Open main menu

Wikimedia Commons β

User talk:Rudolphous

For archive User talk:Rudolphous/Archive

Mappings and a questionEdit

{moved 3 revision by Rudolphousbot/Rudolphous as of 22:07, 2 January 2018 from my talk, as started here, quotation start:}

Hello Roland zh, to improve the gps location mapping of RudolphousBot I check the reverts. Most of the reverts are of your hand and are correct. I will fix those errors. Of each error I make a test and those tests should run fine before the bot can run again. Hopefully you see improvement over time in the mappings. Mappings to places instead of regio's should not happen anymore. A difficulty is that not all commons categories are mapped to Wikidata yet and therefore I cannot always map to villages yet. To check if a mapping is correct I need the gps location of the category. If the gps location is not known it's to risky to map to it.

Question: The English Wikipedia had the following articles: [1] + [2]. On Wikipedia we have a redirect for this category, see [3]. Do you know if this is correct? Chirakkara has a seperate article on wikidata but is missing a commons category, maybe because of the redirect. Further, do you know if Chirakkarathazham is also a village? The English Wikipedia mentions this. Some photo's are taken in this place and officially not in Chirakkara. Rudolphous (talk) 22:05, 2 January 2018 (UTC)

{quotation end}
Hi Rudolphous, thx for notifying: on 1/2 January's night I recategorized about 1000 files related to South India, as above indirectly mentioned, mainly (panoramio) uploads on Andhra Pradesh, Karnataka (KN, and Goa, a handful which Rudolphousbot categorized within Uttara Kannada KN-district) and Kerala, latter on districts and towns (mainly district-related, btw, see also User talk:Rudolphous#Categorization by district). Including five additional 2 January Rudolphousbot-edits on Karnataka, about 25 files were undone, {categorise}-related as above mentioned. There's a more practical reason to do so, since these files are mostly part of serial-uploads (panoramio and flickr), and so it's much easier to categorise files (which otherwise remained uncatogorized) geographically in more adequate subcategories.
Related to your question: both villages are in Kerala's Kollam district; Category:Polachira bases on EN and ML Wikipedias, Chirakkara (just EN), one of assumably (ten)housands villages in Kerala, I redirected since latter also links to Polachira; from my side a more adequate/practicable procedure than to establish categories for every village, regardless if there are GPS coordinates.

As the last time I was very hurried, my best wishes for 2018 and happy editing, Roland zh (talk) 16:15, 5 January 2018 (UTC)

Hi Roland zh, best wishes for 2018! Oke, I now understand why Chirakkara is like it is now. To be honest, I'm not sure what if I get your Karnataka point fully. We want these images in deeper cats right? Maybe you can give an example link to illustrate your point. Rudolphous (talk) 16:12, 7 January 2018 (UTC)

High Street, BirminghamEdit

Please stop your bot from repeating this edit, which is incorrect. I have already reverted it twice. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:22, 23 December 2017 (UTC)

Oke, will have a look into this. RudolphousBot (talk) 19:48, 23 December 2017 (UTC)
The following data comes back from google maps:
{"types":["route"],"short_name":"High St","long_name":"High Street"}
{"types":["administrative_area_level_2","political"],"short_name":"West Midlands","long_name":"West Midlands"}
{"types":["country","political"],"short_name":"GB","long_name":"United Kingdom"}
{"types":["postal_code"],"short_name":"B6 4US","long_name":"B6 4US"}

Both Birmingham and Aston seems to have both a High Street. I did not see your reverts. However I fixed the problem. Let me know if you see more problems :-). RudolphousBot (talk) 19:48, 23 December 2017 (UTC)

Park AvenueEdit

Hi, like several users above, I undo the edits of your bot and would like not to do it anymore. From what I understand, because some uncategorized images are located on the "Park Avenue" of ramdom locations around the globe, your bot categorizes them on the category Park Avenue, New York. Worst, some images that were categorized in the right category (by actual human users) are decategorized and miscategorized by your bot (examples: [4] [5]). Not sure if you improved your bot since the comments above but hope not to do the work again. Best and happy holidays. --Deansfa (talk) 19:06, 25 December 2017 (UTC)

