Wikipedia:Bots/Requests for approval/Yobot 47

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

Denied.

Yobot 47

Operator: Magioladitis (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 13:07, Thursday, February 2, 2017 (UTC)

Automatic, Supervised, or Manual: Automatic

Programming language(s): AWB

Source code available:

Function overview: Add non-breaking between units and numbers per WP:NBSP

Links to relevant discussions (where appropriate):

Edit period(s): Daily

Estimated number of pages affected: Thousands per day.

Exclusion compliant (Yes/No): Yes

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

Function details: For example: It will change 48 km to 48&nbspkm. The units covered are: ([cmknuµ][mgWN]|m?mol|cd|mi|lb[fs]?|b?hp|mph|ft|dB|[ap]\.m\.|[kGM]?Hz|m/s|°[CF])\b(?<!(\d?\.?\d+)mm) (all the units covered by FixNonBreakingSpaces function of AWB). It will ignore filenames. -- Magioladitis (talk) 09:36, 3 February 2017 (UTC)[reply]

Discussion

Support. A minor change, but one supported by guidelines and which changes how the page looks. Similar to changing a hyphen to an endash. ~ Rob₁₃^Talk 13:29, 2 February 2017 (UTC)[reply]
I note that the task does not include general fixes. Is the task exclusion compliant? What are the function details? — Carl (CBM · talk) 13:48, 2 February 2017 (UTC)[reply]
I will have to agree. This task seems fine, but much like many of the other Yobot BRFAs, it'd be nice if we could fill in the "function details" so we know exactly what the bot is going to do, along with the other required fields — MusikAnimal ^talk 01:35, 3 February 2017 (UTC)[reply]
- MusikAnimal I added an example. -- Magioladitis (talk) 09:34, 3 February 2017 (UTC)[reply]
  - @Magioladitis: Would you consider also covering ft, lbs, ml, kg, and mg? Omitting some due to likely false positive issues (like in for inches), but these seem safe when preceded immediately by a number and succeeded by a space. ~ Rob₁₃^Talk 11:07, 3 February 2017 (UTC)[reply]

BU Rob13 Thanks. I'll provide the full list of unit covered as soon as possible. -- Magioladitis (talk) 12:25, 3 February 2017 (UTC)[reply]

How does this fit in with WP:COSMETICBOT? That policy is saying Scripts that apply cosmetic changes, such as cosmetic_changes.py, should be used with caution and then giving removeNonBreakingSpaceBeforePercent as an example. I think the task described and removeNonBreaking can be considered very similar. Is there consensus established for this task, or will Cosmetic changes (such as many of the AWB general fixes) should be applied only when there is a substantive change to make at the same time apply; does this edit only happen when the bot is already editing and finds that this change needs to be made. TheMagikCow (talk) 13:25, 4 February 2017 (UTC)[reply]
- Until this is addressed, I am inclined to Oppose; obviously changing the !vote if it appears that the bot fits the policy. TheMagikCow (talk) 15:51, 4 February 2017 (UTC)[reply]

This task adds non-breaking spaces. Moreover, CONSENSUS > COSMETICBOT. -- Magioladitis (talk) 19:36, 12 February 2017 (UTC)[reply]

BU Rob13 AWB current regex supports the following: ([cmknuµ][mgWN]|m?mol|cd|mi|lb[fs]?|b?hp|mph|ft|dB|[kGM]?Hz|m/s|°[CF])\b(?<!(\d?\.?\d+)mm)"); -- Magioladitis (talk) 19:39, 12 February 2017 (UTC)[reply]

@TheMagikCow: A non-breaking space prevents the page from rendering with the number from a percent and the actual percentage sign on separate lines. This does create a rendered change, just only on certain screen resolutions (depending on all the text on the page). In the past, COSMETICBOT has been interpreted as requiring a rendered change on any single browser/screen resolution, not all of them. ~ Rob₁₃^Talk 19:52, 12 February 2017 (UTC)[reply]

@BU Rob13: I would certainly agree that there are many interpretations of COSMETICBOT here, the definition of Cosmetic fixes is rather loose and open to interpretation. However, I don't think this change is substantive in its nature as COSMETICBOT implies changes must be. I would consider a substantive change to be a change that appears to all users, instead of just to a small subset of users. Magioladitis I completely understand that CONSENSUS > COSMETICBOT, but you have not shown a discussion where consensus has been established, so COSMETICBOT is all we have to go on. I feel that a discussion to establish consensus for this task should be held before approval is given. TheMagikCow (talk) 10:08, 13 February 2017 (UTC)[reply]

TheMagikCow Don't you think that Manual of Style reflects consensus? -- Magioladitis (talk) 10:10, 13 February 2017 (UTC)[reply]

