread here. --Dirk Beetstra ^{T C} 14:16, 26 October 2011 (UTC)
• Oppose most. The only exceptions are those that could be done by a bot, there in consensus as to exactly what the bot should do, and (unless fixing something that causes a broken view) be combined with a substantive edit, and with the provision that exactly what task is being done needs to be in the edit summary. Among tasks 1-20 this is 4, 6 (only applying to visible whitespace, meaning not within references, as the examples indicate), 7 (only if the link is checked manually, as opposed to by a script), 8 (with the indicated restrictions), 9, 12 (and I might exempt this from the substantive edit requirement, but the references must be identical), 13, 18, 20 (and I might exempt this from the substantive edit requirement, as well.) Strong oppose 10, 11, 15, and 19, and separately oppose 16 and 17, as not having a reasonable consensus for the action. The example in 19 is not appropriate. — Arthur Rubin (talk) 22:33, 26 October 2011 (UTC)
• Oppose all The restrictions are appropriate given this editor's long history of controversial behavior. (That he has been permitted to edit at all is quite astounding to me.) ElKevbo (talk) 17:30, 1 November 2011 (UTC)

Δ meta discussion

• I have a broad concern with the phrasing of these tasks. They are situations where they are OK, but there are nevertheless exceptions or situations where they should not be applied. If a task is "authorized", that has to mean that exceptions will be recognized too. If a few days from now, people say the task shouldn't be applied in some circumstances, I don't want to see an argument claiming "it's authorized so it can be done". Gimmetoo (talk) 20:15, 24 October 2011 (UTC)
• Likewise, in the other direction, there shouldn't be people using the restrictions as bludgeoning tool because they disagree with Δ, despite approval here, and insist he must comply with their interpretation. There will always be differences of opinion between all editors. --Hammersoft (talk) 20:33, 24 October 2011 (UTC)
• I support kittens and puppies and oppose running out of tea bags. -- fg^TC 22:07, 24 October 2011 (UTC)
• Comment. It's amusing that his friends are petitioning here for trivial tasks, when clearly these are not the ones that have caused the ruckus.
  talk
  ) 04:57, 25 October 2011 (UTC)
• I'm worried about the extremely persistent use of "Cleanup" as an edit summary -- if semi-automated tools are being used, they can be written to provide a more useful summary; if the changes are being made manually, it shouldn't be a problem to explain them. Given the nature of the sanctions, more specific edit summaries might be very helpful. – Luna Santin (talk) 04:58, 25 October 2011 (UTC)
  • Agreed.
    talk
    ) 05:00, 25 October 2011 (UTC)

• Seeking clarification. Apparently, this thread is about a particular user, named Δ. That much I've figured out on my own. I've successfully avoided knowledge of the User:Δ, but since you, Hammersoft, have made his existence and problematic behavior so impossible to ignore, I feel compelled to ask for clarification of this huge thread (around 60K bytes so far).

It seems that Δ is a user so unable to get along with his fellow Wikipedians that he's been to AN/I and ArbCom, blocked and banned, and even needs

his own personal template

so we can even address or refer to him. Apparently, he has a penchant for making long series of edits with uninformative edit summaries; he seems to use a script to help him make many edits in a short space of time, whether the edits are seen as useful or desirable to the WP community or not. Often, I guess, they're not.

Now despite his egregious behavior he has a chance to avoid a complete ban, if (among other things), he gets approval for any pattern of edits to more than 25 pages. It seems an obvious oversight that the editing restrictions don't specify the timespan for the 25-page limit, but I assume that'll get cleared up somehow. But what I don't understand is what you're trying to accomplish here with all of these "Δ Task" threadlets. You seem to be pre-emptively seeking permission for all the kinds of behavior that got Δ into trouble in the first place (except using obscure edit summaries; you don't seem to have addressed that yet). That seems to go against how I see the editing restrictions.

The way I understand it, Δ is allowed to edit (or would be, if he hadn't gotten blocked again), but if he sees a particular set of articles he wants to make similar changes to (the "pattern of edits"), then he needs to then seek specific approval. So, if he says,"I just noticed the articles on Law & Order episodes, and I see that the last four seasons' pages all have X, when they should have Y," then he can come and seek permission to do X-to-Y changes on those articles (still observing the 4 edits/minute in 10 minute restriction when he gets consensus).

Instead, you want to get permission in advance to change X to Y on any article (all articles, in fact), just in case Δ gets the urge to perform that pattern of editing. Do I understand your objective correctly? Because if I do, it seems to be misinterpreting the intention of the editing restrictions and the allowance to avoid the ban. But maybe I'm the one misinterpreting things. — JohnFromPinckney (talk) 09:56, 25 October 2011 (UTC)

• Δ is required to ask permission to make repetitive edits. Since some people are construing that requirement very broadly, the situation now is one in which any edit of Δ's needs to have basis in there being a request here to perform that edit. He's already been doing these edits. So it's past and future. If he's done enough of them that the next one of a type will be some number above 24, some editor will cry foul when he does a similar edit if there's not a specific request here. This is the silliness that we are forced to contend with. Since he's required to make these requests, and he's authorized to make these requests on his behalf, the requests are being made to cover every possible kind of edit that he might do that someone could remotely construe as "pattern". --Hammersoft (talk) 13:02, 25 October 2011 (UTC)

I have suggested on AN/I that when someone construes something as a pattern - whether before the 25 or after the 25, that that person has to bring that to AN or AN/I, and notify Δ. Δ is at that moment stopping with that part which is construed as a pattern, and either awaits whether the construed pattern is not deemed a pattern, or asks for permission (latter wise anyway) - no blocks applied, even if there are already hundreds of edits that fit the pattern (but blocks can be applied when a significant number of edits with that pattern are after the notification). --Dirk Beetstra ^{T C} 13:07, 25 October 2011 (UTC)

