Wikimedia Commons

Alexis Jazz

Joined 13 October 2007
This page may contain some profanity. If you are PC Principal you probably shouldn't be reading this page. Try going here instead. Everyone else should probably take this essay with a grain of salt.

Or a truckload.

🖕


Babbel userinfo
en
This user is rumored to understand English.
nl
HEB IK WAT VAN JE AAN OFZO?
de
Ah Scheiße
Users by language

Commons has governance issues.

(this bit will be removed once the issue is fixed. is it still here? guess it's not fixed then)

Admins are appointed by a mostly democratic process. No big issues there. But what if the community loses faith in an admin over time? Well we're kinda screwed then.

Do not become this again

Admins are appointed for life. Most countries recognize there should be some limit (China sucks), term or other way to remove people from office. Not on Commons though! An admin has to severely abuse their powers or piss off the WMF to be desysopped. Community consensus resulted in a requirement of a 50% majority to desysop an admin. Well, that's acceptable, isn't it? It would be, but to actually launch the desysop request you need "some consensus for removal" which is generally interpreted as roughly 75% or even more, otherwise the request can be closed by a bureaucrat as inadmissible.

De-adminship is massively crippled by this. This de-adminship analysis shows we had 15 successful de-adminship requests in 11 years (with 200+ admins that's already ridiculous), 10 of those were formalities and 4 had no proper consensus. For INeverCry, there was the opposite of consensus to start a request. In Jcb 2 it was even mentioned, and Rd232 said something sensible:

Taken literally, that badly worded and vague policy requires an informal deadmin discussion hardly less voluminous and unpleasant than the formal one, in order to establish consensus for launching the formal one. The spirit of that policy is clearly to ensure that there is discussion of issues before launching into a deadmin procedure. You can't say that hasn't happened, given the prior deadmin discussion and discussions around the place as well directly with you. If that policy means anything, it's that frivolous requests may be swiftly shut down by bureaucrats; that hasn't happened, partly because it's very clear that the concerns are widely shared. Frankly, the only thing that taking the policy literally will achieve is Commons:Administrators/Requests/Jcb (de-adminship 3) being opened in short order, with this discussion being evidence of "some consensus for removal".

— Rd232
Bureaucracy

But today we are here. Bureaucrats refer to this line in policy (which was boldly added in 2007) and as a result, holding admins accountable is incredibly difficult and creates loads of drama.

There is some hope though. De-adminship (without the admin in question socking, wheel warring or obviously vandalizing the wiki) has been proven to be not impossible. The long period of rather intense drama that it took to get there was disproportionate, but it happened. This means the system is merely unbalanced, not broken.

Wikimedia developers, horny for features, less interested in fixing broken design edit

Ooooooh shit you made it to my user page! A highly questionable honor. You're in my crosshairs now.

You know how they say "if you want a lawnmower, don't strip down a car to its bare essentials and stick dangerously sharp blades to it, just buy a fucking lawnmower"?

What, no? Just me? Well fuck you, I got the coolest lawnmower on the block. That was the story of flow. If you want forum software, use forum software for fucks sake, don't hack up a wiki.

And after "Structured Discussions" (Flow), we now face "Structured Data", in what may be the least structured implementation of anything, ever:

File captions are the first bit of structured data that has been activated on Commons. So what's the problem with it?

..

What isn't the problem with it!

  • Which languages are shown depends on what you entered in your babel box. Yes, really. It's not a preference. Good luck figuring that out. This is seriously dumb. This behaviour was inherited from Wikidata.
  • File Captions generate a huge white box on every single file page, pushing down the {{Information}} template. Exactly how huge depends on the number of languages in your babel box, but even with a few languages it's massive due to ridiculous padding. The community had to come up with Compact Captions to fix this.
  • From launch (10 January 2019) to 13 February 2019, users were not informed about the license for the captions they enter. With some goodwill, these could be considered BY-SA 3.0 (like wikitext, despite the captions box not showing any information), but that was not good enough for the developers. They insist the captions must be CC0. Which is impossible, because the developers don't own the rights. They claim it's {{PD-ineligible}}, which is questionable for the US and downright wrong for many other countries, but PD-ineligible does not translate to CC0.
  • Many descriptions get copy-pasted from the description field. It was even proposed to mass-copy descriptions using a bot. (an idea that was swiftly put to rest) Many of those are licensed as Creative Commons ShareAlike. Not CC0.
  • File captions become duplicates of file descriptions. The file captions can't be inserted in Wikitext. Maybe later this year. So duplicating information is unavoidable.
  • There is NO USE CASE for this shit. None. Zip! Bupkis! All suggested use cases are imaginary.
  • They were misdesigned. There is no single caption that fits an image in every context. When an image is added to an item on Wikidata, file captions can be entered there, and that makes sense. It allows you to provide a caption for that image in the context of that Wikidata item. So why didn't we just stick with that?
  • File captions are not allowed to contain wikitext, so even if you close your eyes and wish away all of the above, virtually no wikiproject will have any use for these.

