The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at
WT:BRFA. The result of the discussion was Approved.
Already has a bot flag(Y/N): Y
VVVV See revised proposal VVVV
Function details:
If has an unref tag or similar will simply remove the cat
If has evidence of references will simply remove the cat
There are some "long tail" items i.e. months with a single item or a few. These may be consolidated into the oldest reasonable non-empty month.
Discussion
VVVV See revised proposal VVVV
Background. Erik9bot created this category, which should probably have been simply applying an appropriate tag. CiterSquad has provided cites for a good number of articles, and a couple thousand others have had cites added and the cat left in - I have fixed these already, though there may be more by now. The category fails for a number of reasons:
it is hidden, so it is not seen and it gets left behind.
it is too large.
it stops articles getting tagged properly.
it doesn't follow the dating pattern of other categories
will this be a hidden version of the template, what I don't want to see is a lot of stub articles getting plastered with annoying obnoxious templates. is there some way of adding it so that its not visible yet still in a hidden category?
βcommand 01:47, 31 October 2009 (UTC)reply
We can make the template hidden but it is rather against consensus. Stub articles need to be referenced too. RichFarmbrough, 18:51, 31 October 2009 (UTC).reply
I too dislike the idea of very short stubs being tagged. Could the bot be programmed to recognize stubs, perhaps with a regex, and not tag those articles with unreferenced? Or maybe recognize them by byte size? --
Ϫ 18:17, 1 November 2009 (UTC)reply
It could simply skip stubs. or we could have a stub=yes parameter. But stubs are dangerous - they hang around a long time un-reviewed and unreferenced - they are smaller and get less traffic but that doesn't mean they are less prone to error. In theory stubs should be easy to find references for since they contain very little. Also the act of finding reference number 1 should go with some basic assessment/fixing of other aspects of the stub. One of the things that does concern me is that we get copied and used all over the place and can easily end up using something derived from a WP article as a reference for the article itself. RichFarmbrough, 22:02, 1 November 2009 (UTC).reply
One problem area with stubs is with organism stubs. The orphan templates and the unreferenced templates make the text of the article hard to see. It looks like there is no text at all. When an organism article is a stub it can be a one-lined stub, maybe someone didn't add a reference, and it certainly should be referenced, but by putting a template that obscures the text, you've made the article completely useless. You might as well speedy delete the article. But with the single line of text and the taxobox the organism stub is at least a starting point for a reader finding more information. Is it possible to categorize these articles? In particular, could they be categorized unreferenced by phyla or divisions or orders? It might be easy to get wikiprojects to reference the articles, then. I add references to organism articles by the boatload while cleaning them up.
I don't see that with single line stubs, as most organism articles are, that the information is copied into reliable sources from wikipedia. Wikipedia is pretty much forbidden as a source for taxonomic information, because it uses mixed taxonomies, among other reasons. So, this is a non-issue for these articles. --
69.225.3.198 (
talk) 16:08, 2 November 2009 (UTC)reply
The majority of organism stubs are created in chunks and can be referenced in chunks - if they aren't already, most recent ones are. And yes they can be categorised, most are already, I was looking at some beetle stubs the other day and they really fit my ideas for dynamic taxonomic categorisation. RichFarmbrough, 22:35, 2 November 2009 (UTC).reply
Is there a page where I can find unreferenced by category organism stubs? I used to have lists of articles I was going to write or reference, and I think they were in categories like this. Having them categorized by phyla, in my opinion, would be incredibly useful. I don't know about referencing in chunks, but if they can be that would be great. Still, categorizing by phyla at least, better by class, etc., would make future referencing easier. --
IP69.226.103.13 (
talk) 08:38, 14 November 2009 (UTC)reply
Unreferenced stub
I like this proposal; I particularly like the small separate {{Unreferenced stub}} would these articles be identifiable in the (same or different) category? As mentioned above finding references (or not) for short stubs is an easier body of work then full articles that are unreferenced. I would support the automated use of {{Unreferenced stub}} placed at the bottom of the article for any article with a stub marker.This is a good compromise between all the different views on unreferenced templates and stubs.