No, that leaves the door open for Delta to, for instance, make 2000 rapid edits to articles in what very blatantly constitutes a pattern, have someone complain about it, have the community decide 'yes that's a pattern', then move on to another 2000 rapid edits different enough that he can claim they're not a continuation of the same pattern. Part of the problem with Delta's editing, from what I can see, is that he makes questionable edits in such numbers that it's almost impossible for someone to check his work afterwards, or worse to clean up a mess that is caused as a result. My reading of the sanctions is that the '25 articles' figure is a throttle, to slow him down so that others can review his edits more easily, and that if he wants to do more than that he needs prior approval so that people can review his intended changes beforehand. Giving him free reign only until such time as someone complains is effectively lifting the sanctions altogether, something I doubt will gain much traction.

talk

) 23:34, 25 October 2011 (UTC)

Under his editing speed restriction, (40 edits in 10 minutes average), 2000 edits would take him 500 minutes - a bit more than 8 hr - to complete if he stayed within that speed. Someone will notice that first. So he has a throttle already. The idea of the original 25 article piece was to prevent Delta from instituting a common set of changes that did not have community approval - eg, the equivalent of running an unauthorized bot without having gone through the bot request process. The problem is that there is a grey line between a pattern of edits and making similar edits in several articles. What one may consider a pattern will not be by another. Therefore, if Delta does start doing something that he himself doesn't consider a pattern but someone else does, he needs to be alerted to it and discussion needs to ensue to determine if that is a pattern of edits. Since he's already pointed to VPP to request patterns of edits, it makes the most sense that the discussion take place there. But importantly, as soon as someone says to him directly (talk page) that they think it is a pattern and they're opening up discussion on it, Delta should be expected to stop that specific action. --MASEM (t) 13:29, 26 October 2011 (UTC)

I've been watching this set of proposals grow and am horrified. It now looks as if directed by a council of

No one editor should have special rights. If edits are good keep them. If intentions are good but edits aren't: undo and get over it. If intentions are bad: undo, warn, warn again, block. All these policies and/or guidelines are in place.

A continuation of discussing any change in policy or guidelines should be done without basing (more like debasing) them around one editor. The whole pattern of 25 edits business is also in place for seemingly good reason. Why should any one editor get a free pass? I hope this set of proposals is considered unacceptable under some other policy or guideline because they are outrageous. -- fg^TC 06:05, 26 October 2011 (UTC)

Blocking doesn't work when someone has a vocal group of supporters endlessly arguing for unblock at ANI because they see the restrictions as unfair to begin with.

talk

) 06:32, 26 October 2011 (UTC)

Unfair? Who said that? I think I would use the word 'ambiguous' to describe my concerns. --Dirk Beetstra ^{T C} 14:03, 26 October 2011 (UTC)

One question has come up repeatedly: should Δ be performing edits which can be handled equivalently by current bots? Several of the proposals below have objections relating to this question. Some recent discussion at #Δ proposed task #18. – Luna Santin (talk) 23:16, 26 October 2011 (UTC)

And since that question is likely to need its own sub-thread, here's another question: what about reverts? If someone (i.e. a regular editor of the article) disagrees with a set of changes, especially those approved only as accompanying some other substantive changes, what do they do? Is the onus on them to sort out only those changes they disagree with from among an a large set of separate changes in the diff, preserving the neutral changes and the one good bit of visible change? Or can they revert all of it? Can the whole thing be re-reverted as "community approved task, discuss specific problems on talk page". This will come up sooner or later. Not to obscure your own question, which does need separate consideration. Franamax (talk) 02:14, 27 October 2011 (UTC)

• Well, you have to admit it's an insanely tangled mess we've got ourselves in, don't you think? The locus of all of this is Δ, is it not? So let's take the easy road; let's just ban Δ and be done with it. No more debate, no more worries, no more endless threads. Nice, neat, convenient. --Hammersoft (talk) 12:47, 27 October 2011 (UTC)

No, the locus is Beta's unrelenting desire to make code-assisted edits that cause the wiki to conform to their own idea of what it should be, as fast as possible. Is and always has been. A deficient style of communication only becomes a factor because the inevitable errors, times the fast pace, result in multiple and prolonged questions, which is where Beta inevitably snaps. Franamax (talk) 21:43, 27 October 2011 (UTC)

Actually I rarely rarely ever snap, if a user comes to my talk page to raise an issue. I only get short with someone after repeated cases where the person does not listen, will not read what I ask them to read and makes continuous personal attacks/and or insults against me. Anyone will snap if pushed too far. I do not snap unless pushed extremely far. ΔT _{^{The only constant}} 02:08, 28 October 2011 (UTC)

• Which of course is all completely Δ's fault. So again, why not just ban him from the project and be done with it? --Hammersoft (talk) 02:04, 28 October 2011 (UTC)

Manual vs. automatic

Δ claims that all 8000 or so cleanup edits were checked before saving, and that they were not made in a bot-like fashion (i.e. not simply pushing "save" without checking). Anyone who has seen the speed of these edits, and the huge diffs they generate, may have doubts about this, but in the end the only thing we have to check is the actual result of these edits. At the end of september, I checked a number of these cleanup edits, and fould quite a few errors in them.[5][6] These were mostly corrected in the script afterwards, but suggest to me that these edits weren't really checked manually. Looking at a small random sample of other cleanup edits show that e.g. here a bare url is given the description "Yasni-Ergebnis für http://northtexan.unt.edu/content/nathan-kelly". The fact that this is in German is due (probably) to some user preference of Δ, when I check the same link the title it is in Dutch, and for most of you it would be in English. Manual checking and editing would have seen and corrected this. One can also wonder whether bare links aren't better than promotional title pages like (from the same edit) "Books, Discount Books, Cheap Books, Australian Bookstore - Emporium Books". Perhaps something to take into account at proposed task 13? In this diff, for some unknown reason a cite web template is replaced by a bare link plus title, resulting in the loss of the accessdate and the gain of a promotional title, changing ^ See "photoquotations.com". Retrieved December 29, 2010. to ^ See photoquotations.com ⁄ world's largest photo quotation resource. Not really an improvement... This is an example of adding titles to bare links, even though titles are already present. A bot can't detect this, a human should (e.g. it changes .<ref>[http://www.uuchurch.org/history/middle-years]“The Middle Years,” (March 2011).</ref> . to <ref>[http://www.uuchurch.org/history/middle-years The Middle Years | University Unitarian Church]“The Middle Years,” (March 2011).</ref> ).