I think it reflects consensus for that style to be used at when making edits to a page. I can see no consensus for retrospective fixes to pages. TheMagikCow (talk) 10:13, 13 February 2017 (UTC)[reply]

Interesting approach. -- Magioladitis (talk) 10:14, 13 February 2017 (UTC)[reply]

When two policies are in conflict with each other, as it seems here with COSMETICBOT and MOS, I would like to err on the side of caution and establish consensus from the community about how to proceed. TheMagikCow (talk) 10:15, 13 February 2017 (UTC)[reply]

TheMagikCow Exactly. COSMESTICBOT is in conflict with other policies. This is the reason is subject to change. It seems many people think that this is a policy only in case no conflicting policy exists. You are right in underlying this. -- Magioladitis (talk) 11:10, 13 February 2017 (UTC)[reply]

I completely agree with the fact that this needs to be clarified. However, until then I still feel that the bot violates the policy. If it seems like many people only apply this policy when no others are present can we just sort this issue with a simple RfC? TheMagikCow (talk) 12:20, 13 February 2017 (UTC)[reply]

TheMagikCow Under this analysis, many bots violate COSMETICBOT. The bot that removes duplicated parameters, the bot that removed the Persondata template, the bot that renames files, etc. It looks like there is some unwritten rule... -- Magioladitis (talk) 12:29, 13 February 2017 (UTC)[reply]

Indeed it does appear to be. I will change to a Support here, and I will start an RfC to update the rules to avoid any future confusion. TheMagikCow (talk) 12:42, 13 February 2017 (UTC)[reply]

TheMagikCow there is an obvious confusion. I know for good. The entire process BRFA-BOTREQ-COSMETICPOT-various projects creates multiple decisions all "by consensus" but there are cases that they may conflict each other. There was a recent ArbCom about this. --Magioladitis (talk) 12:48, 13 February 2017 (UTC)[reply]

Completely agree. Also, I am adding some comments to your proposal to clarify this whole policy mess. I strongly support that proposal and the new rewording. TheMagikCow (talk) 12:50, 13 February 2017 (UTC)[reply]

TheMagikCow I know that something good will come afterall. All this ongoing discussion for a month of more it should end to something good. -- Magioladitis (talk) 13:02, 13 February 2017 (UTC)[reply]

Yeah, just me being a bit sloe to come up to speed with all of the relevant context! Sorry, for all the confusion here! TheMagikCow (talk) 14:25, 13 February 2017 (UTC)[reply]

@TheMagikCow: I don't see nearly as much of a contradiction as you appear to here. Persondata removal was supported by consensus and necessary because the template was to be deleted (resulting in red-linked templates on articles if they were left behind). A bot that renames files is necessary because our policies say we generally should not maintain file redirects at all. A bot that removes duplicate parameters doesn't exist to my knowledge due to WP:CONTEXTBOT issues, but if it did and could intelligently figure out which value to keep, it would be necessary because it often results in changes to output of the page. We've always applied COSMETICBOT as needing consensus specific to a bot to override. This has been clearly explained to the bot operator here many times in the past, so I'm rather curious why this idea is novel to them here. We've also applied COSMETICBOT as not being applicable in cases of accessibility issues, which affect small subsets of users, so I think anything that produces a rendered change (or non-rendered accessibility change) for any subset of users should be fine. ~ Rob₁₃^Talk 13:43, 13 February 2017 (UTC)[reply]

@BU Rob13: I think that is where the confusion comes from. It is how the policy has been applied. There is nothing explicit in the policy to really define what cosmetic actually is, it just gives examples. Most of the BAG people apply it in the same way - which is great, but that is no substitute for actually defining what cosmetic means in the policy. Currently, different people can have different interpretations that are equally as valid. Surely this should be avoided and consensus reached about what cosmetic actually is. TheMagikCow (talk) 14:25, 13 February 2017 (UTC)[reply]

@TheMagikCow: Yeah, there's currently some discussions ongoing to better define things in our policy. I forget where these discussions were being held, but they are ongoing. You'd probably find them if you looked in the usual places (WT:BRFA, WT:BAG, WT:BOTPOL), but I'm short on time at the moment, so can't find it myself. ~ Rob₁₃^Talk 14:27, 13 February 2017 (UTC)[reply]

Thanks Rob! I have left my thoughts at WT:BOTPOL. TheMagikCow (talk) 15:16, 13 February 2017 (UTC)[reply]

BU Rob13 [1]. -- Magioladitis (talk) 13:47, 13 February 2017 (UTC)[reply]

Also recall that Wikipedia:Bots/Requests for approval/Dexbot 6 only affected categorisation. In fact a very useful one because it allowed Wikidata-Wikipedia comparison that could be very beneficial for both projects. -- Magioladitis (talk) 13:50, 13 February 2017 (UTC)[reply]