JeepdaySock (
AKA, Jeepday) 16:35, 2 November 2009 (UTC)reply
As it stands same categories. RichFarmbrough, 22:35, 2 November 2009 (UTC).reply
The reason this category was used at all is because there was not agreement to just add the unreferenced tag automatically (see
Wikipedia:Bots/Requests for approval/Erik9bot 9). The arguments made there (I did not participate) would apply to this request as well.
On the subject of stubs, the stub tag itself indicates that the article requires significant improvement. There is no reason to add "unreferenced", "expand", or similar templates to articles that are already marked as stubs. The stub tag itself indicates that the article is likely to be deficient in many ways. — Carl (
CBM ·
talk) 12:54, 19 November 2009 (UTC)reply
Would those who would prefer not have stub articles tagged, be willing to jointly pursue expanding {{stub}} (This article is a stub. You can help Wikipedia by expanding it.) to something like "This article is a
stub. You can help Wikipedia by expanding it with
cited sources.
Jeepday (
talk) 11:31, 22 November 2009 (UTC)reply
Opposed
I oppose this task, as it is currently formed. I find the over-use of these cleanup templates to be garish, distracting, and unnecessary. Worse, the addition of these templates often seems arbitrary, which has led to the use of them becoming semi-permanent on many articles, which is problematic for several reasons. I bring all of these up because I feel strongly that this is not an appropriate bot activity. Bot actions should be essentially uncontroversial and easily reverted, and I don't think that this activity qualifies. Something should definitely be done with the category and the pages that are in it, but this doesn't seem to be the most effective solution in my view. We could, and likely should, have a wider discussion (outside of the RFBA process) on this issue as a whole. —
V = I * R (
talk to Ω) 13:43, 19 November 2009 (UTC)reply
I agree - in fact I had rather optimistically assumed that the bot was adding this category to articles as an alternative to the ugly and misleading "unreferenced" tags. We certainly don't want a whole lot of automatically generated instances of this tag, that serves no purpose except to tell people what they can already see, and falsely implies that articles which cite references are somehow reliable.--
Kotniski (
talk) 13:45, 19 November 2009 (UTC)reply
Well there are items that have remained unreferenced for 8 years. The tagged items only backlog 3 years. If you think the tag is ugly and misleading go and fix the tag. RichFarmbrough, 22:18, 19 November 2009 (UTC).reply
Having said that a smaller tag is suggested for stubs, or indeed a sane compromise is to make the tag invisible for stubs. Why is this sane - compared with what was done before? Because it addtresses items 2,3,4 and 5,and if 1 is still found to be a problem it can be addressed later on. RichFarmbrough, 22:47, 19 November 2009 (UTC).reply
See though, this won't actually be dealing with the real issue here at all. Namely, those articles that have been unreferenced for 8 years will still be unreferenced regardless of how many tags, categories, or anything else that is added to them. I don't object to giving smackbot or any other bot as many useful tasks as possible, and this certainly isn't about any aesthetic concerns (on my part), but this task really just plain doesn't seem useful. I don't think that either this or the original Erik9Bot approval to add the cat really took the criticisms that are being offered to heart. Is this proposal (or the existing erik9bot category) really helping the encyclopedia in some manner? —
V = I * R (
talk to Ω) 00:10, 20 November 2009 (UTC)reply
As proposed all the articles will be addressed by
Wikipedia:Unreferenced articles which will work to systematically address each article based on age. Articles get references or they get deleted, if they can't be referenced. I would say that qualifies as "helping the encyclopedia in some manner"Jeepday (
talk) 00:27, 20 November 2009 (UTC)reply
Assuming that said project is actually active (which is quite an assumption), is it clear that they are supportive of having their workload essentially doubled overnight? Has anyone actually notified the project, or talked to them at all? I should reiterate that the usefulness of categorization or tagging is really an entirely separate question, since we should be discussing the bot editing here. The two issues naturally conflate together somewhat, but the basic question is "should a bot do this", and my reply to that question tends to be "no". —
V = I * R (
talk to Ω) 00:39, 20 November 2009 (UTC)reply
The project has completed 219 articles so far this month, which puts them on track to finish all the presently-tagged articles 31 years from now if they keep up the same rate. — Carl (
CBM ·
talk) 00:48, 20 November 2009 (UTC)reply
OK, well I just aded a notification about this discussion to
Wikipedia talk:Unreferenced articles. I don't want to speak for them, and this is really a secondary concern for me regardless. My main issue is essentially the same as the objections that were ignored in regards to Erik9Bot. The simple fact is that this is clearly not an uncontroversial task, and therefore it should not receive approval. —
V = I * R (
talk to Ω) 12:45, 20 November 2009 (UTC)reply
I have some familiarity with the
Wikipedia:Unreferenced articles, I expect there will be no objects to adding these articles to the tasks. We will start with the oldest ones and move forward.