Six edits checked, three with problems. All in all, these are not the signs of thorough human checks, and make all the comments of 8000 edits without problems, all checked by a human, a bit less convincing.

• So apparently there's a retrieval of what the page title is? That could be a definite problem if no verification is happening. But, it would be very hard to review and see what pages were viewed and which ones took the title directly, without modification. Without that, hard to ascertain an error rate. But, like I said, taking the bare title without review is a problem. By the way, the one edit of yours I checked (your posting in this section) contained an error. I don't think "fould" is word. Hmm. 1 edit checked, 1 error found. 100% failure rate? Wow. I'm sure there's no problem with my sample size, or at least no problem any different than your sample size :) (humor please, humor). --Hammersoft (talk) 17:41, 26 October 2011 (UTC)
  • No problem with humour, but errors in typing are normal (and I check my talk page posts less than my article edits), errors in scripts get repeated over many edits and pages. That doesn't mean that the occasional error shouldn't be expected, but the error rate should be very low. It isn't normal that I could identify many different types of errors when I checked a small number of cleanup edits at the end of September. Most of these errors shouldn't have been found when testing the script, and if not then when checking the affected pages.
    Fram (talk
    ) 19:42, 26 October 2011 (UTC)
• By the way; humans are prone to errors. Where do we set the bar for Δ and how do we evaluate performance? --Hammersoft (talk) 17:43, 26 October 2011 (UTC)
  • These issues don't exist in a vaccuum, one can't argue in one breath that Delta is experienced and careful enough to exceed his restrictions in 20 broad cases, and in the next breath say that he's so prone to mistake that "Books, Discount Books, Cheap Books, Australian Bookstore - Emporium Books" completely slipped his attention on a carefully and manually reviewed edit, as is required of him by his editing restrictions. Sure, humans make mistakes. But most humans haven't made so many mistakes that they're under strict editing restrictions. If Delta can't pick up his game and make error-free edits even when he's compelled to do so by mandatory, careful, manual review, what faith can the community have that he'll be able to apply a similar level of quality control if his restrictions are ever lifted?
    talk
    ) 22:13, 26 October 2011 (UTC)

• Nobody can make error free edits in perpetuity. By the way, you spelled vacuum wrong. Can't you pick up your game? --Hammersoft (talk) 12:50, 27 October 2011 (UTC)

• I certainly can. But then, I'm not under editing restrictions. Delta doesn't need to make error free edits in perpetuity, he needs to make error free edits until such time as he proves that his volume (not frequency) of errors is in line with community expectations.
  talk
  ) 22:19, 27 October 2011 (UTC)

• Right, in perpetuity. You do know the history here, yes? By the way, you're perfect? Really? --Hammersoft (talk) 02:05, 28 October 2011 (UTC)

• "Can't you pick up your game?" "I certainly can." doesn't strike me as overly confusing, Hammersoft.
  talk
  ) 03:00, 28 October 2011 (UTC)

• The humorous point is that you're insisting Δ pick up his game and be perfect, yet you can't spell vacuum correctly. Contrary to popular belief, Δ _IS_ human. Insisting that he be perfect in order to rehabilitate himself in the community's eye is completely unreasonable. --Hammersoft (talk) 13:13, 28 October 2011 (UTC)

• Delta is human, but the edits he makes (or at least those under discussion) have no human input beyond saving them. When people type text, they will make speling errors once in a while. When bots or scripts make edits, they will make the same errors over and over again. The benefit of someone adding content generally far outweighs the occasional typo they make. The benefit of most of the cleanup tasks Delta (and other bots) does is so small that any error in it often actually makes the article worse than it was before. And that is why the bar is clearly higher for automated or semi-automated edits than it is for manual, content-adding editing.
  Fram (talk
  ) 13:24, 28 October 2011 (UTC)

• Bull. So the only error he is allowed to make is if he's typing something directly? Perhaps you're unaware of the range of possible human errors? You're still expecting perfection from a process that CAN'T be perfect by definition. What you're asking for is impossible and unreasonable. --Hammersoft (talk) 13:27, 28 October 2011 (UTC)

(note, funny to make a spelling error in the term spelling error, Fram). It keeps being said, and though I have to confess I have no proof to the opposite (I'll ask Δ - well actually, going through the edits, there are edits there, where the edit summary is 'Cleanup' but which are clearly made by hand), some statements here do have no proof in favour either. Here quoting Fram: "have no human input beyond saving them". Are you sure that these edits are 100% scripted? Do you have any proof that Δ is not going through them, and adapting or removing certain things and then saving? Do you have any proof that Δ does not here and there think 'hmm, that is not correct', and does not save, or that he thinks 'oh, I have to do that as well, I'll be back in a second'? I'm sorry, that if a script returns a name for a site, then what reason in beginning do you have to think that that name is wrong - if I type 'user:Fram' in the searchbox above, what reason do I have not to think that I get to the correct page. Yeah, funnily enough the toolserver is in Germany so in some rare cases it will return the localised German version (and not some strange user preference) - but that is not wrong (but, granted, not optimal either). But do realise that documents which are hosted in Germany and linked to in Germany but used as a reference on the en.wikipedia may not have an English version, they only have a German version and those documents do return a German title. And there will also be documents based in the US, which are written in German, and hence do return a German title. He could even have thought 'hmm, funny, a German title, well, apparently someone who can read German added the reference' (Note: it may even be that some en.wikipedian based in Germany actually used the German version to write the text in English and added the reference). So I do agree, this could have been, and maybe should have been detected .. but it is not an eternal mistake, something that is so blatantly wrong that it MUST be seen and adapted. So yes, there will always be suboptimal things in edits, and I agree, also avoidable outright mistakes, even cases where things go so wrong that they actually break pages. Yes, we should have high expectations - and I do have high expectations for Δ as well - but not 100% error-free.

