Wikipedia talk:Files for discussion

This is the talk page for discussing improvements to the Files for discussion page.

Put new text under old text. Click here to start a new topic.
New to Wikipedia? Welcome! Learn to edit; get help.

WT:FFD

Archives: 1, 2, 3, 4, 5, 6, 7, 8: 31 days

RfC: Feedback on the process

The consensus to merge the WP:NFCR into "Files for deletion" led to formation of "Files for discussion" in October 2015. WP:PUF was also merged per consensus in April 2016. Now I would like your feedback on the current process, "Files for discussion". What are your thoughts about this? --George Ho (talk) 08:55, 19 December 2016 (UTC)[reply]

I think it's working fine. Do you have any concerns? Ramaksoud2000 ^{(Talk to me)} 22:30, 21 December 2016 (UTC)[reply]

@Ramaksoud2000: There have been a fair share of backlogs lately as explained above. Some participation, but most of the nominations usually have one or two voters. Also, I have to use Twinkle or manual setup to nominate every file. Also, I have to add captions notifying readers/editors about the discussion nomination on files, or no one else would be aware that the files would be deleted. Also, the images at Tales of Eternia are still under discussion since June 2016. File:Finally 12-inch Choice Mix.ogg has been still under discussion without participation since May 2016. Also, a chuck of files have been nominated for discussion/deletion in one day or another. George Ho (talk) 01:53, 22 December 2016 (UTC)[reply]

There is a definite participation problem. FFD needs more admins/commentators, but since Wikipedia is a volunteer service, I don't know of a solution. Files with no opposition to deletion are already deleted. Ramaksoud2000 ^{(Talk to me)} 02:03, 22 December 2016 (UTC)[reply]

Also, the other issue you describe is more a lack of tools than a problem with FFD. Commons has great JavaScript for nominating a file for deletion, but enwiki doesnt. Ramaksoud2000 ^{(Talk to me)} 02:06, 22 December 2016 (UTC)[reply]

Not just that, Ramaksoud2000, the file names... are not exactly memorable to type. I have to copy-and-paste the whole name. There might be other concerns, but there's a long list. Another is organizing nominations by day, not by subpage like AFD. George Ho (talk) 02:12, 22 December 2016 (UTC)[reply]

What do you mean by copy-pasting the name? Twinkle does that for me. Or are you talking about the image notice? Ramaksoud2000 ^{(Talk to me)} 03:36, 22 December 2016 (UTC)[reply]

I'll rephrase. Twinkling single-file nominations is okay. However, I must do multi-file nominations; Twinkle doesn't do that. --George Ho (talk) 03:47, 22 December 2016 (UTC)[reply]

True. Unfortunately, I don't know JavaScript to write a tool for that. Ramaksoud2000 ^{(Talk to me)} 04:11, 23 December 2016 (UTC)[reply]

You wouldn't need to write a batch tool from scratch—it would just need to be imported from Commons and activated only in certain namespaces (e.g., on category/file pages). Sometimes all it takes is a request at the technical Village Pump and someone will port it over. czar 07:18, 29 December 2016 (UTC)[reply]
I used to participate in the old NFCR much more than I do at FFD. The problem with FFD is that editors bring forth uncontroversial/unchallenged actions. There should be a guideline to be WP:BOLD first, and come to FFD only if it's an unclear case or it's challenged (similar to WP:BRD). – Finnusertop (talk ⋅ contribs) 03:26, 17 January 2017 (UTC)[reply]

Vinyl record labels (LPs)

How are we handling 2D scans of the text-only circle stickers in the middle of vinyl LP records? Are we saying they have no compositional originality? What about labels from the UK and its low threshold of originality? I know there have been a few cases for precedent, but I haven't kept track. Might be useful to have a template for these cases, both for us and for readers czar 18:29, 12 January 2017 (UTC)[reply]

Same principles would apply to these. Most cases, the wording is all text and even for the UK that would be uncopyrightable. But often you'll find production logos too and if those are copyrightable in the host country, then the label is copyrightable too as a derivative work. --MASEM (t) 20:37, 12 January 2017 (UTC)[reply]

Mass listings with identical rationales

Files for discussion nominations were mentioned at Wikipedia talk:Miscellany for deletion#Mass listings with identical rationales, disruptive by SmokeyJoe. The nominations by とある白い猫 began on Wikipedia:Files for discussion/2017 January 6, and by my estimate about 200 files have now been express closed as delete mostly by Fastily.— Godsy^(TALK_CONT) 10:03, 16 January 2017 (UTC)[reply]

Yes. I do not see the disruption. -- A Certain White Cat ^chi? 14:57, 22 January 2017 (UTC)

Batch tool

Is there a tool for batch nominations to FfD? Like commons:Help:VisualFileChange.js for enwp? czar 21:14, 16 January 2017 (UTC)[reply]

@Czar: No there is not. It might be possible to port over that script but the targets would all have to changed. DRs at Commons are all their own page (like AfD here) while FFD is part of a single subpage. There are also different templates and talk page notices but those should be easily changeable. --Majora (talk) 22:05, 16 January 2017 (UTC)[reply]

