User talk:Fæ

From Wikimedia Commons, the free media repository
(Redirected from User talk:Fae)
Jump to navigation Jump to search
Notice If you want to see Python source code that supports some of my projects, go to Github and help yourself. The code is not written with reuse in mind... -- (talk) 15:57, 15 May 2018 (UTC)
Notice

If you are concerned that a category gets flooded with automated uploads, check that a template like {{Disambig}}, {{Photographs}}, {{Categorise}}, {{CatDiffuse}} or {{CatCat}} has been applied before complaining. In the case of my batch upload projects, any category marked this way will not be added to new photographs. -- (talk) 16:32, 20 September 2018 (UTC)

Archives.png

2017
2018
2019

File:Noah Silliman 2016-11-02 (Unsplash).jpg[edit]

{{Please link images}}

File:Noah Silliman 2016-11-02 (Unsplash).jpg (edit|talk|history|links|watch|logs) Stombari7 (talk) 21:56, 25 May 2018 (UTC)Stombari7

Silverton[edit]

Recatagorized all images in the Category:Silverton to Category:Silverton, New South Wales.

Image Deletion[edit]

File:Internet Archive book plate project, blank example 1.jpg[edit]

Elisfkc (talk)

Thanks~Cessna_208_Caravan#/media/File:Cessna_208_Caravan_I,_Seawings_(Jet-Ops)_AN1347237.jpg[edit]

It is helping young students learn that the cockpit is not really that complicated ~ and all those fancy buttons and such do have a purpose and are very easy to learn ~ Mitchellhobbs (talk) 21:25, 13 April 2019 (UTC)

— Preceding unsigned comment added by Mitchellhobbs (talk • contribs) 21:30, 13 April 2019 (UTC)

Poor duplicates from trainpix.org[edit]

Hi Fæ!
You uploaded a lot of images from trainpix.org. Very good job overall. However, my trainpix images are weak duplicates and they hit Commons. I have a trainpix account and i uploaded a large photo on Commons and a small copy on trainpix. Trainpix has a limit to images no more than 1280 pixels wide only.
All my uploads on trainpix can be viewed here: [trainpix gallery]
For sample: original image and poor copy from trainpix.
Look at your uploads from trainpix and make deletion request for poor duplicates, please.
With best regards, --George Chernilevsky talk 07:55, 23 September 2019 (UTC)

Housekeeping of this type, where images are later uploaded which are duplicates, is not obviously an action for the uploader of the older image. As you know which files you have uploaded and were also released on trainpix by yourself, this is more an action to be considered by yourself.
You can mark lower resolution duplicates with {{duplicate}}, which is probably the simplest process to use.
Thanks -- (talk) 10:37, 23 September 2019 (UTC)

Seattle Sanborn maps[edit]

I see you uploaded a bunch of Sanborn maps of Seattle (thank you very much for that). However, it looks like a lot of them for 1893 were missed. If you look at https://www.loc.gov/resource/g4284sm.g4284sm_g09315189301/?st=gallery&c=160, we are missing a lot of these images (maybe half? I haven't entirely worked it out). I noticed this because I was looking in Commons for the images that are at https://www.loc.gov/resource/g4284sm.g4284sm_g09315189301/?sp=83 and https://www.loc.gov/resource/g4284sm.g4284sm_g09315189301/?sp=84.

Is there any chance of coming up with a way to upload the missing pages of this? - Jmabel ! talk 02:39, 22 July 2019 (UTC)

