Wikipedia:Bots/Requests for approval/RileyBot 9

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 The result of the discussion was Withdrawn by operator.

Operator:

Programming language(s): Python

Source code available: N/A

Function overview: Looks at every page that transcludes {{Infobox NRHP}}, whenever it finds the | built = parameter, followed by a four-digit number, it encases the number in {{Start date}}. Whenever it doesn't find the parameter, or when it finds the parameter empty or followed by something other than a four-digit number, it does nothing.

Links to relevant discussions (where appropriate): Wikipedia:Bot_requests#Mark_a_lot_of_pages_for_microformatting.2C_round_2 and Wikipedia_talk:Bot_requests/Archive_2#RFC:_Deploying_.27Start_date.27_template_in_infoboxes

Edit period(s): One time run and whenever requested

Estimated number of pages affected: No clue, 1000+ (there is 40,000 pages that transclude {{Infobox NRHP}})

Exclusion compliant (Yes/No): Yes

Already has a bot flag (Yes/No): Yes

Function details: As stated in the function overview, make a list of pages that transcludes {{

Approved for trial (50 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. MBisanz ^talk 04:49, 29 March 2013 (UTC)[reply]

• Trial complete. Sorry, accidently did 60 edits. -- Cheers,
  ey 06:05, 29 March 2013 (UTC)[reply
  ]
  • I checked a random sample - all look good to me. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 12:38, 29 March 2013 (UTC)[reply]

Comment I believe this is the bot requested at

Followup About the sample, it is highly NON-random, being the first alphabetical items in the list of NRHP listings, so happens to be all items starting with numbers. And it omits, relatedly, all churches. There are many thousands of NRHP-listed churches, many started/founded at an earlier date than some later NRHP-listed building was built, e.g. many at First Presbyterian Church (disambiguation), for example.

For one of the items done by the bot, I have just removed the Start date template and corrected information in the article, in this diff. There, the NRHP infobox field was incomplete, showing 1929 when 1929-1931 would have been the correct information. It has not been settled what a start date should be, when a construction period spans years. Usually, perhaps most often, the NRHP infobox "built" field is incorrect in articles, for such cases. A category also was incorrect, stating that the building was completed in 1929 when it was completed in 1931, per corrected interpretation of the NRIS information.

I expect there are other problems in the first 60 done, but there would be more/different problems in a truly representative sample. I request that this go back to the bot request stage, or a general RFC specific to NRHP applications. The bot proponents have cited so-called community consensus on a previous RFC which was conducted with no mention and no awareness by any participant about NRHP application complications. --doncram 18:20, 29 March 2013 (UTC)[reply]

--doncram 18:20, 29 March 2013 (UTC)[reply]

Doncram, The randomness was in my selection from the articles used in the test; once again you've taken something out of context and made a molehill into a mountain. If infoboxes - or categories - contain bad information that's not the fault of this bot, the date template, or this proposal; and the errors can be corrected regardless of them. You have already been told taht an RfC has approved this task for all infoboxes with these datable microformats. You disrupted a previous version of this request with long screeds of opposition, demonstrating no understanding whatsoever of what was proposed or its benefits. Please refrain from doing the same here. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:47, 29 March 2013 (UTC)[reply]

The items done by the bot are NOT random. Look at it, those are the entries with names starting with numbers, and then a first few A acronym ones. That is not random selection from any list.

You, Pigsonthewing, brushed off valid objections in the previous bot request. At the current bot request, I ask the quite serious question to you, of whether you care or not about achieving accurate start date information. Seriously, there are some thousands of churches, schools, ships, libraries, other articles having NRHP infoboxes, where the infobox built date, even if accurate for its purpose there, is not clearly the relevant start date for the church, school, ship, library, etc. And is obviously the wrong date in the context of probably hundreds of these articles. That is different from the also-vexing problem of inaccurate content of the built date field. --doncram 19:15, 29 March 2013 (UTC)[reply]

Your first point is correct, but as no-one has claimed that the items done by the bot are random, tilts at windmills. Your tl;dr objections are not valid, and are about content, not this bot activity or the template used. Take them to a relevant venue. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:21, 29 March 2013 (UTC)[reply]

The bot says the 1935 Labor Day hurricane started in 1937! In this bot edit. Because the NRHP infobox correctly gives 1937 as the built date of a memorial covered in the hurricane article. Fairly often NRHP designations are included in articles in a secondary way, as here. This is not a "content error" in the NRHP built field, which is accurate. This is an error introduced by the bot.