Same goes for the returning remarks that these edits are fully automated, and not checked. Although I do not have any proof that Δ is not fully automating his edits - there is no proof to the opposite either - only a suggestion that it may be here and there. If there are 10 edits in a very short burst which all 10 give a 20-page diff then one could think - but that is still possible with tabbed browsing 'open 10 pages - check 10 pages - copy-paste ten edit summaries - save 10 pages' - then the saving will be in an extremely short burst while all 10 edits are thoroughly checked (and I edit that way as well sometimes). And as I said, there are actually edits which are clearly manual, even with the 'cleanup' edit summary. That, to me, suggests that Δ ran the script to check if all was fine, nothing changed/all was fine, but saw something else which needed adding, which he subsequently did.

I do expect Δ to solve problems which break pages immediately (or to strictly avoid saving them), and I do expect a reasonable answer/discussion when it gives mistakes which maybe are avoidable (link titles, null-edits). I also expect that if something throws suboptimal results that extra care is given (spammy link-titles, null-edits) after being alerted to it. --Dirk Beetstra ^{T C} 14:25, 28 October 2011 (UTC)

In reply to Hammersoft, your attempt to establish fallibility by drawing comparison between a spelling mistake I made and Delta's errors is disingenuous. I don't use automated or semi-automated processes so when I make a mistake, it affects one edit. When Delta makes a mistake, it tends to affect large numbers of edits, and requires a lot more effort by others to repair. There's no comparison between the two. And as I said, I'm not under editing restrictions. If I were, I would have no problem with people being strict about the kinds of edits I made, and I would have no problem doubling or tripling scrutiny of my own edits. If I can't show that I can raise the bar on my editing quality while under restrictions, I cannot reasonably expect that those restrictions would be lifted. I expect no more from Delta here than I would of myself.

talk

) 23:32, 31 October 2011 (UTC)

Δ proposed task #1

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [7]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to articles to remove links to images which have been deleted (example, first line of diff)

On behalf of Δ, --Hammersoft (talk) 17:28, 24 October 2011 (UTC)

• Comment An edit summary more revealing than "(Cleanup)" would be useful if Delta is about to start doing very large numbers of edits like this. I'd also be more happy if Delta was making the request himself -- part of the process here is supposed to be reforming of Delta's attitudes, so he discusses with the community first -- and shows that he considers such discussions valuable, before he sets off on editing projects, which unfortunately the wheels have come off from too many times in the past. Jheald (talk) 17:57, 24 October 2011 (UTC)

• Please keep meta discussion regarding these requests in the appropriate section above. Do you have an objection to this type of edit in particular? --Hammersoft (talk) 18:16, 24 October 2011 (UTC)
• No! What evil will befall this project if Beta can remove those commented out dead links is impossible to imagine... Ok, sarcasm mode off. Obviously, I support this proposal, I am just aghast we have to waste our time discussing what should be a non-issue. --_{Piotr Konieczny aka Prokonsul Piotrus| talk to me} 18:30, 24 October 2011 (UTC)

Two questions. 1. Shouldn't this be done by a bot rather than by hand? 2. Why is it useful to have Beta (who has recently been prohibited from some other image-related work [8]) perform this task, rather than another bot operator who is less likely to draw criticism? — Carl (

• (1) Is there a reason a human shouldn't be allowed to do it? (2) The sanction you refer has to do with non-free images. This request has nothing to do with non-free images. I think it needs to be clarified that Δ is not a bot, and this is not a request to operate a bot. --Hammersoft (talk) 18:38, 24 October 2011 (UTC)
  • It would be better for Beta to avoid image-related cleanup entirely, if he is trying to steer clear of further arbcom motions. Moreover, if an image is deleted for NFCC reasons and Beta then removes the link to the image, this task has an air of non-free image maintenance. I agree he is not a bot, but why should this be done by him in particular versus someone else or a bot? — Carl (
    CBM · talk
    ) 18:43, 24 October 2011 (UTC)

• So you're extending the ArbCom imposed sanctions to include anything to do with images then? Do you have ArbCom approval for such an extension? If you are objecting to the edit because Δ would be doing the edit, then why don't you just make a request to have him banned from the project? Your opposition to all of these proposals is preventing him from editing anyway. --Hammersoft (talk) 18:45, 24 October 2011 (UTC)
  • The arbcom motion only covers non-free image enforcement "broadly construed". But it is in Beta's interest to avoid image-related work entirely, I believe. I support his ability to edit, I am not asking for a ban. But I do not see a compelling argument why he should receive special permission to perform this type of edit at this time, given his recent arbcom motion. — Carl (
    CBM · talk
    ) 18:50, 24 October 2011 (UTC)

• Do you see a compelling reason for him to edit? --Hammersoft (talk) 18:51, 24 October 2011 (UTC)

• Support with modifications. Well, I do think Carl is right that it might be as well for Delta to stay away from images entirely (especially in the light of Wikipedia:Requests for arbitration/Betacommand_2), but on the other hand I can think of no defensible reason why he should be unable to mass-remove images that have already been deleted. I think this would be the limiting factor, so I would like to suggest a reword to:

Δ may undertake non-controversial edits to articles to remove links to deleted images. This does not extend to NFCC enforcement activities or to removal of links to images currently under community discussion ("community discussion" being broadly construed). --Tristessa (talk) 02:22, 25 October 2011 (UTC)

• Support modified version
  Fram (talk
  ) 07:11, 25 October 2011 (UTC)
• I think the modified version is asking for trouble. Delta is doing these in a fast, nearly automated, way and has shown little interest in changing. It seems likely he'll not notice the exact issue and get blocked again. I'm not big fan of Delta, but I see no reason to effectively "set a trap" for him either, especially wrt images. Hammersoft, what do you think? In any case I oppose the unmodified version as skating too close to his editing restrictions. Hobit (talk) 10:19, 25 October 2011 (UTC)
• Non-controversial? There are people in these discussions that think anything Δ does is controversial. That's a blatantly subjective term that will be used to bludgeon him if he does any edits of this type. --Hammersoft (talk) 13:46, 25 October 2011 (UTC)
• Oppose. When viewing a diff, it is not apparent that the image has been deleted, and the edit summary does not explain why it was deleted. Further, in some cases, the name of the image file may help a subsequent editor figure out what the image was about and find a substitute image that improves the article. Jc3s5h (talk) 15:04, 25 October 2011 (UTC)