@Jmabel: this would just get re-archived by ArchiverBot on the next run. See [1], Fæ, I've increased the archive time from 12 to 30 days. You can reconfigure it however you like when you're back. - Alexis Jazz ping plz 02:52, 12 August 2019 (UTC)
@: I see you've been answering some questions lately, so I'm pinging you again. Is this something you can help with, or should I look elsewhere? - Jmabel ! talk 15:53, 30 August 2019 (UTC)
@Jmabel: I briefly looked into this when you raised it, but I cannot tell what the cause is. It could be that the LOC image API (IIIF) has some bugs we don't know about, such as not returning all the files in a scanned album. The 'multisheet' or locbooks uploading routine I use should be comprehensive, and has been rerun several times across the collection. It's a slow but intelligent routine, which notices where files have been uploaded under different filenames. It might be that some of the files you are noticing have actually been uploaded under other names previously, either by me or others, or it could be that the LOC have dummy images getting returned, or it could be that the images are failing within the IIIF image generator on the LOC server side because of something odd happening with the original image scan.
If you can further refine the pattern as to what is missing, then this would greatly help the chance of an easy diagnosis. For me to spend a lot of time digging in to case examples, I would need to know that we were talking about thousands of missing images, rather than a hundred which might be fixable via manual uploads.
My time on uploads is low right now. I should probably try an RFA again, which would warm up my interest as a volunteer with access to deletions and history manipulation, things that have held me back from improving what happens around Commons deletions for several years. It's the main reason that the Commons Fair Use Upload Bot was abandoned as a project, shockingly this was four years ago, itself a rather depressing thing to recognize. -- (talk) 16:09, 30 August 2019 (UTC)
  • It's definitely at most hundreds. And I suspect it was something along the lines of uploading the same number of images as there were plates, but each plate is actually two images. As I say above, for Volume 1 the pattern of what is missing is very straightforward. For both TIF & JPEG, we cut off at "Sanborn Fire Insurance Map from Seattle, King County, Washington. LOC sanborn09315 003-40"; I'm pretty sure that the remaining plates can be had by continuing exactly the same pattern, incrementing that '40', to get images 41-84. Would that much be easy to do? - Jmabel ! talk 16:37, 30 August 2019 (UTC)
If it's detectable, then it should be possible to add a trigger in the upload script to compensate for it. Not going to investigate today, I'm adding a DNAU for a few months to ensure we don't lose it as a wanted fix. -- (talk) 16:40, 30 August 2019 (UTC)

Category:Internet Archive (uncrop needed)[edit]

Hi Fæ. It looks like Category:Internet Archive (uncrop needed) hasn't been tended to for at least a couple of weeks. All the best, --Animalparty (talk) 20:55, 27 September 2019 (UTC)

I'm in the process of swapping my 11 year old HDD to a larger SSD, albeit a cheap refurbished one. It may take a while for me to revisit how this works, but it's caused by "retiring" my old upgradable Macmini which was running these sorts of tasks as a slightly inadequate headless server, so they should be migrated to my X220 Thinkpad. If you notice Faebot activities, you may want to endorse my request at the WMF hardware donation programme, presuming that it's still active. I've cut down a lot on my volunteer time this year due to travelling and a lack of suitable kit. -- (talk) 11:03, 28 September 2019 (UTC)
SSD working, doing a one off catch up run for IA backlog. Due to some odd error with login in to Faebot, which I cannot be bothered to debug right now, having to do these from my main account. -- (talk) 17:40, 28 September 2019 (UTC)
Now running as a Faebot task, every 6 hours. -- (talk) 13:18, 2 October 2019 (UTC)

@Animalparty, Ruff tuff cream puff: If you have some time to investigate, could you examine why these 404 errors might be happening? It may be the cause of a 'residual' number of unprocessed images in the uncrop backlog.

File:Image from page 1120 of "The Saturday evening post" (1839) (14784846062).jpg

IA link: saturdayeveningp1933unse/saturdayeveningp1933unse#page/133/mode/1up
IA page: 133
IA name: saturdayeveningp1933unse 
Download: https://archive.org/download/saturdayeveningp1933unse/page/133.jpg 
Sleeping for 38.2 seconds, 2019-09-29 12:58:04
WARNING: API error http-bad-status: There was a problem during the HTTP request: 404 Not Found
 APIError("http-bad-status", "There was a problem during the HTTP request: 404 Not Found", {u'help': u'See https://commons.wikimedia.org/w/api.php for API usage. Subscribe to the mediawiki-api-announce mailing list at <https://lists.wikimedia.org/mailman/listinfo/mediawiki-api-announce> for notice of API deprecations and breaking changes.'}) 
 Exception of this type not handled! 