The bot has not been approved by any general consensus, for the NRHP infobox application. A general RFC concluded that bot runs would have to be considered on a case-by-case basis: Wikipedia:Requests for comment/Microformats, as brought up by HellKnowz in Wikipedia:Bot_requests/Archive_51#Break for convenience. Pigsonthewing replied then with reference to anohter RFC, also not recognizing NRHP infobox content. I don't think it is right for a bot to march through 30-40,000 NRHP articles adding frequently-erroneous information. To the bot programmer, too, do you want that? --doncram 13:59, 30 March 2013 (UTC)[reply]

Your claim that "the bot says the 1935 Labor Day hurricane started in 1937" is false, as the subject of the infobox in which it was made (and therefore the name parameter of the relevant microformat) is "Florida Keys Memorial"; and you refer to the wrong, earlier RfC, which was about a different subject; and not bot runs. The bot in this BRFA will not add any information; it will mark up the information which is already present. Your continued objections, based on your false assumptions and gross misunderstandings, are disruptive. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 14:53, 30 March 2013 (UTC)[reply]

I may in fact have an incorrect understanding or two, but this would be the result of inadequate, rushed discussion to push through something not ready to be implemented. I do indeed believe that the "start" designation gives the microformat information that the topic of the article title, not the topic of the infobox title, was started in the given year, and Pigsonthewing asserting I am wrong in that, only here, does not convince me otherwise. If this were to be discussed properly I would certainly like to see clarification about this. And, there are many cases where the NRHP infobox, referring to a newer building, has the same title as the article title, and where microformat labelling would seem to imply the wrong thing. I am trying to stand up for accuracy of information in the NRHP articles, and I don't understand the nature of the accusations here.

I do feel the discussion here, and before, is being railroaded. At each of several discussions (one at wt:NRHP, one or two at botrequests, the BRFA, it has been asserted that everything was discussed and resolved elsewhere, and that this forum too is not the right forum to have any discussion. But, IMO, there has not been adequate discussion and consensus by NRHP editors of what the "start" template is supposed to do, and how NRHP editors should support it, if they should. There are a number of unanswered questions, about what to do when the NRHP-listed building is secondary within an article addressing a broader topic, etc. What about multiple NRHP infoboxes in one article? I don't see the urgency to run a bot which will probably introduce a lot of factually inaccurate information. Which could be improved by some delay, discussion, and improvement of the NRHP infobox and of editors' understandings and of the bot run to be implemented (e.g. maybe leave out all church articles, or leave out all articles having more than one infobox, or other changes to a bot run). --doncram 22:00, 31 March 2013 (UTC)[reply]

You now attempt to blame other people for your lack of understanding; and go on to repeat it. That won't wash. I repeat: The bot won't introduce any information, accurate or not. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:13, 31 March 2013 (UTC)[reply]

And you repeatedly have said you are not reading the comments posted with derogatory "tl;dr" statements, and your comments seem to me to show continued lack of understanding on your part (e.g. that the objections are not just about the fact that the built data field is sometimes incorrect). This does not seem like a productive discussion here.

I do think the bot should be disallowed, because there are open questions, not yet addressed civilly. --doncram 23:21, 31 March 2013 (UTC)[reply]

Your questions have been addressed civilly; just not repeatedly. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:34, 31 March 2013 (UTC)[reply]

Civilly, no, the questions have not been answered, and your assurance that they all have been answered is not helpful. There is a way in which repeated assertions like yours can seem incivil, too, in that you are completely dismissing / disregarding another editor's good faith views, without explanation, without justification. That can be regarded as incivil. Thanks. --doncram 00:08, 1 April 2013 (UTC)[reply]

OK, Doncram, let's talk about churches, because I've been working on several of them in Minnesota. First Baptist Church (Saint Paul, Minnesota), for example, is listed as being built in 1874. The building is listed on the National Register of Historic Places. I think it's perfectly legitimate to list the start date as 1874, because the article is about the building, not simply the congregation. If you want to distill down the actual start date of the congregation, does it date back to 1847, when Harriet Bishop came to Minnesota? Or to 1849, when the church formally organized itself? Or to 1851, when they built the first church on Baptist Hill? That sort of thing gets nebulous. But, you know what? We have those facts in the article. We are not lying to the reader when we say the present church building was built in 1874. All we're doing is adding a machine-readable standard format to the article.

Similarly, look at the

Cathedral of Saint Paul, National Shrine of the Apostle Paul. Now, if we added {{start date}} as 1841 to the article, as Doncram proposes, one might wonder why such a grand cathedral was built in 1841 when the settlement was still called "Pig's Eye". (Well, maybe travelers landing at Minneapolis-Pig's Eye International Airport might not be confused, but I digress.) Again, if you want greater accuracy in an NRHP article, do more research, find more sources, and put that information into the article. Humans can easily read the text and know that the building is the fourth cathedral in Pig's Eye -- er, St. Paul. Again, we aren't lying to the reader when we say the building was built in 1904. Now would you please stop being disruptive with the whole start date discussion? --Elkman ^(Elkspeak) 20:10, 30 March 2013 (UTC)[reply