• Support. --Dirk Beetstra ^{T C} 15:09, 25 October 2011 (UTC)
• Oppose unless Δ checks to make sure there's no trivial replacement available. --SarekOfVulcan (talk) 15:32, 25 October 2011 (UTC)
• Conditional support for Tristessa's modified version, provided that Δ checks for trivial replacements and provides a helpful edit summary. – Luna Santin (talk) 22:25, 25 October 2011 (UTC)
• Waaaaay too close to the original nexus of all Beta-drama. The community simply does not trust Beta to remove links to images. Allowing him to do so en masse is simply asking for trouble. The sky will almost certainly not fall down if this has to be left to the rest of us. Chris Cunningham (user:thumperward) - talk 14:10, 26 October 2011 (UTC)
  • I see there are some separate ArbCom restrictions related to that (NFCC images). I'm not sure if the community may override those without consulting with the ArbCom—it's like firing them otherwise. So, I think you're right, the should be consulted before Δ is given approval for this particular task.
    talk
    ) 16:56, 26 October 2011 (UTC)
• Note:
  talk
  ) 18:08, 26 October 2011 (UTC)
• Oppose per my arguments in task #3. These attempts to punch holes in Delta's editing restrictions are not in line with the intention of those restrictions and until Delta demonstrates he can work within his restrictions as they stand currently, I see no reason to relax them.
  talk
  ) 00:17, 1 November 2011 (UTC)
• Oppose per
  TechnoSymbiosis. Besides, isn't there already a bot that does this? Kaldari (talk
  ) 06:25, 1 November 2011 (UTC)

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [9]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to articles to change links to templates where the original pointer was to a redirect, and the corrected pointer is to the template where the redirect was pointing (example, first line of diff)

On behalf of Δ, --Hammersoft (talk) 17:28, 24 October 2011 (UTC)

I don't see a problem if another valid edit needs to be made, and redirects are resolved in combination with the other edit, but resolving redirects doesn't usually seem worth doing as an edit of its own. Is there some other reason for resolving the redirects? Gimmetoo (talk) 17:39, 24 October 2011 (UTC)

Especially given

CBM · talk

) 18:29, 24 October 2011 (UTC)

• So what would you have me do, lump everything together in one giant request? As to WP:NOTBROKEN, what harm is being caused by the edit? Is there a reason to now allow it? --Hammersoft (talk) 18:34, 24 October 2011 (UTC)
  • We generally do not permit large-scale changes of this sort by bots, and AWB users are supposed to make a more significant edit before making this kind of "general fix". It is not clear why Beta would need to make this change to a large number of articles without making any other change to them. — Carl (
    CBM · talk
    ) 18:40, 24 October 2011 (UTC)

• Great! Could you please point out to me the guideline/policy that forbids people from performing this type of edit unless they are using AWB or are having a bot do it on their behalf? Do you have a specific objection to this task other than you think it should be done by a bot or an AWB user? --Hammersoft (talk) 18:44, 24 October 2011 (UTC)

Carl did link to the guideline: "Do not "fix" links to redirects that are not broken". This one is for everyone. Why do these redirects need to be resolved? If there is no reason other than "they are redirects", then it should not be done except in conjunction with another edit. Gimmetoo (talk) 18:49, 24 October 2011 (UTC)

• Fair enough. So, are we agreed it's ok to do this sort of edit in conjunction with other approved edits? --Hammersoft (talk) 18:51, 24 October 2011 (UTC)

• That seems fine, it is already done that way by AWB so there is a precedent. — Carl (
  CBM · talk
  ) 18:57, 24 October 2011 (UTC)

• Support as described in this thread, provided that Δ provides a helpful edit summary. – Luna Santin (talk) 22:27, 25 October 2011 (UTC)

In general, so long as resolving a redirect to a template is done "in conjunction" with other non-trivial edits, I don't really object, but be aware that there may be some opposition to orphaning a redirect to a template in particular cases, and this task might draw criticism. Gimmetoo (talk) 19:04, 24 October 2011 (UTC)

• Support - not at all controversial. --Tristessa (talk) 02:24, 25 October 2011 (UTC) Changed to Support modified version, taking the above comments into account (otherwise non-controversial I think):

Δ may undertake edits to articles to change links to templates where the original pointer was to a redirect, and the corrected pointer is to the template where the redirect was pointing, except in circumstances where community consensus may indicate that the alias should remain in use. --Tristessa (talk) 02:30, 25 October 2011 (UTC)

• Oppose, no foreseeable benefit. Hammersoft, you asked 'what harm is being caused by the edit'. I would argue that the threshold for making an edit on Wikipedia is 'what good is caused by the edit'. If the answer is 'none', the edit shouldn't be done.
  talk
  ) 05:01, 25 October 2011 (UTC)

• See also my arguments in task #3. These attempts to punch holes in Delta's editing restrictions are not in line with the intention of those restrictions and until Delta demonstrates he can work within his restrictions as they stand currently, I see no reason to relax them.
  talk
  ) 00:18, 1 November 2011 (UTC)

• Oppose as it stands. Could support if only those replacements as agreed upon for AWB (see
  Fram (talk
  ) 07:11, 25 October 2011 (UTC)

Δ may undertake edits to articles to change links to templates where the original pointer was to a redirect, and the corrected pointer is to the template where the redirect was pointing for templates which are also changed in the current version of AWB. (example, first line of diff)

Otherwise this is going to get to the point of 'you changed 26 templates in your last 100 edits', and if not, then 'you've removed 26 images during your last 100 edits' is also not a pattern - obviously support my version. --Dirk Beetstra ^{T C} 15:12, 25 October 2011 (UTC)

• Oppose per NOTBROKEN and the fact that inspection of Beta's own comments will show that they see the benefit of doing this as tracking down templates which use non-free images. Beta is topic banned from non-free image work. Franamax (talk) 17:46, 25 October 2011 (UTC)

• Im going to have to {{
  Fact}} that comment. Ive never used this task for NFC issues, its just used for simplifying the wikitext, and making templates easier to work with. ΔT _{^{The only constant}}

