User talk:Zhuyifei1999/Archive 38

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

21:55, 11 June 2018 (UTC)

Thank you for undeleting the other two uploads of the uploader of the subject file for discussion. Please consider treating the subject file the same way.   — Jeff G. ツ please ping or talk to me 13:35, 17 June 2018 (UTC)

Considering the DR has already been started, right now I'm kind of tired of creating or participating in more drama. I'll just mention the file in the DR and see what the community wants --Zhuyifei1999 (talk) 14:34, 17 June 2018 (UTC)

21:47, 18 June 2018 (UTC)

Removing a user from the Flickr bad author list

Commons:Administrators' noticeboard#Please remove The Huntington from the Flickr bad author list

Could you remove 59832923@N02 (The Huntington) from the list, at the very least temporarily, so I can use the bot to check if the license has changed? - Alexis Jazz ping plz 18:32, 21 June 2018 (UTC)

Thanks for pinging Steinsplitter. I was wondering: is there a separate list for Flickr users that the bot should review (just to verify the license on Flickr because that's not interesting for humans) but also adds a category for human review because the account is known to have FoP/DW or other issues sometimes? This would be more useful for The Huntington. It would also be useful for barbarabrannon (76343652@N04) who usually shares own work on Flickr, but a while ago shared a bunch of book covers and photos of their writers. She had permission to share them for free (as in beer), but it was highly questionable if she actually had permission to share them as Creative Commons. Also she seemed to have permission from the writers, not from the photographers/cover designers. Her uploads of own work could also use a FoP check. (because US) So her uploads should be given a human review, but calling her a "bad author" isn't really justified. Flickr2commons doesn't really need to block these accounts either, but it would be helpful if it warned the user for these accounts and not allow "blind" uploads. (mass uploading by uploaders who do not inspect every single picture) - Alexis Jazz ping plz 22:37, 21 June 2018 (UTC)
Currently there isn't such a list. However, 'bad author' don't imply a final 'failed', LRs are free to override that, if the image is proven free. --Zhuyifei1999 (talk) 05:38, 22 June 2018 (UTC)
I know they can override it, but in some cases (generally when there is good faith on the side of the Flickr user but images need a human check for things like FoP/DW) it would be nice if the bot would turn them down a bit less harshly and handy if Flickr2commons wouldn't completely reject them. Flickr users that upload 100% copyvio get the exact same treatment as Flickr users that sometimes upload bad images that could easily be filtered by the uploader if they don't blindly import their whole Flickr stream. - Alexis Jazz ping plz 06:13, 22 June 2018 (UTC)
'Sometimes'... flickrwashing? No, they should stop flickrwashing altogether. If an LR uploader is sufficiently confident that an image is indeed good, imo F2C should allow them to override the blacklist. --Zhuyifei1999 (talk) 06:49, 22 June 2018 (UTC)
In an ideal world they should never Flickrwash again, but copyright is complicated. If an account uploads 90% own work and 10% photos with issues (FoP, DW, insufficient permission by our standards) it should not be possible to blindly upload their whole photostream and FlickreviewR should not blindly without any note accept their photos. But for barbarabrannon for example, the only checks really needed are for FoP (US) and things that are not own work. (book covers or anything that looks like she didn't take the picture herself) Generally there would be no need for the human reviewer to even visit the Flickr page if the bot tags them with the license that it finds on Flickr. But it's just an idea. An option (at least for experienced users, perhaps for everyone with some sort of limit/warning) to override the blacklist when using F2C sounds sane. - Alexis Jazz ping plz 07:07, 22 June 2018 (UTC)

The bot doesn't delete some files

In the last few days I have seen this a few times. It seems that the bot doesn't delete files with embedded executables which are not recognized as self-extracting archives, e.g. File:PIC FIL.xcf or File:Peazip 2018.png. Is it intentional? Because it seems to be theirs new way of uploading warez. --jdx Re: 23:12, 24 June 2018 (UTC)