Jeepday (
talk) 12:38, 21 November 2009 (UTC)reply
Age
What about the fact that erik9bot tagged these articles a long time ago and that many of them may have references now? Will articles with references automatically be excluded, even if erik9bot has categorized the article as needing references? - ʄɭoʏɗiaɲτ¢ 18:15, 19 November 2009 (UTC)reply
I have run a clean-up and taken out those with references, and would do so again. This shows why the hidden cat was a bad compromise. RichFarmbrough, 22:18, 19 November 2009 (UTC).reply
I ran some code this morning to check, and it looks like about 70,000 of the 115,000 articles in the category are tagged as stubs. — Carl (
CBM ·
talk) 18:18, 19 November 2009 (UTC)reply
That excludes those maths stubs where you removed the category? Stubs do need references. Certainly they don't need expand tags which I have removed many of. RichFarmbrough, 22:21, 19 November 2009 (UTC).reply
Of course; I ran the counting code again this morning to get a fresh count. Although the number of math articles I can manually inspect is much less than 70,000. The stated goal of the Erik9 category was to let people review the articles by hand, and I followed that goal by looking at the wiki source code of a collection of math stubs in the category to make sure the articles really were tagged as stubs before removing the category. To be fair, I also removed {{unsourced}} from stubs when I come across it. Stubs do need to be verifiable, but they do not need most maintenance templates. — Carl (
CBM ·
talk) 22:54, 19 November 2009 (UTC)reply
"The stated goal of the Erik9 category was to let people review the articles by hand", very true, This was I believe mostly because there was a concern that many of the articles could have poorly formatted reference that the bot would not recognize. I personally have reviewed and tagged a few hundred or a couple thousand from the category working from
Wikipedia:CiterSquad. I found maybe one in 200 that had a poorly formatted references. Having also worked
Wikipedia:Unreferenced articles I can tell you that the number of articles that currently have {{Unreferenced}} or a variation and also have numerous well formatted references is far in excess of any that may be incorrectly identified. When the process was suggested at
Wikipedia:Bots/Requests for approval/Erik9bot 9 I was hesitant to have a bot adding {{Unreferenced}} or {{Unreferenced stub}}. I have now lots of experience in these bot identified articles and what has been marked is appropriate for tags in the unreferenced family.
Jeepday (
talk) 23:40, 19 November 2009 (UTC)reply
Over half the articles in the Erik9 category are tagged as stubs already. Both in the Erik9bot BRFA, and higher above on this one, various people have expressed the opinion that stubs should not be automatically given "unreferenced" tags when they are already tagged as stubs.
In particular: in the previous BRFA, Erik9 said, "Based on comments by Gimmetrow, Antandrus, Keith D, and Geogre opposing the automated addition of template:unreferenced to articles, I am revising the task for which approval is requested: ...". In this BRFA, Betacommand, OlEnglish, Ohm's Law, and Kotniski have objected. I also object to the automated addition of "unreferenced" to stubs. So it seems there is quite a bit of objection to the idea of mass tagging. — Carl (
CBM ·
talk) 23:48, 19 November 2009 (UTC)reply
This is a good point, imo. Stubs are already classified by the projects, animal stubs as feline, shark, moth stubs, etc., etc. Plants also. Maybe eliminating stubs from this list is a good idea. --
IP69.226.103.13 (
talk) 00:15, 20 November 2009 (UTC)reply
Just because they have stub tags does not mean they are stubs (RE:
User:Triddle/stubsensor), All articles including stubs require references
WP:V, any article that has no references is appropriate for an unreferenced template
Template:Unreferenced. I appreciate that many users have opinions counter to this statement, but the application of the tags proposed is completely within policy.