• Support. This is beneficial for the project, EOT as far as I am concerned. --Piotr Konieczny aka Prokonsul Piotrus| talk to me 18:00, 25 October 2011 (UTC)
• WP:NOTBROKEN specifically addresses this. Should Hell freeze over and WP:NOTBROKEN be declared to no longer have community consensus, a real bot can easily do this. Chris Cunningham (user:thumperward) - talk
  14:36, 26 October 2011 (UTC)
• Oppose per NOTBROKEN. Karanacs (talk) 16:24, 28 October 2011 (UTC)
• Oppose. This seems like a pointless task. If the redirect isn't broken, why fix it? Kaldari (talk) 06:27, 1 November 2011 (UTC)

Δ proposed task #3

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [10]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to remove external links where such links were used as a failed attempt to include an image in an infobox (example, first line of diff)

On behalf of Δ, --Hammersoft (talk) 17:28, 24 October 2011 (UTC)

In general, this is OK, but not optimal. Often these external links exist because someone tried to replace an acceptable image. Simply removing the external link leaves the page without an image. This sort of edit is really better done by a human who examines the edit history. Gimmetoo (talk) 19:11, 24 October 2011 (UTC)

• Point of fact; contrary to popular belief Δ is human :) --Hammersoft (talk) 19:39, 24 October 2011 (UTC)

Granted. Will delta look at the edit history, find the most-recently-used non-deleted image, and restore it? That could even be scripted. Gimmetoo (talk) 20:02, 24 October 2011 (UTC)

• Support provided edits are fully reviewed. --Tristessa (talk) 02:31, 25 October 2011 (UTC)
• Support provided Delta continues to follow his other restrictions (no automated edits or edits that appear automated; manual, careful, individual review of each edit).
  talk
  ) 05:03, 25 October 2011 (UTC)

• Oppose per discussion elsewhere in this overall wall of text. Hammersoft has successfully convinced me that Delta has respect for neither the letter nor the spirit of his current editing restrictions, and that he has no intention of working within his restrictions to redeem himself in the eyes of the community. The entire process has instead attempted to undermine and chip away at his restrictions with no assurance whatsoever that he will be able to correct his past mistakes, nor even an acknowledgement that he made mistakes and that the community found this problematic. I oppose any increase in his editing privileges until such time as he demonstrates genuine understanding and genuine respect for the community, and takes steps to redeem himself, within his current limitations.
  talk
  ) 00:13, 1 November 2011 (UTC)

• Support.
  Fram (talk
  ) 07:11, 25 October 2011 (UTC)
• I can imagine a case when a clueless newbie tries to update the infobox logo of a company after an official change, but instead of uploading the new picture the newbie puts an external link there. In that case it would preferable to revert. If the removal is a semi-automated task and does not check the recent history, perhaps a message should be automatically left on the talk page.
  talk
  ) 14:19, 25 October 2011 (UTC)
• Oppose unless Δ checks to be sure the external link didn't replace a valid internal link.--SarekOfVulcan (talk) 14:49, 25 October 2011 (UTC)
• Conditionally support with same restriction as SarekOfVulcan suggested at 14:49 UTC. Jc3s5h (talk) 15:09, 25 October 2011 (UTC)
• Support - but try to explain the editor how to actually get the image there. --Dirk Beetstra ^{T C} 15:14, 25 October 2011 (UTC)
• Support. This is beneficial for the project, EOT as far as I am concerned (agreed with SoV and DB though). --_{Piotr Konieczny aka Prokonsul Piotrus| talk to me} 18:00, 25 October 2011 (UTC)
• Conditional support with same restriction as SOV suggested at 14:49 UTC, otherwise oppose. As Gimmetoo notes, this is not amenable to script-based editing. Franamax (talk) 18:05, 25 October 2011 (UTC)
• Conditional support per Franamax, Sarek, Gimmetoo, provided that Δ provides a helpful edit summary. – Luna Santin (talk) 22:28, 25 October 2011 (UTC)

Δ proposed task #4

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [11]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to move content existing below category tags on an article to a position above the category tags in accordance with

Wikipedia:CAT#Categorizing_pages. (example

, bottom portion of diff)

On behalf of Δ, --Hammersoft (talk) 17:29, 24 October 2011 (UTC)

Does "content" exclusively mean templates? Gimmetoo (talk) 17:42, 24 October 2011 (UTC)

• Construe it as in compliance with the directive linked above. --Hammersoft (talk) 17:46, 24 October 2011 (UTC)

I don't see how the sample diffs follow from the guidelines, nor how this task could be automated without more specification. Why should templates be moved? Which templates? Without more details I really don't know what's being proposed here. Gimmetoo (talk) 17:58, 24 October 2011 (UTC)

• The appropriate line from
  Wikipedia:CAT#Categorizing_pages is "By convention, category declarations are placed at the end of the wikitext, but before any stub templates (which themselves transclude categories) and interlanguage links." The before and after links above demonstrate that. I didn't indicate the task or any of the proposed tasks should be automated. --Hammersoft (talk
  ) 18:17, 24 October 2011 (UTC)
  • We generally do not allow AWB users or bots to make this sort of edit en masse. Why would it be appropriate for Beta to do it automatically when they cannot? Usually we say that these things should only be done if there is some more significant reason to edit the page. — Carl (
    CBM · talk
    ) 18:27, 24 October 2011 (UTC)

• If you would be so kind, could you please indicate where I said "automatically" or even implied "automatically"? Thank you. As to performing the edit, it is asked for by the
  Wikipedia:CAT#Categorizing_pages. Should we remove the above quoted line from this guideline? --Hammersoft (talk
  ) 18:35, 24 October 2011 (UTC)
  • If Beta is given permission to do this, then he will have permission to do it on a large-scale basis. I don't see your argument for why it is necessary, nor why he should be authorized to do it. Even if it was necessary, I would think AWB could handle it. — Carl (
    CBM · talk
    ) 18:37, 24 October 2011 (UTC)