-- (talk) 12:05, 29 September 2019 (UTC)

These could be a glitch in the archive. The uncropped image is stored at jp2, but I thought that all formats should be rendered as part of the archive. Based on this one(?) case, I'm not planning to try to rewrite code to fix it or work around it. For the time being I have manually emptied the category without uncropping the files, not a solution but we are so far unclear what the problem is, or if there is a fixable problem. -- (talk) 10:35, 30 September 2019 (UTC)

Rollbacks in watchlists[edit]

@Jeff G.: With regard to User_talk:Fæ/2019#Rollback, I went with this for all wikis:

.mw-changeslist-line .mw-rollback-link { visibility:hidden; }

.mw-watchlist seems to not exist as a class on my version, this might be down to other javascript mods in my layout. Previously my javascript swapped rollback link text with "-", but that has not worked for a couple of years, presumably due to restricting how javascript is allowed to behave. I could not get :after in css to work within the anchor itself which might have effectively done the same thing. -- (talk) 10:40, 14 October 2019 (UTC)

Thanks.   — Jeff G. please ping or talk to me 13:49, 14 October 2019 (UTC)

File:Mt. Rainier from Roof Garden, Hotel Lincoln, Seattle, Wash (NYPL b12647398-68897).tiff[edit]

What (if anything) is the basis for 1898 as a date for File:Mt. Rainier from Roof Garden, Hotel Lincoln, Seattle, Wash (NYPL b12647398-68897).tiff? Looks based on a photograph, not just an architect's visualisation drawing, but the building had not been built at that date (completed 1900, burned in 1920). - Jmabel ! talk 06:49, 27 November 2019 (UTC)

The date is simply pulled from the NYPL catalog. In this case the date appears to be for the collection not the item. Hence "1898: Issued (Approximate)" is actually the earliest date for objects in the collection, so if we can deduce more based on the building itself, that can at least be added. The dates for the postcards should probably all be adjusted to a date range or at least qualified to say this is the earliest date.
As the upload was 2016, I have no precise memory of the mappings used, and the current NYPL catalog page looks very different to what it was back then and may be displaying different data. -- (talk) 07:04, 27 November 2019 (UTC)
About what I thought. Sounds like if we can still identify what was in that batch, anything that says 1898 should instead say {{other date|after|1898}}. I'll correct the date for the one I was looking at. - Jmabel ! talk 07:13, 27 November 2019 (UTC)
Probably 5,775 matches search. -- (talk) 07:19, 27 November 2019 (UTC)
✓ Done dates replaced. -- (talk) 09:35, 30 November 2019 (UTC)
Hmm, I notice, in looking at half a dozen of these Detroit Publishing Co. postcards, that they have serial numbers. Maybe those numbers are in strict chronological order, in which case a category listing them in that order would allow interpolating from the dated ones to date all of them.
— Preceding unsigned comment added by Jim.henderson (talk • contribs)
The categories already have keys by image ID. I hestitate to replace those keys. However I am running a routine to add postcard numbers from the NYPL catalog to the description, ref search. This makes it easier to do some smart report later, without having to read from external sites.
As a test, it may be that DPC>5000 ⇒ year>=1901 and DPC>6000 ⇒ year>=1902 -- (talk) 17:04, 1 December 2019 (UTC)
Ah, and I see you have started adding "Postcard series number" from the publisher's printed serial number. Very good. Clearly the files already have many numbers, but what I failed to notice is that they are already sorted chronologically, more or less, because that's how the "NYPLb" number was assigned. This suggest that all year I have been going about this the wrong way. Instead of continuing what someone else started, by making and filling geographic subcats, I should have started with annual subcats, partly to break down the large category that hinders my operation with slow DSL and other impediments, and partly as inboxes for geographic and other category work. It might help if the sources and uses of the various numbers were described on a page, perhaps the Detroit Publishing category talk page or some other that the ignorant are likely to stumble across. Jim.henderson (talk) 19:44, 1 December 2019 (UTC)

Detroit postcard numbers[edit]

Elle of Ganado, DPC 9999, published "prior to 1906"

Looking at the first 2000 DPC numbers, filtering for dates (and end of range dates where they appear), gives:

Year Min Max
1901 132 497
1902 5011 7636
1903 6002 6999
1904 60523 8704
1906 9011 9999
1908 11000 12157
1909 12026 12999
1915 60007 60538
1918 71002 71981

Empirically that's probably very good evidence to use the postcard number to deduce a latest date of publication where this is missing, noting that some have ranges like 1905-1915, so choosing "1915" is definitely a latest date that might be refined later.