That one should probably be challenged; consensus was not built for it. ~ Rob₁₃^Talk 14:25, 13 February 2017 (UTC)[reply]

You keep commenting even against yourself. Please stop. -- Magioladitis (talk) 19:27, 13 February 2017 (UTC)[reply]

I'm not sure what you're asking me to stop doing here. I'm commenting in support of this task. If you're asking the community (including me) to divest itself of its role in overseeing bots, then I will decline to do so. ~ Rob₁₃^Talk 04:12, 14 February 2017 (UTC)[reply]

Approved for trial (20 edits). Please provide a link to the relevant contributions and/or diffs when the trial is complete. — HELLKNOWZ ▎TALK 20:15, 13 February 2017 (UTC)[reply]

Diffs: [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21]. -- Magioladitis (talk) 21:13, 13 February 2017 (UTC)[reply]

Trial complete. -- Magioladitis (talk) 21:15, 13 February 2017 (UTC)[reply]

Trial looks good. In my opinion, a proper scope of approval here is "changing normal spaces to non-breaking spaces between numbers and units of measurement" with the usual caveat to perform proper testing after any change in code before running it automated. @Magioladitis: (k|M?p)si would be a good addition to the regex, as would mmHg. ~ Rob₁₃^Talk 04:10, 14 February 2017 (UTC)[reply]

These (1, 2) edits don't feel right - from the guideline Whether a non-breaking space is appropriate depends on context ... it may be counterproductive in a table. How are you evaluating the "context" of these characters? — xaosflux ^Talk 04:34, 14 February 2017 (UTC)[reply]

Xaosflux I can exclude tables the same way I exclude filenames. We have to make sure that this BRFA is now just brainstorming between us but that it really reflects consensus so we have to stick on the safe side. -- Magioladitis (talk) 10:11, 14 February 2017 (UTC)[reply]

The MOS is generally reflective of broad consensus - and non-breaking spaces have appear to have broad support in free text, but much less when parameterized for sections like templates and tables. Hellknowz - as you are active on this BRFA, what are your thoughts? — xaosflux ^Talk 13:02, 14 February 2017 (UTC)[reply]

Everything that could make my life easier is good :) AWB provides two levels of exclusions: Level 1: Exclude external/internal links, images, nowiki and math tags and comments. Level 2: templates, refs, link targets and headings. Level 2 is dependent on Level 1. -- Magioladitis (talk) 14:37, 14 February 2017 (UTC)[reply]

Magioladitis I didn't see anything on that about tables - will this only work for tables produced by templates? — xaosflux ^Talk 00:14, 15 February 2017 (UTC)[reply]

Xaoflux I think if we enable Level 1+2 hide then all tables will be excluded. I'll test it later. Let's see what the consensus says first. -- Magioladitis (talk)

As a side observer, I'm of the opinion that units in tables (and infoboxes) should be left alone by the bot. I'd also be uncomfortable unleashing the bot without a more substantial trial. Also I'd like the opinion of WP:MILHIST on [22][23] since gun-related quantities are often written with their own conventions. Also, a few edits would be fairly confusing to people reviewing them. [24] for instance leave psi with a normal space and [25] leaves inches with a normal space too, [26] leaves gal and g/km with a normal space, [27] leaves V DC with a normal space, etc. There's also the fact that several edits [28][29] will not change the appearance of rendered pages, unless you zoom in to ridiculous levels. Headbomb {talk / contribs / physics / books} 16:57, 14 February 2017 (UTC)[reply]

Headbomb To add to this. If we enable generla fixes it will be even better because the general fixes function also contains various exceptions to various cases, etc. -- Magioladitis (talk) 16:59, 14 February 2017 (UTC)[reply]

Oppose - turning this in to any sort of omnibus (and genfixes) approval. If AWB can't do it, then maybe it isn't the best tool for the job. — xaosflux ^Talk 00:16, 15 February 2017 (UTC)[reply]

xaosflux It can be done without general fixes. With it's going to be more interesting. Still I understand why you are not ready to approve this atm. -- Magioladitis (talk) 00:19, 15 February 2017 (UTC)[reply]

xaosflux All the test edits were done by AWB. -- Magioladitis (talk) 01:17, 15 February 2017 (UTC)[reply]

{{BAGAssistanceNeeded}} -- Magioladitis (talk) 23:12, 22 March 2017 (UTC)[reply]

Denied., per my 16:57, 14 February 2017 (UTC) post, basically. While there are some benefits, they are too occasional and incomplete to be worth doing on their own by bot, or worth flooding watchlists with. Headbomb {t · c · p · b} 20:06, 17 April 2017 (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.