Just a note that the script can be found here: c:MediaWiki:VisualFileChange.js. There are quite a few things that need to be changed since it does a lot. Talk page notices being the big one. Perhaps asking at WP:VPT would be the best option? --Majora (talk) 22:20, 16 January 2017 (UTC)[reply]

If nobody else volunteers to do this, then ping me and I will put one together from scratch. Been busy IRL, so I can't promise I'll get around to it immediately, but I can certainly try :) -FASTILY 01:09, 18 January 2017 (UTC)[reply]

Since this was pointed out to me, I have used it quite a bit on the commons and it is a really useful time-saving tool. Fastily, if you repurpose it for here, even if FFDs were the only function that would be great. ww2censor (talk) 16:27, 22 January 2017 (UTC)[reply]

Non-free use rationale bot?

Throwing this here to get comments from file namespace folks. What are your thoughts on a bot that goes through Category:Non-free images for NFUR review and marks images as having a rationale if they have a completed rationale?

Example: Consider {{Non-free use rationale logo}}. That template is "complete" when it has its "Article" and "Use" parameters completed, as those are the only mandatory parameters. I could create a temporary tracking category to track all completed non-free use rationales for logos and then run a bot through to ensure that all transclusions of {{Non-free logo}} on those file description pages have the "image has rationale" parameter set to yes.

The goal would be to clear out a lot of the crap in the 180k+ NFUR review category to allow us to actually review non-free files without appropriate rationales. Thoughts on this? ~ Rob₁₃^Talk 23:45, 25 January 2017 (UTC)[reply]

While this could be helpful, keep in mind that the rationale is just a mechanical thing that can be added and there still needs to be human evaluation and potentially some bot-aided review (eg each use of an image has its own rational, etc.) of the rationale to make sure it is appropriate. The suggested both would do one function but its not a complete solution nor can it be by the nature of NFC. --MASEM (t) 23:49, 25 January 2017 (UTC)[reply]

@Masem: Very true, but the current situation is no review for any of those files. The prospect that we'll ever get to looking at 180k files in a timely fashion is unlikely, to say the least. I could add an auto parameter to separate out those reviewed by the bot, but the main point here would be to get the existing review category down to the images that urgently need human review (I.e. With no rationale whatsoever). ~ Rob₁₃^Talk 00:47, 26 January 2017 (UTC)[reply]

Oh, sure, it still makes sense, but just to make sure that it's not affirming that templated rationales are necessarily valid. Also in writing any supporting language for the bot, it should be clear that templated rationals are not required, just that they do require more manual review than without. --MASEM (t) 00:57, 26 January 2017 (UTC)[reply]

Yes, absolutely. This bot would never be designed to review anything in the negative (i.e. nominate for deletion, etc.) due to the absence of a templated rationale. It will just review in the positive for a templated rationale which appears to have appropriate fields completed. I like the idea of adding an "auto" parameter set to yes for everything the bot reviewed so that any interested editor could provide human eyes on the bot-reviewed files. ~ Rob₁₃^Talk 01:21, 26 January 2017 (UTC)[reply]

I like the idea, I have little experience with NFUR, having not really go into it, but it sounds like a "bot-assessed move to commons" sort of deal, with the massive backlog stopping people from trying to fix it, which makes the backlog grow unabated. Iazyges Consermonor Opus meum 04:28, 27 January 2017 (UTC)[reply]
I support the idea too. I think it is important to have the bot somehow categorize any files it has reviewed as bot-reviewed. -FASTILY 01:04, 28 January 2017 (UTC)[reply]
Add my support as well. Ronhjones ^(Talk) 14:20, 30 January 2017 (UTC)[reply]

Admin bot to delete old revisions of non-free files

Thoughts on this? An adminbot to delete all old file revisions of images in Category:Non-free files with orphaned versions more than 7 days old if there's only one editor in the file history (to filter out cases where the file is later changed to non-free, which might require admin review). It seems pretty safe to automate deleting old revisions of rescaled images marked as non-free by the original uploader. ~ Rob₁₃^Talk 04:59, 29 January 2017 (UTC)[reply]

I wonder how common that scenario is. My impression is that most files tagged for removal of old revisions were edited by 3 people, the uploader, the reducing bot and an user who tagged it for reduction. And bots and users cleaning up new uploads. I also wonder how many admins do in-depth reviews of such requests. Jo-Jo Eumerus (talk, contributions) 09:58, 29 January 2017 (UTC)[reply]

Even then, it should be pretty safe to process these via bot. There's a weeklong delay before the old revisions are deleted, and any sort of vandalism/abuse of the system is typically resolved within that timeframe. See also: here & here. Also, ping for @Ronhjones who originally suggested the idea. -FASTILY 00:44, 30 January 2017 (UTC)[reply]