A place to keep specific dating tips for the NYPL collection is at User:Fæ/Project list/NYPL, it may even be worth a shortcut link. -- (talk) 11:50, 2 December 2019 (UTC)

„File:Kazuenokami Kato Kiyomasa Observing a Monkey with a Writing Brush LACMA M.84.31.456 (2 of 2).jpg“[edit]

Hello, Fæ! Searching for the prints of Utagawa Toyonobu’s series „Shinsen Taikōki“ I came accross this file. Years ago you uploaded the file „File:Kazuenokami Kato Kiyomasa Observing a Monkey with a Writing Brush LACMA M.84.31.456 (2 of 2).jpg“ but it makes no sense; parallel you upoaded „File:Kazuenokami Kato Kiyomasa Observing a Monkey with a Writing Brush LACMA M.84.31.456 (1 of 2).jpg“ which is the correct content (1 of 2) of Yoshitoshis print. The uploaded print (2 of 2) is the left sheet of a Toyonobu diptych you can find at Commons under „File:End of Baba Minonokami.jpg“. The file (2 of 2) should better be deleted. Best. --Horst Gräbner (talk) 16:07, 28 November 2019 (UTC)

I don't see a need for deletion. If the work can be identified correctly, renaming of the file and adjusting the description is what should be done here. De728631 (talk) 21:09, 28 November 2019 (UTC)
It’s an unusable file. Which sense makes it to have the half of a diptych when only the two prints depict the composition? But on the other side here are tens of thousands unusable files. So please do what you mean. Regards. --Horst Gräbner (talk) 12:34, 29 November 2019 (UTC)
Just to reiterate, it could be that LACMA only have this one page of the original diptych. As such it's a reasonable photograph of the specific print. If you can find evidence that both parts are held by LACMA, that would be useful to add to the description, even if we cannot find the equivalent extra photograph. -- (talk) 12:38, 29 November 2019 (UTC)
Six years is a very long time ago for upload projects. Vaguely I recall some issue with wrong images appearing in the catalogue at LACMA, and if you check the source, the error is still there. However the "new" website has since dropped these images from public view, ref https://collections.lacma.org/node/1349418, consequently I doubt the curators have any interest in correcting the old format catalog. If we can find the other half of the diptych that matches, that would be ideal. However it may never have been made public at lacma.org, or perhaps LACMA only ever had this single page of the diptych in their collections, and a conventional Google site search shows nothing. Certainly the textual description on Commons should be corrected, do be bold and have a go.
-- (talk) 21:34, 28 November 2019 (UTC)
File moved to 長篠合戰 - Nagashino kassen - the battle of Nagashino.jpg and description amended based on the new identification. -- (talk) 09:53, 30 November 2019 (UTC)

Used laptop[edit]

Thinkpad X220 with last of the old-style IBM keyboards

In which country (or at least, what continent) do you live? I'm not sure what I can do, but I tinker with hardware sometimes. If you live (or can receive packages) in the EU, I could think about this. Also, what are you looking for in a used laptop? It's not clear to me from m:Hardware donation program/Fæ if it should be chunkier, if IPS is required, processing power, battery life, keyboard layout.. - Alexis Jazz ping plz 14:34, 30 November 2019 (UTC)