Jeepday (
talk) 00:36, 20 November 2009 (UTC)reply
WP:V requires that all articles are verifiable, not that every article must explicitly list references right now. Of course the goal is to eventually reference all articles, and the goal is also to expand all articles so that they are not tagged as stubs. In the meantime, articles that are still tagged as stubs can be expected to have many problems, so adding extra cleanup tags to them is just redundant. When the articles are expanded beyond stubs, references will be added, and if they are not added, then an unreferenced tag is reasonable. — Carl (
CBM ·
talk) 00:45, 20 November 2009 (UTC)reply
Huge category
This category is so huge it's worthless.
Yes! That's exactly the point #2 above. RichFarmbrough, 22:19, 19 November 2009 (UTC).reply
Can orphans be subcategorized some way, like orphaned angiosperm stubs? --
IP69.226.103.13 (
talk) 19:48, 19 November 2009 (UTC)reply
We aren't talking about orphans - that's another matter - see my Village Pump proposal to de-deprecate orphans. But yes it is possible to categorise unreferenced angiosperm stubs if that would be useful. Or by classis, or regnum or whatever. RichFarmbrough, 22:18, 19 November 2009 (UTC).reply
Anything that makes it possible for an editor to jump in and start dealing with a list of 30,000 or 70,000 articles should be done. No one's going to tackle a list they can't make a dent in, or a list where 27,600 articles are outside of their area. But, sometimes an editor will be willing to tackle a list that is manageable.
I will ask at the projects how they would like their stubs classified. Animals lists probably by order, but plants and fungi may be different. Removing at least stubs of living things to lists that are workable would be really useful, imo. --
IP69.226.103.13 (
talk) 22:57, 19 November 2009 (UTC)reply
Year articles and lists
I notice that a lot of the articles are years or other navigational pages (are there disambigs in this category too?). I was under the impression that these do not require a list of references.
OrangeDog (
τ •
ε) 13:31, 20 November 2009 (UTC)reply
Disambigs will have the cat removed (I think I have already done this). List articles may need references, or they may de facto delegate some referencing. For example
List of harmoniun players might be a pure bulleted list of links, in which case provided each article supported the harmonium hypotheses it would be fine (but better to use a category?) - if however it included un unlinked bullet:
Dave Farmbrough
Then a citation would be needed. The same applies to year articles - they probably mostly don't need citations, just checking with the articles they link to. RichFarmbrough, 14:44, 20 November 2009 (UTC).reply
Strong Support
Jeepday (
talk) 12:40, 21 November 2009 (UTC)reply
OK with me, although I don't think there is much reason to think the template is useful. — Carl (
CBM ·
talk) 13:05, 21 November 2009 (UTC)reply
If it categorizes it in some useful way, yes. If not, as in just puts it in unreferenced category, it's still useful to gather unreferenced articles together knowing they require work. --
IP69.226.103.13 (
talk) 07:16, 22 November 2009 (UTC)reply
Support, but can understand reluctance by some to embrace even {{Unreferenced stub}}. Some people believe the {{stub}} sufficiently indicates that the article is in need of significant work. Other people feel more strongly about highlighting the lack of references, as warning to novice readers and as a inspiration to improve the article by adding references, while providing links for how and why.
Jeepday (
talk) 12:50, 21 November 2009 (UTC)reply
Numerous editors have disagreed with this, both on this BRFA and the Erik9 BRFA. The stub tag already indicates that the article requires a lot of espansion. When the article is expanded until it isn't a stub, if it still has no references, then tagging it as unreferenced is more reasonable. But when we are talking about a typical one- to two-sentence stub, the lack of references is overshadowed by the lack of content in general. Anyone who adds to the content is likely to add at least one reference, and anyone who adds a reference is also likely to know enough to expand the content if they like. So, for stubs, expanding the article and adding references are two sides of the same coin. For that reason, the stub tag itself is enough of a maintenance tag on stub articles. — Carl (
CBM ·
talk) 13:10, 21 November 2009 (UTC)reply
Again, unreferenced echinoderm stub is useful, but, once an article is tagged as a stub, then as an unreferenced stub, and the main space templates overpower the text, while the fail to meaningfully categorize, I don't see the point. --
IP69.226.103.13 (
talk) 07:17, 22 November 2009 (UTC)reply
I would support this but there is no consensus for this. What about using the tag, but hiding it in some way. A |bot=yes parameter? This would categorize articles in the dated cats and make it easier for wikiprojects or editors who work on specific categories to identify the articles needing work using tools like
WP:CATSCAN or
Cleanup listings, and at the same time hide the big bad tag.