• Is there a reason he can't do it? Nothing on Wikipedia is "necessary". People edit according to their interests. Nobody is required to edit anything. The fact he is finding such errors shows me at least that all the various AWB users aren't finding all instances of this error. By performing the edit, he is helping the project and doing so in accordance with guidelines. Is there some reason he can't do it? --Hammersoft (talk) 18:41, 24 October 2011 (UTC)
  • Presumably, if Beta is asking for special permission to do a large-scale maintenance task, there should be a compelling reason why he needs to do it. This particular task would not be approved for a bot to do, because requests to perfrom only trivial maintenance like re-arranging the bottom of the article are not approves. It isn't clear to me how this proposal is different enough at different standards would apply. Is there a special standard that would be used to select articles to edit? — Carl (
    CBM · talk
    ) 18:47, 24 October 2011 (UTC)

• There is no compelling reason Δ needs to edit here at all. Why not just permanently block him? I note you haven't opposed #3 yet. Do you support #3? --Hammersoft (talk) 18:49, 24 October 2011 (UTC)

• Oppose - as per Carl I'm not convinced that this should be done on a large scale without wider, specific consensus for such a mass change; this task is also prone to mistakes and, without adequate edit review, is something of a Pandora's box. --Tristessa (talk) 02:34, 25 October 2011 (UTC)
• Oppose, this is the kind of thing that needs to be done one a case-by-case basis, often text added below the cats is vandalism, or comments that belong on the talk page.
  Fram (talk
  ) 07:11, 25 October 2011 (UTC)

• Support - we are not talking about automated edits per sé here, we are talking about patterns of edits. Of course, if it is vandalism, it needs reverting, but if {Δ would follow the AbuseFilter and get to 25 articles which are not vandalism and move the text up then for sure, {Δ should be able to do that. --Dirk Beetstra ^{T C} 15:16, 25 October 2011 (UTC)
• Oppose per "we are not talking about automated edits per sé here". We are.
  talk
  ) 16:01, 25 October 2011 (UTC)
  • No .. this can also be done fully manually - scripted, maybe, but not automated. Δ is still obliged to check. --Dirk Beetstra ^{T C} 14:19, 26 October 2011 (UTC)
• Comment. Most of that is vandalism, not all... should be done on case by case basis, automated move on likely vandalized text into main article would not be helpful. --_{Piotr Konieczny aka Prokonsul Piotrus| talk to me} 18:00, 25 October 2011 (UTC)
  • Not only that. I've seen mis-formatted categories using ";" instead of ":" and what not at AfC. Generally one needs to take a look at what the heck is the stuff at the end trying to do.
    talk
    ) 18:12, 25 October 2011 (UTC)

• comment All of my edits are manually reviewed before saving and I check for many issues, I often see reference sections added below interwiki stubs and categories. the correct layout is categories, stub templates, and finally interwiki links. ΔT _{^{The only constant}} 19:15, 25 October 2011 (UTC)

  You say that, but what evidence do we have to support it? In the past you reportedly provided screenshots which just showed you hammering "Y" to approve edits without actually verifying them (I didn't see them, but I'm going off an old AN/I thread on that), and a few months ago you verified that you weren't actually viewing the pages when you reverting, simply looking at the contents of the diffs (in relation to you reverting someone who put a link to an image and not the actual image on a talk or project page).

  Crossmr (talk

  ) 07:33, 26 October 2011 (UTC)

• Support seems sensible and is good cleanup. -- Eraserhead1 <talk> 19:48, 25 October 2011 (UTC)
• This task is not suited to rapid editing, per Carl, and the community does not have enough faith in Beta to accept his vouching for personal inspection of each case, per Crossmr. Chris Cunningham (user:thumperward) - talk 14:17, 26 October 2011 (UTC)
  • Rapid? Still less than 40 edits in 10 minutes .. and certainly every edit should be checked (Δ is not allowed to run a bot, scripted edits alone). --Dirk Beetstra ^{T C} 14:20, 26 October 2011 (UTC)
    • One edit every fifteen seconds for ten minutes is blazingly fast. The specific speed limit is codified so he has a bright line, rather than because we've decided that four edits a minute is scientifically the right value. If we do not allow regular AWB users to carry out this "task" then there is certainly no rational for making an exception for BetaCommand. Chris Cunningham (user:thumperward) - talk 14:46, 26 October 2011 (UTC)

Oppose per Chris Cunningham. Karanacs (talk) 16:28, 28 October 2011 (UTC)

• Oppose per my arguments in task #1 and #3.
  talk
  ) 00:20, 1 November 2011 (UTC)
• Oppose per Chris Cunningham; all kinds of text ends up below the categories (including templates and other valid edits; vandalism; attempts at discussion; various unhelpful stuff from inexperienced users, &c) and each case therefore requires a little thought rather than rapid-fire repetition of a single change with little regard to the end result. I lack confidence that this would be done properly. bobrayner (talk) 20:57, 2 November 2011 (UTC)
  • Also: The strict sequence of things at the bottom of an article (categories, interwikis, stub templates, &c) is just a matter of convention and does not make a great deal of difference to the average reader. Making a great effort to fix these without actually fixing real content in an article strikes me as pointless busywork; but if conscript Δ is actually volunteering to dig a trench from the fencepost until lunchtime, that's no reason for me to stop them. bobrayner (talk) 21:11, 2 November 2011 (UTC)

Δ proposed task #5

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [12]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to apply appropriate reflist template parameters when stylization to fix layout is appropriate. (example: before and after)

On behalf of Δ, --Hammersoft (talk) 17:29, 24 October 2011 (UTC)

See

CBM · talk

) 18:25, 24 October 2011 (UTC)

• If you would be so kind, would you please point out where I said "automatically"? Thank you, --Hammersoft (talk) 18:35, 24 October 2011 (UTC)
  • The only reason Beta would need to request this is if he was planning to do it on a large-scale basis, presumably using automated or semi-automated tools like his "cleanup" script. Even AWB does not do this task, so it is not clear why it would be appropriate for Beta to do it. — Carl (
    CBM · talk
    ) 18:38, 24 October 2011 (UTC)

• Carl, are you going to oppose every request? I'm sensing a pattern here and growing increasingly unsettled that this is the case. To this request specifically, is there some reason he can't do it? Or is this just a global exception to Δ editing, period? --Hammersoft (talk) 18:42, 24 October 2011 (UTC)
  • I support Beta editing, but there is no reason why this is the sort of editing he should be doing. — Carl (
    CBM · talk
    ) 18:44, 24 October 2011 (UTC)