@Jdx: Could you check what kind of (format) self-extracting archive it is? I could blacklist PE format altogether though. --Zhuyifei1999 (talk) 03:09, 25 June 2018 (UTC)
I have hesitated to do it because at the moment I do not have a sandbox PC. Anyway, it seems that both files mentioned above contain 7-Zip self-extracting archives created by 7-Zip ver. 18.05. --jdx Re: 04:30, 25 June 2018 (UTC)
Ok I'll try to implement 7z and zip in Magic detector & also blacklist PE format. --Zhuyifei1999 (talk) 04:32, 25 June 2018 (UTC)
Did the latter (blacklist). The former needs some time to investigate on the formats. --Zhuyifei1999 (talk) 04:38, 25 June 2018 (UTC)
Note to self: 7z archive offset 0x00032400 --Zhuyifei1999 (talk) 04:58, 25 June 2018 (UTC)
I think I found a programmatic way to find that offset (of Ending detector) for PIC FIL.xcf. @Jdx: Is there some self-extracting archives created by other tools (like winrar), so I can check if that works for those SFX as well? --Zhuyifei1999 (talk) 05:11, 25 June 2018 (UTC)
Is File:Peazip YG.xcf one? --Zhuyifei1999 (talk) 05:21, 25 June 2018 (UTC)
No, this file contains a 7-Zip SFX archive with a Rar archive inside. The Rar archive contains peazip.exe – instalation package of PeaZip. I think that the bot finds the EXE and the Rar archive because in such cases 7-Zip is smart enough and doesn't compress the archive but just stores it. BTW. Hmmm, for some reason I was convinced that the bot detects SFX archives created by WinRAR… Now I'm not sure, but I'm too lazy to go through automatically deleted files in order to find an example. --jdx Re: 06:34, 25 June 2018 (UTC)
Hmm, the reason might be because of the Magic detector working on Rars. --Zhuyifei1999 (talk) 08:09, 25 June 2018 (UTC)
The bot also doesn't delete gziped files: Special:Undelete/File:سليبل.gif. BTW. Do you remember JPEG files with attached text which looks like base64 encoded data, just like this one: Special:Undelete/File:FFC_5441.jpg? They are prepared by a tool called Free File Camouflage. It seems that the tool adds yet another EOI marker (FFD9) and then VVGEC8…, so you might try to use FF D9 56 56 47 45 43 38 as the signature for this type of files. --jdx Re: 13:33, 25 June 2018 (UTC)
Ok, will look into it in the next days. --Zhuyifei1999 (talk) 15:31, 25 June 2018 (UTC)
@Jdx: By the base64 one, so you mean User_talk:Zhuyifei1999/Archive_36#Special:Undelete/File:صورة_فيس_بوك.jpg? That one starts with 'HXc+UA' instead of 'VVGEC8' after two 0xFFD9. --Zhuyifei1999 (talk) 15:41, 25 June 2018 (UTC)
Most likely this is encrypted data produced by the tool. I have "camouflaged" the same data using the same "cover image" and when the data is not encrypted, it starts with VVGEC8 (actually this constant string seems to be a bit longer); encrypted data starts with a random string, i.e. different strings for different passwords. --jdx Re: 16:05, 25 June 2018 (UTC)
Hmm. The base64 decoding of this 'unencrypted version' is still random trash 00000000: 5551 840b cb86 0d0b 8d86 193a 14a8 3420 UQ.........:..4 . It's probably related to the use of AES in that program. I don't think it's a good idea to only blacklist the non-encrypted (those with VVGEC8). However, I can't think of a way to 'blacklist' base64s that make no sense. --Zhuyifei1999 (talk) 16:38, 25 June 2018 (UTC)
It just looks like base64 but it isn't. If you open such data in a text editor you will see two long lines, the first seems to be a base64-like encoded header and the second seems to be base64-like encoded actual data. Anyway, you can easily extract data from such JPEGs using Free File Camouflage. --jdx Re: 17:06, 25 June 2018 (UTC)
I don't use Windows regularly so will do that next time I setup a Windows VM, hopefully within this week; my last one broke. The problem with using their program to extract such data is that I don't know how it works, no specs or anything I could implement into the bot. This is why I tried to disassemble it. --Zhuyifei1999 (talk) 17:13, 25 June 2018 (UTC)
As for freefilecamouflage, the disassembly has a unusual amount of bad instructions... --Zhuyifei1999 (talk) 15:43, 25 June 2018 (UTC)
Probably because it is a .NET application, i.e. it runs on a virtual machine instead of "bare metal". --jdx Re: 16:17, 25 June 2018 (UTC)
Argh. VMs making everything more complicated. If it were written in C it would be so good. --Zhuyifei1999 (talk) 16:38, 25 June 2018 (UTC)