I shall have a look into this. Rudolphous (talk) 08:09, 29 December 2017 (UTC)
I could fix this by converting the commons category to the wikidata entry. Looking up the gps coordinate and calculate the distance. If the distance is too far the category is not used anymore. Rudolphous (talk) 12:05, 30 December 2017 (UTC)


Hi Rudolphous - your bot has also added several files to Category:Pinaceae, which have nothing to do with that! If you could check that too please. Thanks! - MPF (talk) 01:13, 30 December 2017 (UTC)

I shall have a look into this.
This is fixed now. Rudolphous (talk) 12:22, 30 December 2017 (UTC)


  • In this case, Category:Karlsruhe is completely wrong. The location is in Waldprechtsweier (a district of Malsch (Landkreis Karlsruhe)) within the county Landkreis Karlsruhe. Karlsruhe (a city) is not an umbrella term for Landkreis Karlsruhe. Karlsruhe isn't even part of the county of Karlsruhe. (And the location isn't really relevant in this picture). Also, if your bot adds the first cat to a formerly uncategorized file, please remove Category:Photos from Panoramio needing categories as of XXXX-XX-XX. Keep Category:Photos from Panoramio ID XXXXXXX needing categories though, as this one replaces a "Check category" tag. --Sitacuisses (talk) 04:57, 30 December 2017 (UTC)
Thanks for your report, will look into this. Rudolphous (talk) 07:25, 30 December 2017 (UTC)
So you just ignored my request to keep Category:Photos from Panoramio ID XXXXXXX needing categories. --Sitacuisses (talk) 10:07, 30 December 2017 (UTC)
Ah sorry, I misread this one. I will put that category back. Rudolphous (talk) 10:29, 30 December 2017 (UTC)
They are back now, right? Rudolphous (talk) 12:57, 30 December 2017 (UTC)
These [6] [7] are strange. What made the bot decide to add them to a random district of Karlsruhe (city), when the coordinates point to a different city (Baden-Baden)? Is it the name "Weststadt", that exists in Baden-Baden as well? The correct category can be easily guessed from the coordinates and the files' names: Category:Michaelsberg (Baden-Baden). --Sitacuisses (talk) 21:37, 1 January 2018 (UTC)
Here [8] [9] the reading of the coordinates lacks just a little precision. As you can see on the map, all were taken on the Baden-Baden side of the municipal border, while the bot added some files to Category:Baden-Baden and others to Category:Gernsbach. --Sitacuisses (talk) 22:12, 1 January 2018 (UTC)
For the mountain I get back the following result: {administrative_area_level_1=Baden-Württemberg, administrative_area_level_2=Karlsruhe, administrative_area_level_3=Rastatt, country=Germany, country_code=de, county=Baden-Baden, hamlet=Eckhöfe, locality=Baden-Baden, pitch=Gleitschirm-Startplatz West, postal_code=76530, postcode=76530, premise=Merkurturm, road=Merkurstraße, route=Merkuriusberg, state=Baden-Württemberg, state_district=Regierungsbezirk Karlsruhe, street_address=Merkuriusberg 2, 76530 Baden-Baden, Germany, street_number=2, suburb=Oberbeuern}. This is indeed a sligh error. To be honest: these two photo's give currently Category:Baden-Baden as result. I don't know why earlier today Category:Gernsbach was processed..... Maybe Open Street Map has been changed..... Rudolphous (talk) 22:34, 1 January 2018 (UTC)
"Administrative_area_level" seems to stand for a border which is part of two entities. Maybe this field contains one of these two entities alternatively, but not necessarily the one that corresponds to the actual location? This would be one more reason to use "county", "locality" and maybe even "suburb" instead. --Sitacuisses (talk) 23:38, 1 January 2018 (UTC)

Thanks for your message. I combine data from google and open street map. The overall results is:

  • administrative_area_level_1=Baden-Württemberg,
  • administrative_area_level_2=Karlsruhe,
  • administrative_area_level_3=Rastatt,
  • country=Germany,
  • country_code=de,
  • county=Baden-Baden,
  • locality=Baden-Baden,
  • postal_code=76530,
  • postcode=76530,
  • road=Solmsstraße,
  • route=Solmsstraße,
  • state=Baden-Württemberg,
  • state_district=Regierungsbezirk Karlsruhe,
  • suburb=Weststadt,
  • town=Baden-Baden