• Do you have a reason why he shouldn't; do this type of edit or are you just opposing because it's him? --Hammersoft (talk) 18:46, 24 October 2011 (UTC)
  • This particular edit should not be done because of WP:CITEVAR. Even AWB does not perform this type of edit (changing the number of columns in the reference list). — Carl (
    CBM · talk
    ) 18:48, 24 October 2011 (UTC)
    • RefLinks does automatically add columns if it detects a large number of references. Alpha_Quadrant ^(talk)
      19:00, 24 October 2011 (UTC)

• Oppose - sorry, this particular one must be done by hand as there is a tendency for an automated/semi-automated application of this to result in all sorts of errors and unforeseen consequences; not least, however, is the objection that this should not be changed wholesale per
  WP:CITEVAR. --Tristessa (talk)
  02:38, 25 October 2011 (UTC)
• Oppose per Tristessa.
  Fram (talk
  ) 07:11, 25 October 2011 (UTC)
• Oppose, this should not be automated. --SarekOfVulcan (talk) 14:53, 25 October 2011 (UTC)

• Support - we are not talking about automated edits per sé here, this should be applied with the mentioned common sense. May get to 25 in a row, don't see any reason to disallow the common sense, manual, improvement even when it does get over 25 in a row. --Dirk Beetstra ^{T C} 15:18, 25 October 2011 (UTC)
• Oppose per "we are not talking about automated edits per sé here". Gimme a break with the hypocrisy.
  talk
  ) 15:57, 25 October 2011 (UTC)
  • Hypocrisy? Thank you, Have mörser, will travel. --Dirk Beetstra ^{T C} 14:21, 26 October 2011 (UTC)
• Oppose – even the before/after example is lame. Someone subsequently fixed it by hand. Dicklyon (talk) 17:21, 25 October 2011 (UTC)
• Support. This is beneficial for the project, and my experience with Beta and ref fixing has been quite positive. --_{Piotr Konieczny aka Prokonsul Piotrus| talk to me} 18:03, 25 October 2011 (UTC)
• Oppose per Tristessa. Franamax (talk) 18:24, 25 October 2011 (UTC)

• Comment this isnt done too often. I only remove multi columns if there are less than 8 references (its pointless at that few references) and I only add multi columns if there are more than 30 references (where mulit-columns really become effective.) So please take that into consideration. ΔT _{^{The only constant}} 19:18, 25 October 2011 (UTC)
  • Which articles do you plan to apply it to? Do you have a list, or are you proposing to apply it to all articles that meet those criteria? If so, why can't a bot do it instead, if there is consensus? — Carl (
    CBM · talk
    ) 19:23, 25 October 2011 (UTC)
    • A bot could do it, however no such bot currently does it, and it probably wouldnt get approval as it is a very trivial edit that I do in conjuncture with other edits, and I do it on an article by article basis as needed. Im just seeking approval so that this part of my edit cannot be considered a block-able pattern down the road. ΔT _{^{The only constant}} 19:31, 25 October 2011 (UTC)
  • Support Δ's description right here, provided a helpful edit summary is used. Opppose description above (too vague). – Luna Santin (talk) 22:33, 25 October 2011 (UTC)
• Oppose per
  WP:CITEVAR
  . This is bound to cause problems for some editor somewhere and there are editors out there who are deeply attached to their citation style, regardless of how idiosyncratic when compared to Wikipedia as a whole. You are in for the fight of your Wikipedia career if you mess with them.

Also, the example you gave added a citation in a format that is not documented in

harvnb}}, etc.) before suggesting an automated edit that effects citations. ---- CharlesGillingham (talk

) 23:17, 25 October 2011 (UTC)

• One Im not a bot, two Ive made over 8,000 edits with this gen fix enabled without a single complaint or issue raised about this item. ΔT _{^{The only constant}} 00:34, 26 October 2011 (UTC)

• I'm sorry, I misread the diff the first time out. The change was not nearly as large as I thought at first glance. Is this request just to change/add the |col= in {{reflist}}? I think you should ask specifically about that, if that's the only change you will be making. ---- CharlesGillingham (talk) 02:43, 26 October 2011 (UTC)

• Oppose per
  talk
  ) 00:22, 1 November 2011 (UTC)

Δ proposed task #6

In accordance with

• The first bullet item of Δ's editing restrictions,
• Agreement by Δ for me to operate on his behalf in making requests in accordance with said restriction [13]
• and in compliance with other editing restrictions which may apply to this request

the following request for editing is being made:

Δ may undertake edits to remove white spaces as appropriate in accordance with Help:Whitespace. (example, first part of diff affecting infobox)

On behalf of Δ, --Hammersoft (talk) 20:50, 24 October 2011 (UTC)

Of course Delta can fix the sort of whitespace issues described in Help:Whitespace, but the issues are varied and wouldn't seem likely to me to form a "pattern". Am I missing something here? The diff in question, and your explanation, seems to suggest removing linebreaks, which is a separate issue; in that diff they are not actually producing visible whitespace. Gimmetoo (talk) 22:26, 24 October 2011 (UTC)

• Δ was previously accused of conducting patternistic edits in removing white spaces. Thus, this request. --Hammersoft (talk) 22:42, 24 October 2011 (UTC)
  • But that was referring to the removal of spaces inside the wikicode, rather than to removing whitespace from the rendered article. There is no reason for him to go around doing the former, while
    CBM · talk
    ) 01:18, 25 October 2011 (UTC)

• Splitting hairs about white space? Is that was this has devolved to? So now I have to make a request about different types of white space? First ANY white space is a pattern, and now we have to delineate what kind of white space? That blade has two edges. --Hammersoft (talk) 01:43, 25 October 2011 (UTC)

• Conditional support if in accordance with Help:Whitespace (that is, article whitespace) is interpreted strictly, or else Δ proposes a change to those guidelines and obtains consensus for this change to be implemented at scale. --Tristessa (talk) 02:43, 25 October 2011 (UTC)
• Conditional support per Tristessa.
  talk
  ) 04:27, 25 October 2011 (UTC)