@Jdx: What is going on with File:PIC FIL.png? A new program? --Zhuyifei1999 (talk) 06:21, 30 June 2018 (UTC)

I have no idea. I will look closer at it in the evening. --jdx Re: 07:45, 30 June 2018 (UTC)
According to TrID File Identifier it is a Windows Journal's file. Indeed, Windows Journal Viewer opens it. This particular file contains only a picture – our abuser probably looks for tests new possibilities. --jdx Re: 11:40, 30 June 2018 (UTC)
Oh great... a proprietary format. --Zhuyifei1999 (talk) 13:25, 30 June 2018 (UTC)

Bot question

Hi, your Flickr review bot (which seems to be doing a great job, btw!) har tagged the image File:Wilhem Rediess at 8. Parteitag – Gefallenenehrung 27 September (cropped).png, which is just a cropped version of File:8. Parteitag – Gefallenenehrung (8616649598).jpg. The latter image is marked as Flickr Commons. The permission link is dead, though.

As I assume that the wartime photos from the Norwegian national archives are properly licensed, may I remove the tag placed by your bot? Asav | Talk 14:02, 25 June 2018 (UTC)

If the uploaded image is a DW of a reviewed Flickr image, yes I think you can remove the flickreview tag. --Zhuyifei1999 (talk) 15:29, 25 June 2018 (UTC)

Structured Data on Commons IRC Office Hour, Tuesday 26 June

Greetings,

There will be an IRC office hour for Structured Data on Tuesday, 26 June from 18:00-19:00 UTC in #wikimedia-office. You can find more details, as well as date and time conversion, at the IRC Office Hours page on Meta.

Thanks, I look forward to seeing you there if you can make it. -- Keegan (talk) 20:54, 25 June 2018 (UTC)

23:10, 25 June 2018 (UTC)

What properties does Commons need?

Greetings,

Structured Commons will need properties to make statements about files. The development team is working on making the software ready to support properties; the question is, what properties does Commons need?

You can find more information and examples to help find properties in a workshop on Commons. Please participate and help fill in the list, and let me know if you have any questions. Thanks! -- Keegan (WMF) (talk) 18:53, 28 June 2018 (UTC)

SignBot

https://commons.wikimedia.org/w/index.php?title=User_talk:Alexis_Jazz&diff=308507250&oldid=307528060

Was not signed, user has less than 800 edits, did not opt out, no {{speedy on my user page. - Alexis Jazz ping plz 19:50, 28 June 2018 (UTC)

Looks like an outage; I don't see this edit in the logs. On 19:32 the bot fataled and restarted twice. Then on 19:46 to 19:48 it timed out waiting for RC and restarted itself 4 times. That edit was probably caught in the restarts or somehow missed, --Zhuyifei1999 (talk) 20:12, 28 June 2018 (UTC)

HEY

Yo puedo hacer eso de poner la barra de los idiomas encima de lo que te escribo ???