]

This seems like badgering, and putting words that I did not say, into my mouth. I am sure reasonable people, in a reasonable discussion, could come up with some guidelines about when the "start" date should be given as a date of founding of a church, vs. when it should be the beginning-of-construction date of a church, vs. when it should be the end-of-construction date, or other options. It is certainly possible that the correct "start" date to convey will be different than the NRHP built date in many cases. I expect this is too adversarial-seeming, here, to have that discussion here. I think it has been asserted that conversation should not happen here. --doncram 22:08, 31 March 2013 (UTC)[reply]

I see. Well, I'm not a reasonable person, nor am I participating in a reasonable discussion. I'm just a knuckle-dragging idiot who's too stupid to know the start date of anything. You've told me just as much on several other occasions. --Elkman ^(Elkspeak) 01:20, 1 April 2013 (UTC)[reply]

• Could we try to keep this on topic please? :) -- Cheers,
  ey 16:40, 1 April 2013 (UTC)[reply
  ]
  • More importantly, can we get on with this task? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 22:57, 2 April 2013 (UTC)[reply]
    • No, don't do that. For one reason, this bot run is entirely unnecessary as the effect of the bot can efficiently be done by a simple edit to the NRHP infobox code, without disrupting 30,000 articles. For fuller discussion, I've opened an RFC. --doncram 01:41, 7 April 2013 (UTC)[reply]

Please see Wikipedia:Requests for comment/Start date in NRHP articles, about running a bot to implement "start date" and "end date" microformatting into NRHP infoboxes. --doncram 01:41, 7 April 2013 (UTC)[reply]

Can we please have BAG either approve this bot now because it does what it should, or disapprove it because it's misbehaving? The above discussions are completely unrelated to the question of whether the bot will do what it's meant to do, and as has been noted here and elsewhere, consensus has already been established to have this template added in this kind of situation. Nyttend (talk) 02:58, 8 April 2013 (UTC)[reply]

BAG looks at both the technical accuracy of the bot and whether the task has community consensus to be performed. I understand there is past (and possibly present) consensus for the task, but it does appear that an RFC has been started on it. It would generally be premature to approve a bot if an RFC is ongoing regarding the task. MBisanz ^talk 03:59, 8 April 2013 (UTC)[reply]

So this is how a lone editor can effectively veto a BOTREQ. How many more hurdles must we jump though before this task is done? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 16:19, 8 April 2013 (UTC)[reply]

The burden is on the bot operator to show a task is technically sound and consistent with community consensus. An active RFC in which two people support the task and two people oppose the change is not consistent with consensus. Please wait for the RFC to be closed by an uninvolved editor with a result in favor of the change before seeking approval. MBisanz ^talk 21:33, 8 April 2013 (UTC)[reply]

You ignore the fact that we've already had a successful RfC with overwhelming support for applying this fix across all relevant infoboxes. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:40, 9 April 2013 (UTC)[reply]

MBisanz is not ignoring any facts, he is taking everything in to consideration but this task cannot be approved with the open RfC at hand. So until the RfC is closed, I ask that no more comments be made that involve "jumping hurdles", how there has been a previous RfC, etc. (any comments about how the bot works are welcome of course). Until the RfC is closed, this request is

ey 19:22, 9 April 2013 (UTC)[reply

]

"MBisanz is not ignoring any facts" Really? Where does he address it? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:58, 9 April 2013 (UTC)[reply]

I'm not sure what facts I'm ignoring. The policy that controls my actions states:

Once the request has demonstrated its conformance with the community standards and correct technical implementation, the BAG may approve the task. The BAG may also decline a request which fails to demonstrate community consensus to perform the task. There is an active RFC that states: Should articles with U.S. National Register of Historic Places be bot-tagged . . . The RFC has not been closed as a bad-faith or otherwise improper action. While such an RFC is pending, it would be improper to approve the bot because such an RFC precludes the operators from being able to demonstrate community consensus to perform the task, as that is what the RFC itself will demonstrate. An RFC isn't required to demonstrate consensus at BRFA, but when one has been started and is ongoing, BAG cannot ignore it. MBisanz ^talk 11:19, 11 April 2013 (UTC)[reply

]

As I thought I'd made clear above; "You ignore the fact that we've already had a successful RfC with overwhelming support for applying this fix across all relevant infoboxes. ". It is therefore already demonstrated that this request is "in conformance with the community standards". Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:32, 11 April 2013 (UTC)[reply]

New RFCs can change the results of old RFCs if they demonstrate that consensus has changed since the old RFC. We can't ignore such a possibility and must act in a conservative manner to avoid approving bots that lack support. Assuming the new RFC closes in favor of the same course of action as the old RFC, the bot will be approved. MBisanz ^talk 12:03, 11 April 2013 (UTC)[reply]