Rettetast (
talk) 11:59, 22 November 2009 (UTC)reply
The articles already have a hidden marker, namely the category that Erik9bot placed on them. How does changing from one hidden marker to another help? — Carl (
CBM ·
talk) 19:00, 27 November 2009 (UTC)reply
Part of the rationale is based here
Wikipedia:Administrators'_noticeboard/IncidentArchive566#Erik9_appears_to_be_the_sock_of_a_banned_user, while there is no question that the actual work of the bot was within policy, having a category that includes the name of a banned user has created controversy. If the stub articles remain marked as unreferenced (new category) then they will be available for work in a category for
WP:CSQ or other projects. There are stubs with references and there are stubs without, only stubs without references will be in the catogory. These are relatively easy to address and will be good for users not up to referencing projects like
WP:FRC. If I have not fully addressed your question, please let me know.
JeepdaySock (AKA,
Jeepday) 20:45, 27 November 2009 (UTC)reply
Revised proposal
Function details:
If already has an unref tag or similar will simply remove the category
If has evidence of references will simply remove the category
If it is a stub an invisible tag will be added. This will include a parameter to show it was automatic, and if it contains a taxobox the appopriate taxon or taxa. The template will categorise the stub accordingly.
Otherwise a simple tag will be added, and the category removed.
I still fundamentally fail to see the point of any of this. How does it help the encyclopedia to add tags to thousands of articles, informing people of something they can see already? --
Kotniski (
talk) 13:09, 27 November 2009 (UTC)reply
It is a question of workflow - and a significant amount of cleanup work is being driven by these categories . If you object to visible tags that is a different question - see for example
VP Orphans. That answers your question. Now the hidden assumption that Joe Random will look at an article and say "Ooops, not cited." seems to me false. Joe may never have read a learned journal, or even a book with footnotes - for a whole bunch of reasons, age, socio-economic background, accessibility, education, interest, etc. RichFarmbrough, 11:24, 28 November 2009 (UTC).reply
Well yes, I don't object to the categories, but it is the question of visible tags that concerns me. Looking at the orphans thread you link to I see I'm not alone in being opposed to the "drive-by tagging" of which this bot's proposed activity would consist. Unless there's been a proper wide community discussion that shows there is general consensus for this activity (i.e. that we opponents are a small minority) then I don't think it's appropriate for a bot to do it.--
Kotniski (
talk) 12:03, 28 November 2009 (UTC)reply
Do we have community consensus for each of the tags? --
IP69.226.103.13 (
talk) 20:26, 28 November 2009 (UTC)reply
I thought someone else would understand what "each of the tags" means and answer that question. If you mean for using unreferenced tags (none other are actually relevant here) then yes, de-facto. There are a significant number of tagged articles, the tagging has been happening for nearly five years. The principle was established on wikien-l. The template survived a TfD. So yes there's consensus. RichFarmbrough, 19:34, 10 December 2009 (UTC).reply
I have no idea why I asked this question. I appreciate the thorough answer. I must have meant to post this elsewhere. The task looks reasonable and bot-appropriate, takes care of existing problem, experienced bot operator, who responds to concerns, good community input. --
IP69.226.103.13 (
talk) 17:47, 11 December 2009 (UTC)reply
I'll note that I am likely to support the deletion of the current category. What ever is decided here should allow the code the generates the project 'Cleanup listing' pages to include these articles in those bot generated work lists. If that means replacing the category with a template, then so be it.
Vegaswikian (
talk) 20:05, 10 December 2009 (UTC)reply
The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at
WT:BRFA.