UK based. However, I do tinker with hardware as well. The current X220 was a slightly damaged one from eBay for £100 and has been upgraded with a refurbished SSD and a small format m2 sata card, both from eBay. It's a great laptop, especially the keyboard, but the processor is soldered on so not replaceable, and I have had overheating problems. I like the old ThinkPads as they are so upgradable and repairable but I have had other machines. My request for the WMF scheme was a bit of a test as whether it really works, but based on the history of Commons upload projects I could bypass the Wikimedia organizations and try opening a crowdfunding project online to buy a beefier fixed machine to act as a home server and a more mobile easy to carry a laptop to fiddle with programming while in different places or travelling. If that worked, I would probably buy new kit rather than second hand or refurbished.
Vaguely thinking of specs:
Desktop or long term server
  1. Low-ish power consumption to be a bit greener
  2. Low noise as it'll have to be used as a desk machine
  3. Supporting 2 or 3 monitors, though this might be tacked on in various ways
  4. OS & processor has a good reputation for image processing and power management using standard Ubuntu install
  5. As it'll be used for experiments or maintenance on graphics, complex image processing (like mass image hashing) and video processing, something like 32GB or more of RAM
  6. Storage less of an issue, as 500GB disk space would be plenty and archives can be put out to connected drives
Laptop something nice and light to carry around in hand luggage and user maintainable enough to last at least 5 years (Thinkpad, Dell?)
  1. Probably no more than 14-inch screen with high resolution, 13-inch is ideal
  2. Core i7 within the last few years (my X220 has i7 but it's a rather old one) or better or user upgradable
  3. Good reputation for running with Linux (I have not used Windows for a decade)
  4. 16GB RAM and preferably upgradable to more
  5. Good reputation for efficient cooling, rather than fans maxing out when video processing is going on
I have not thought about UPS, it's probably not essential as power outages are rare. Sorting out an automated backup is probably more critical. There is also the caution that whatever is always-on has credible green credentials and would probably still need to be powered off if unattended for weeks for safety. -- (talk) 15:11, 30 November 2019 (UTC)
Not UPS, IPS. Better colors, but more expensive. Core i7 laptops aren't really something I see much. And unless you're talking about quad cores (not every i7 is a quad core), I'm not sure there is much of a benefit over an i5. 16GB of DDR3L laptop memory can be found on AliExpress for 40 euros or so, I didn't have to pay import tariffs but YMMV, but this is probably useless advice as any laptop from the last few years uses DDR4. In general, used laptops are best value when they are at least 4 years old in my experience. That's roughly the time after which companies write them off. If you want something newer, buying new is probably a better idea. When a new 14" Ryzen 5 quad core laptop can be found for 500 quid or so, a relatively new used machine isn't great value.
The cheapest desktop with 32GB is probably a fake X79 and a Xeon like ali(remove this)express.com/item/33031262183.html?storeId=604710 (Kllisre X79 motherboard set with Xeon E5 2650 V2 4x8GB=32GB 1600MHz DDR3 ECC REG memory ATX USB3.0 SATA3 PCI-E NVME M.2 SSD) for 140 quid. Don't know what the import tariffs for that would be. (in NL it would cost about 45 euros to import, assuming they declare the actual value.. but they probably declare less) - Alexis Jazz ping plz 17:04, 30 November 2019 (UTC)
Some other notes..
  • Video processing on a laptop is not the most amazing idea (obviously), but if you do it some external cooling solution would be recommendable. You could put something together yourself or get something off the shelf like File:Laptop cooler.JPG.
  • If your laptop heavily depends on its underside for cooling (your X220 doesn't seem to, but my Dell gets airflow from the bottom), lift the laptop an inch or so from the surface it's sitting on. Not always practical, but for my Dell it makes a huge difference.
  • In a 14" or smaller laptop, you will generally not find a user upgradable CPU. Upgrading a laptop CPU is usually not worth it, though there are a few exceptions.
  • 16GB of RAM can be had in a DDR3 laptop (even your X220 supports that), more means DDR4 and drives the price up.
  • A third, fourth or fifth gen i5 laptop (X220 is second gen) can be found under 150 quid, sometimes as low as 50. Newer generations cost more and usually have DDR4. (sixth gen can be either DDR3 or DDR4) Seventh gen is barely worth the additional cost second hand IMHO, compared to new.
Most of the stuff I have laying around atm is mostly as old or older than your X220. As long as your X220 still works probably not worth the trouble. - Alexis Jazz ping plz 18:17, 30 November 2019 (UTC)
And I forgot: your cooling problems. Probably caused by either dust or a bad thermal connection between your CPU and heatsink. In case of dust (assuming you don't want to take the entire machine apart), you can try blocking the fan with a thin screwdriver or something and blow into it with compressed air. But be careful with the pressure, even when the fan is blocked, extremely strong airflow can bend stuff. If the problem is the connection, you would have to take it apart, remove the cooling paste or pad and apply new cooling paste. If you're out of luck, the fan is running on its last legs. Finding a replacement may or may not be easy. - Alexis Jazz ping plz 19:46, 30 November 2019 (UTC)
Thanks for the suggestions. I have looked at how to get at the fan, but it's a bit of a nightmare on the X220. The problem was mainly (probably) caused by doing FFmpeg recompression for hours at a time, sadly something that I have failed to find a way to "officially" run on labs both because of complaints about the storage needed and finding the IP address blocked by YouTube. Without that heavy loading, the core temperature hovers around 64-74C with medium fan noise.
If the WMF get back to me in the next couple of months, I'll be genuinely grateful for whatever they can let me have from their old development quality laptops and adapt my plans to whatever that make is. If they don't, I may start looking at cheapest second-hand 440p prices with the option of eventually upgrading the processor and maybe think again about the effort involved in trying my hand at crowdfunding to raise enough to buy something more modern. The T440p is an old machine, but has good reviews for this sort of video processing and can probably easily handle real-time Zoom video conferencing, which the X220 definitely is *useless* at, though I would probably keep the X220 handy as a devoted basic Ubuntu word processor and occasional code editor for its lovely old-style keyboard. -- (talk) 22:38, 30 November 2019 (UTC)
A quick look at the hardware maintenance manual for the X220 makes it appear as if the fan should be blockable from the outside, but I don't have an actual X220 to check. Note that for video processing with ffmpeg only one thing counts: the CPU. Nothing else. So you'd probably want to avoid T440p models with a GeForce (Intel HD Graphics 4600 is not bad anyway) and try to find a model with a quad core. (so not the i7-4600M) A quad core running at a lower frequency is more power efficient than a dual core at a higher frequency if the task is multithreaded. Though a second hand quad core machine doesn't seem to be very attractive unless you can find a really good deal. A Lenovo V330-14ARR 81B1000FMH (different class, I know) with AMD Ryzen 5 2500U is 500 quid. AMD really shook things up. - Alexis Jazz ping plz 23:50, 30 November 2019 (UTC)