So what can we learn from this pile of dog poop?

  • Work more closely with the community. Get the community excited for a new feature before deploying it.
  • Show the feature in action on Beta Commons or a similar environment. Listen to criticism and adjust accordingly.
  • If possible, deploy it for a limited user group first.
  • If you can't explain what benefits your feature will bring to the community, you fucked up. Back to the drawing board.
  • Make sure the feature has a use case from the very start.
  • If at all possible, make sure it can be disabled for those who don't want it.
  • If you're going to do anything with licensing, get WMF legal (preferably multiple members of the legal team) to make public statements that they fully understand and support what you are doing. Licensing can't be an afterthought like it was for file captions.

Here's the seriously worrying bit. File captions are just the start of structured data. Oh no, this ain't over. And if this was so incredibly poorly thought out, it doesn't bode well for the rest of it. Seemingly, this crap is being made by people without a vision. Hey, I'm available. Just ping me or something.

Other issues edit

Things on this page may not be related to each other. If you'd rather look at pictures you might enjoy Great images with captions.

Chipping away at images edit

If a new version of a file is uploaded that is not obviously worse, nobody cares. Compression could be increased, file headers invalidated, poor crops be done or micro-graffiti added somewhere in an image. Especially with a vague comment like "improved version" or "better source" these don't get noticed. If it does get reported eventually, the report will go away without administrator action in a few days.

Images removed edit

There is an easy way to remove images. Add {{subst:npd}} to an image and 9 times out of 10 it will go away after just one week. Especially if the uploader hasn't been active on Commons for a few months or more. If a user removes the nonsense template, an admin will put it back.

Scans of coins and similar objects could be removed by finding a non-free scan of the same object that is older than the file on Commons. The file on Commons could be nominated for deletion by claiming COM:DW.

It's also possible to nominate a file and linking any other place on the web where it occurs. When a file is marked {{copyvio|source=URL}} the odds of it being removed are even bigger. This will also not trigger a DR.

Images kept edit

If an image can't be found with Google image search (search by picture) or TinEye and has no visible watermark it is very unlikely to be removed. Social media photos often can't be found this way. A copyrighted photo that is photoshopped can also be hard to find.

Breaking descriptions edit

It's possible to remove (parts of) the description on a page, then remove the license information, then put the description back and add {{subst:nld}} to the page. Either the image will be removed or the description will be broken by an admin after 7 days.

Useful selfies edit

Commons doesn't have a lot of typical selfies from reliable sources. Whenever we do get one, they likely get deleted as "unused personal photo". Photos of journalists may also be removed as "unused personal photos". Just any unused photo of people who are not famous in the US may be nominated for deletion as "unused personal photo". If you have never heard of them, they can't possibly ever be notable!!

Closed file formats edit

 
Unhelpful.

When trying to upload a video with .mp4 extension, the upload wizard bluntly informs you it's not gonna happen. You shot the video yourself? It's a video you downloaded from YouTube that is actually in the public domain? Don't care, get lost. Sure, the wizard could offer you to convert the file to an open format that is allowed on Commons. Maybe link to video2commons. But this is much easier!

We have the same problem with .heic files. (Phabricator ticket)

Stressing users to lie edit

When media is uploaded after (local) midnight and the user enters the correct date, the upload wizard will tell the user that "The date that you selected is in the future.". The user must lie about the publishing date to make this message go away. It can be ignored, but it's just wrong.

Crosswiki uploads edit

Shit.

OTRS is broken by design, the lawyers did it edit

Very broken.

Templates and release generator edit

There is a Release Generator that works quite well. (although it's only available in English) But various places (including the {{No permission since}} template!) are still referring to mail templates. Increasing the workload of the OTRS team and reducing the chances of ever getting proper permission.