Digo, para trasladar de uno a otro idioma --Cristian cda (talk) 21:08, 28 June 2018 (UTC)

Translated: "I can do that by putting the Language bar on top of what I write to you??? I mean, to move from one to another language"
@Cristian cda: no, you have to copy the text into https://www.bing.com/translator/ or https://translate.google.com/. (if I understand your question correctly) Also, SignBot and Zhuyifei1999 can't help you with this. - Alexis Jazz ping plz 21:24, 28 June 2018 (UTC)

Hi, this user is blocked by you but can still edit and vandalise his talk page. Can you take a look at that? Thanks! - Richardkiwi (talk) (talk) 20:44, 29 June 2018 (UTC)

Hi again, I think you left the talkpage open, but he/she is abusing it with hundreds of edits. The talkpage open during a block is only to talk about the block (as you know). Please block the talkpage asswell. Thnx. - Richardkiwi (talk) (talk) 20:50, 29 June 2018 (UTC)
ok, blocked and deleted. Totally wasn't expecting them to vandalize their own talk page. --Zhuyifei1999 (talk) 06:10, 30 June 2018 (UTC)

Crystal icons

Hi,

Because a certain IP broke the {{Crystal icons}} template, all files using this template (and that's a lot) where assigned to the Media missing infobox template category by YiFeiBot. Is there a way to fix this? Thomas Linard (talk) 20:38, 30 June 2018 (UTC)

The bot should undo itself after some time (when the database 'links' are updated) --Zhuyifei1999 (talk) 00:44, 1 July 2018 (UTC)
Perfect, thanks! Thomas Linard (talk) 07:31, 1 July 2018 (UTC)

FlickreviewR 2 seems to be down

Hi Zhuyifei1999. It seems that your bot has stopped working. As i write, there are 858 files in the queue, and counting. Thank you! --Edelseider (talk) 16:53, 1 July 2018 (UTC)

It's okay now. --Edelseider (talk) 19:22, 1 July 2018 (UTC)
Well, I didn't touch anything ;) --Zhuyifei1999 (talk) 00:03, 2 July 2018 (UTC)
It took a break for an hour or two. Artificial intelligence knows when to rest. :) --Edelseider (talk) 06:42, 2 July 2018 (UTC)

00:46, 3 July 2018 (UTC)

Structured Data on Commons Newsletter - Summer 2018

Welcome to the newsletter for Structured Data on Wikimedia Commons! You can update your subscription to the newsletter and contribute to the next issue. Do inform others who you think will want to be involved in the project!

Community updates
  • Our dedicated IRC channel: wikimedia-commons-sd webchat
  • Since our last newsletter, the Structured Data team has moved into designing and building prototypes for various features. The use of multilingual captions in the UploadWizard and on the file page has been researched, designed, discussed, and built out for use. Behind the scenes, back-end work on search is taking place and designs are being drawn up for the front-end. There will soon be specifications published for the use of the first Wikidata property on Commons, "Depicts," and a prototype is to be released to go along with that.
Things to do / input and feedback requests
Discussions held
Wikimania 2018
Partners and allies
Research

Two research projects about Wikimedia Commons are currently ongoing, or in the process of being finished:

  1. Research:Curation workflows on Wikimedia Commons—a project that seeks to understand the current workflows of Commons contributors who curate media (categorize it, delete it, link to it from other projects, etc.).
  2. Research:Technical needs of external re-users of Commons media—soliciting feedback from individuals and organizations that re-use Commons content outside of Wikimedia projects, in order to understand their current painpoints and unmet needs.
Development
  • Prototypes will be available for Depicts soon.
Stay up to date!

-- Keegan (WMF) (talk)

Message sent by MediaWiki message delivery - 21:07, 6 July 2018 (UTC)

Structured Data Newsletter - Research link fix

Greetings,

The newsletter omitted two interwiki prefixes, breaking the links on non-meta wikis as you might see above. Here are the correct links:

  1. m:Research:Curation workflows on Wikimedia Commons—a project that seeks to understand the current workflows of Commons contributors who curate media (categorize it, delete it, link to it from other projects, etc.).
  2. m:Research:Technical needs of external re-users of Commons media—soliciting feedback from individuals and organizations that re-use Commons content outside of Wikimedia projects, in order to understand their current painpoints and unmet needs.

My apologies, I hope you find the corrected links helpful.

- Keegan (WMF) (talk) 21:21, 6 July 2018 (UTC)