I see no issue here, sounds like a good idea. -- Iazyges Consermonor Opus meum 02:22, 30 January 2017 (UTC)[reply]
Obviously works for me! Typically we don't delete old versions, we use User:Legoktm/Frescaled.js script (since it takes way too long manually!) which uses revdel to hide all the oversize images (there may be more than one!) and remove the template. This approach therefore still shows the file history. If there is an issue with the reduction, it's not that difficult to revert. I would rather all big files be reduced and manually revert back 0.1% rather than have 99.9% too big (and I think 0.1% could be overestimated!) I'll ping @Diannaa: as she has an interest in this field.Ronhjones ^(Talk) 14:19, 30 January 2017 (UTC)[reply]
I have no objection to a bot doing this task, as like Ronhjones, I have discovered that the number of cases that actually require human intervention is vanishingly small (0.1% is probably overstating the error rate by quite a bit). Current practice is to use revision-deletion of the old revisions rather than outright deletion, as this helps ensure the correct person gets notified if/when the file gets nominated for deletion as an F5 or for some other reason, because the upload history is visible on the front-facing file description page rather than hidden in the back. — Diannaa 🍁 (talk) 14:29, 30 January 2017 (UTC)[reply]
@Sphilbrick: you have done thousands of these lately; what's your opinion? — Diannaa 🍁 (talk) 14:32, 30 January 2017 (UTC)[reply]

I'm traveling - due to arrive tomorrow. Probably supportive, but would like to give it some more thought tomorrow when I am not brain dead.--S Philbrick (Talk) 01:17, 1 February 2017 (UTC)[reply]

I asked DeltaQuad if she could handle this, and she's looking into it. ~ Rob₁₃^Talk 02:22, 31 January 2017 (UTC)[reply]
Support As Diannaa notes, I've processed quite a few of these and whenever doing so, vaguely wondered why it wasn't a bot task. My supposition is that it used to require more thought early on, but that's not important, the important fact is that there are a lot of these, the error rate is exceedingly low, and if a mistake occurs it can be undone. Like Jo jo, I would not restrict it to a single editor. A single editor is quite rare. Most involve a bot but many involve a human doing a reduction, so one cannot even limit it to a single editor other than bots. It would be nice to ensure that the bot is either dedicated to this task, or configured so that if a problem occurs, it is easy to identify the history to undo. --S Philbrick (Talk) 12:50, 2 February 2017 (UTC)[reply]
Support per above, the error rate is low and it is normally those with more than one human editor involved that cause the problems - especially when it's not a reduction but a new version of the image. Bollywood movie posters are where I see edit wars most, over which version and rather than argue of which of two files the edit war is over which revision of one file. Nthep (talk) 20:40, 2 February 2017 (UTC)[reply]

Google Maps screenshots

I stumbled on File:17 High Park Road.png this today which is being used under fair use. To my mind this image is easily re-created by simply going to the location in question and taking a free photograph. There's nothing to indicate that this building is no longer there (i.e. that a new and free image would be difficult/impossible to get). I was wondering if someone could give me some advice on whether these kinds of FU images are legitimate. The Rambling Man (talk) 07:09, 9 February 2017 (UTC)[reply]

It is exactly as you say, The Rambling Man. Fails WP:NFCC#1. – Finnusertop (talk ⋅ contribs) 07:12, 9 February 2017 (UTC)[reply]

Thank you for your quick response! The Rambling Man (talk) 07:13, 9 February 2017 (UTC)[reply]

There seems to be a pretty liberal FOP for buildings in the UK according to c:COM:FOP#United Kingdom. So ,if the residence is still standing, then no reason why a freely licensed equivalent cannot be created by someone. Even if the house has been torn down, the use would fail WP:NFCC#8 because there no sourced discussion of the house itself in William T. Stearn#Retirement (1976–2001) to provide the context required by NFCC#8. On a minor note, the way images are being referred to in the article and captioned also seems to be atypical. Is it common to number images as "Fig. 1", etc. and then refer to them as such in the article content? I don't think I've seen that before. -- Marchjuly (talk) 08:00, 9 February 2017 (UTC)[reply]

No, that's unusual, although I have seen other such "Table 1" etc captions used, but it does make maintaining the article more difficult as removing one of these necessitates a fix to the text and then a possible renumbering of subsequent figures. I've notified the main author of the page of this discussion. The Rambling Man (talk) 08:04, 9 February 2017 (UTC)[reply]

Regarding File:Agnes Arber circa 1916.jpg, if the date can be verified to be correct then it is likely old enough for WP:PD either as {{PD-old-100}}, {{PD-UK}}, {{PD-UK-unknown}} or some other reason per c:Commons:Hirtle chart. If the file can be converted to PD, it will no longer be subject to Wp:NFCCP and can be re-added to the article. -- Marchjuly (talk) 08:24, 9 February 2017 (UTC)[reply]

That's helpful to know, thanks. Once the author gets back onto this, I'll make sure he's aware of these subsequent discussions. The Rambling Man (talk) 08:30, 9 February 2017 (UTC)[reply]

Proposals related to file patrol

Please see Wikipedia:Village_pump_(proposals)#Backlog_of_unpatrolled_files for two proposals related to file patrol. Cenarium (talk) 21:22, 11 February 2017 (UTC)[reply]