File:Russia Fortepan 6412.jpg[edit]

File:Russia Fortepan 6412.jpg (edit|talk|history|links|watch|logs)
Commons:Deletion requests/File:Russia Fortepan 6412.jpg Akela (talk) 15:46, 30 November 2019 (UTC)

File:A Busy July -) (18650072673).jpg[edit]

File:A Busy July -) (18650072673).jpg (edit|talk|history|links|watch|logs)
Commons:Deletion requests/File:A Busy July -) (18650072673).jpg A1Cafel (talk) 07:37, 1 December 2019 (UTC)

File:Too many "unresolved" Itnl Transgender Day of Remembrance -tdor Washington DC USA (15834983091).jpg[edit]

File:Too many "unresolved" Itnl Transgender Day of Remembrance -tdor Washington DC USA (15834983091).jpg (edit|talk|history|links|watch|logs)
Commons:Deletion requests/File:Too many "unresolved" Itnl Transgender Day of Remembrance -tdor Washington DC USA (15834983091).jpg A1Cafel (talk) 03:27, 2 December 2019 (UTC)

laurentthiry[edit]

  • Hi. I would like to overwrite File:KierchSchweebech.JPG with a better quality picture that I took myself. However, this is not possible. It would be awesome if you could help me with this, or tell me why I can't overwrite the file. Thank you in advance. Laurentthiry (talk) 08:16, 2 December 2019 (UTC)
Hi Laurentthiry, please do not overwrite other photos with your own just because you think yours is better. Please have a look at COM:OVERWRITE to learn when overwriting is ok. Instead, you should upload your photo under a new name. Regards, AFBorchert (talk) 09:03, 2 December 2019 (UTC) (in talk page stalking mode)
Continued at User talk:AFBorchert#laurentthiry. --AFBorchert (talk) 09:46, 2 December 2019 (UTC)