Right holders are expected to contact OTRS edit

First of all, OTRS can't contact right holders. So either right holders need to contact OTRS all by themselves (sure..) or a user needs to ask them to contact OTRS. I did that a couple of times, I've never seen anyone else doing that. You would get a lot more response if OTRS could actively contact right holders, but for some stupid legal reason they can't. WMF is afraid it would look like WMF is contacting them and OTRS is somehow legally not WMF.

You still with me? No? Good.

So what is the difference between that and a user contacting a rights holder asking them to send permission to OTRS? Slim to none. When they are sent to wmflabs.org, wouldn't they still think they are dealing with WMF? Yes, they would! But legally, it's watertight.

The problem is WMF is technically just a hoster, not responsible for the content uploaded. YOU are. More on that later.

And this process is completely fucking invisible edit

So when you ask a right holder to contact OTRS, what happens next? Nobody knows! You'll never know if they even tried to contact OTRS if the permission is not obtained. If you contact OTRS, how long will it take to get a response? Your guess is as good as mine, but if you know where to look you would find the current backlog is 7 days.

 
Is this a WMF lawyer?
 
This has got to be one.
 
Two WMF lawyers! (inside the cave)

Where the lawyers at? edit

With all these complicated copyright issues, you need some lawyers right? They are here. Nobody knows where exactly, or how to contact them. They are like Bigfoot. Everybody knows Bigfoot exists, because everybody knows somebody who claims to have seen somebody who saw Bigfoot. The result of this is that when a legal decision needs to be made, it will be made by an administrator who may or may not be able to read.

Upload to Commons, get sued edit

As you just learned, WMF is just a hoster. YOU are responsible. So okay, maybe you upload a logo that you later find out is likely copyrighted by UK law. So you request for it to be deleted. The administrator decides UK law can kiss his ass and keeps it. Now the uploader is at risk of a lawsuit, even though they did nominate it to be deleted. But uploaders have no "delete" button.

Legally, WMF may not have a problem. As long as the image in question is covered by either fair use or not eligible for copyright in the US, WMF can host it. Commons does not want fair use content, but the judge won't care about that. Its origin doesn't matter. What may matter is the country the uploader performed the upload from:

“You have to respect the laws of the country you upload from, that's why we have the "country of origin" requirement, as 99% of works are uploaded from their country of origin (random figure, but I would be surprised if that's not the case).” - As said by an admin

About 87% of statistics is made up, 72% of that is made up poorly. So you upload something from Italy while you are not in Italy or the US? Just uploading a photo that was already here which you only retouched also counts! You are doing something the administrators did not think could ever happen and you could be at risk of a lawsuit.

In Soviet Russia, copyright owns YOU! edit

Law of the Republic of Belarus No. 262-Z of May 17, 2011, on Copyright and Related Rights: "Objects of copyright are: literary works; dramatic or dramatico­musical works; audiovisual works; works of architecture, municipal engineering, and landscape development; works of photography; (yada, yada, yada); other works.'

Other works! (why did they even bother making the list?) Basically you can literally copyright a turd in Belarus. A mother should technically, when interpreting the law literally, be able to copyright her own child. NEVER tell a mother childbearing is not work unless you're suicidal! This shouldn't be too surprising considering they also have an extensive law to patent plants. I'm not saying Monsanto send over free hookers to politicians and then blackmailed them with footage from hidden cameras. How could I possibly know how exactly they would do such a thing?

Pepe the frog can officially go fuck himself. edit

Pepe the frog is a cartoon that is under copyright. We generally remove those.

IS IT CUZ I IZ BLACK????

Donald Duck, Mickey Mouse (YOU'RE NEXT), Bugs Bunny (YOU CAN RUN BUT YOU CAN'T HIDE), Frog. WE DON'T CARE. But Pepe has been adopted by the alt-right movement, and boy, it brings out the very best in humanity!

I was considering to start a discussion on Commons to see if Commons should maybe allow a bit more. After this, I will not be starting that discussion anytime soon now. So instead, I'll just start patrolling to make sure that goddamn frog stays off Commons until the creator sends their permission to OTRS. Which will never happen because the creator has publicly expressed his dismay at Pepe being used as a hate symbol. So really, I'll be doing the creator a favor.

DIOS Y LIBERTAD, YOUR BELOVED LEADER, ALEXIS JAZZ

Retrieved from "https://commons.wikimedia.org/w/index.php?title=User:Alexis_Jazz&oldid=674717575"