File:Rothenberg, Hessen - Stadtansicht und Landschaftsansicht (Zeno Ansichtskarten).jpg[edit]

Hallo Fæ, Diese Karte zeigt nicht Rothenberg in Hessen, sondern Rothenberg bei Stuttgart (jetzt Rotenberg, Stadtteil von Stuttgart, Rotenberg (Stuttgart)) mit dem Berg Württemberg. Auf dem Berg kann man die Grabkapelle erkennen. Dafür spricht auch die Aufschrift am Rande: „Stuttgarter Ausflugskartenserie I“ und der Sitz des Verlags „Kunstanstalt H. Haufler & Co., Stuttgart“. Willst du die Beschreibung und Kategorisierung der Karte ändern? Gruß --Mewa767 (talk) 09:19, 3 December 2019 (UTC)

How to change 6000 handles, and monitor changes in the Naturalis shell collection?[edit]

Dear Fae,

Perhaps for you the following is just routine:

1. Naturalis shell curators would want to monitor metadata changes by shell enthusiasts in the metadata about 110 k files in https://commons.wikimedia.org/wiki/Category:Mollusca_specimens_at_Naturalis_Biodiversity_Center

2. Leiden University Library wants to put permanent handles as backlinks to their database (I think) in https://commons.wikimedia.org/wiki/Category:Media_contributed_by_KITLV in the metadata field | source =

  • Could you give me some hints, are there tools for this, or is a Python script needed (i didn't use a little Python since ... 2000?)?

Thanks a lot, Hansmuller (talk) 11:18, 3 December 2019 (UTC)

Depending on what sort of things they want to monitor, there are many ways of monitoring changes to the collection like recent changes or using Petscan. If I get the GLAM dashboard running again, that generates modest reports.
Mass adding/changing links to more than a few hundred files is something pywikibot is well suited to, though if you prefer you can raise a request at COM:BR (@Hansmuller: recommend this way forward, the folks that hang out on that noticeboard can knock out scripts like this in a matter of minutes, so long as you can specifify exactly how the changes are to be made). If the backlinks are easily generated based on current identity information then it's fairly easy to write a script for. If there are just a few hundred changes at a time, then COM:VFC is worth trying. -- (talk) 19:20, 4 December 2019 (UTC)

AHKP1[edit]

Hi Fæ! Is your script still identifying the photographer? Has he been active since the script was run? (I hope he's not scared away.)

Do you keep the list of his accounts somewhere? It might be interesting for the community to understand his activity over the years, when he stops contributing someday.

I think Hoia Saerda Mea 28 (talk · contribs) is him too, but your script did not match it.--Roy17 (talk) 12:21, 4 December 2019 (UTC)

These accounts are not him:
  1. Anderle (talk · contribs)
  2. Bambinok (talk · contribs)
  3. Enkol (talk · contribs)
  4. Hanankhaleeq (talk · contribs)
  5. Hovorka (talk · contribs)
  6. Riosolar (talk · contribs)
  7. UncleRemus (talk · contribs)
Gentech.spv (talk · contribs) is probably not him either.--Roy17 (talk) 12:26, 4 December 2019 (UTC)
Not planning on doing another run at the moment. I would argue against doing it often or automatically, maybe we should try again in 3 or 4 months time and think about whether other types of contribution pattern could usefully apply. The historical tracking of accounts by behaviour or other data is fine, but as this is not vandalism or in other ways damaging, there is potentially more damage from being over-aggressive in tracking this uploader, plus we are probably always going to some misidentifications which could worry other users too. Note that I can not stop someone going through the category and reverse engineering a list of uploader accounts, but I do think it would be a mistake to publish it.
Accepted your list at face value, related uploads will be gradually removed... could massively speed it up but probably not worth the extra programming thinking time. If you spot others, save up a list and I'll de-cat them as a block. -- (talk) 12:55, 4 December 2019 (UTC)
I agree that he should not be constantly tracked. I was just wondering whether he has contributed since this cat was created and was worried he might be scared off. I was actually thinking of more conservative tagging—his accounts should be matched only after one or two years. For example, a run in 2025 should only match his photos before 2023/4 maybe, so that he's not being followed real-time.--Roy17 (talk) 13:05, 4 December 2019 (UTC)
This might be a more friendly approach. Just re-running one of the reports, it looks like they have created 9 different accounts in the last 2 days, possibly this slight change in behaviour is to put off tracking. -- (talk) 13:21, 4 December 2019 (UTC)

File:Japan 2015 (22936656609).jpg[edit]

File:Japan 2015 (22936656609).jpg (edit|talk|history|links|watch|logs) Batholith (talk) 15:00, 5 December 2019 (UTC)

Sumner Heights and Hazelwood Valley Railroad[edit]

Takes the narrow gauge (NYPL b11707500-G90F222 014B)

Thank you very much for uploading two excellent photos of the en:Sumner Heights and Hazelwood Valley Railroad, which are shown in Category:Sumner Heights and Hazelwood Valley Railroad. After looking at the description shown on the back, I searched for the 11 other photos of this series, but couldn't find them. Are you more experienced (or lucky) in searching for them? --NearEMPTiness (talk) 08:30, 6 December 2019 (UTC)

The best check is at NYPL looking at the Norfolk County sub collection here. There are 52 stereogrammes, but after a brief look, I do not see any in this particular series and so were unlikely to have been preserved. -- (talk) 16:51, 7 December 2019 (UTC)
Thank you for the link and for confirming, what I had expected. --NearEMPTiness (talk) 13:47, 8 December 2019 (UTC)

Adding talk page to the talk page trimmer[edit]

Hi! I agree to add my talk page to the talk page trimmer, but I don’t see an easy way. Could you help me please? -- Ввласенко (talk) 14:55, 7 December 2019 (UTC)

@Ввласенко: Thanks for writing. Your page is on the list, pages are added manually. Your talk page exceeds the maximum transclusion count because of the large number of QIC templates. If you want, I can convert all of those to comments, which should fix the problem, but those particular templates are not part of the talk page trimmer. -- (talk) 16:43, 7 December 2019 (UTC)
@:Thank you! If you can, convert all of those to comments, please. -- Ввласенко (talk) 07:31, 8 December 2019 (UTC)
✓ Done Most but not all QICpromoted templates are being trimmed. There is some variation in use of the parameters, so a few are left intact for the moment. -- (talk) 11:27, 8 December 2019 (UTC)

@: Thank you! -- Ввласенко (talk) 07:22, 11 December 2019 (UTC)

MOTD Notification[edit]

The file you uploaded, is on the main page!

The file File:Monkeys Spinning Monkeys (ISRC USUAN1400011).mp3, that you uploaded, is on the main page today. Thank you for your contributions to this project.

Motd-logo 02.svg

//EatchaBot (talk) 00:03, 9 December 2019 (UTC)

Notification about possible deletion[edit]

Bundle DR:
Commons:Deletion requests/Files found with Toy Ship and Sailors incategory:"All media needing categories as of 2016"

Affected:

And also:

Extended content

Yours sincerely, Gbawden (talk) 12:35, 9 December 2019 (UTC)

FV Promotion[edit]

This video has been promoted to Featured video!

The video File:Healthy Pregnancy Tips From the CDC.webm, that you uploaded is now assessed as one of the finest videos on Wikimedia Commons, the nomination is available at Commons:Featured video candidates/File:Healthy Pregnancy Tips From the CDC.webm. Thank you for your contribution. If you would like to nominate, please do so at this nomination page.

Featured videos logo square.svg

/FVCBot (talk) 14:24, 9 December 2019 (UTC)

Talk page trimmer optin[edit]

Hi, can I opt-in for User talk:Fæ/talk page trimmer? Emoji u1f603.svg — regards, Revi 03:37, 12 December 2019 (UTC)

I have created Category:Talk page trimmer for users to opt-in/out more easily, without needing any programming time too. The category is being processed, there's nothing on your talk page that will show that yet. It's not yet at toolforge, I want to soak test a little to make sure no last minute bugs were introduced. -- (talk) 08:08, 12 December 2019 (UTC)