Normally when a suburb is given also a 'city' is also given. In this case there is no city en then I used administrative_area_level_2. For Baden-Baden this apparently doesn't work. Thanks a lot for your report! I fixed the problem and all other tests run oke. Rudolphous (talk) 22:07, 1 January 2018 (UTC)

Administrative_area_level_2=Karlsruhe could be en:Karlsruhe (region) then, not to be confused with en:Karlsruhe (district) or en:Karlsruhe. But what does "administrative_area_level_3=Rastatt" mean? Baden-Baden is not part of en:Landkreis Rastatt. Maybe this is just an error in OSM, or an incompatible method of classification? On the other hand, "county=Baden-Baden" and "locality=Baden-Baden" is correct and could be utilized.
When I categorize files from Panoramio, I sometimes look at the Commons on OSM layer to find similar files and the way they are categorized. Would it be possible to make the bot do the same, i.e. look for files, categories or wikipedia objects with coordinates nearby, then check the file's title and description for matches? --Sitacuisses (talk) 23:02, 1 January 2018 (UTC)
Rastatt is indeed an error. This information comes from Google Maps and not from Open Street Map by the way. Your OSM Layer link is really helpful. Your mapping approach is very interesting, but a little too complicated to automate at the moment I'm affraidRudolphous (talk) 22:11, 2 January 2018 (UTC)

categoriesEdit,_2016_07_03_(3).JPG_-_panoramio.jpg&diff=prev&oldid=274064423 this looks off... just thought you'd want to know. Jon Kolbert (talk) 11:13, 30 December 2017 (UTC)

Agree, this is already fixed. The bot now checks if categories already exist before adding them. I think this problem will be fixed soon by other bots that remove duplicate categories right? Rudolphous (talk) 11:17, 30 December 2017 (UTC)
Yep, that bot should be around in a bit. Jon Kolbert (talk) 18:01, 30 December 2017 (UTC)

Categorization by districtEdit

Hi Rudolphous. Thank you for your categorization work. When you localize images BY DISTRICT using coordinates (an example), please add the category of the district, not the category of the main town of the district. --ŠJů (talk) 00:29, 31 December 2017 (UTC) (The second problem is that coordinates of this Panoramio uploader are completely wrong.) Btw., to localize images directly by municipality can be a bit better. --ŠJů (talk) 00:35, 31 December 2017 (UTC)

Thanks for your clear message. The current version of the bot gives Category:Moravský Krumlov as result. Do you think this is correct? If the gps coordinates are wrong it will be difficult to handle..... Rudolphous (talk) 00:43, 31 December 2017 (UTC)
Came here to say the same thing, this is another example. --Novarupta (talk) 13:23, 2 January 2018 (UTC)
The bot returns Category:Zermatt now for this image. I rechecked all images in Category:Visp and a few other districts. Rudolphous (talk)
Even better, try to find out the exact location and to identify the subject of an image, because some district categories are already rather full, and the files would be better placed in the category appropriate to the town, village, or even the category for the subject itself if it exists. --Schlosser67 (talk) 13:33, 2 January 2018 (UTC)
This will be done from now on. Do you know some categories which are quiet full? Then I can try to recategorize them further. Rudolphous (talk) 18:43, 2 January 2018 (UTC)

Categorization of the locations in Varanasi by ghat nameEdit

I've noticed several cases of the photos categorised by the bot on basis geo-coordinates in Varanasi, India. It seems this is too precise. There are nearly 100 ghats along the river bank (over 5 km long) which are of the size middle-size railway platform whilst the pictures uploaded from websites as Flickr frequently have only approximate location (eg. File:Varanasi (8748090088).jpg - obviously taken not at the marked location of Digpatya Ghat, as it is indicated on the map). It would be better to categorise the pictures to the level of the city. There were several cases of inaccurate category or pictures of objects without any indication in file name or description that were photographed at the particular ghat (eg.  ). --Oo91 (talk) 23:31, 2 January 2018 (UTC)

Hello oo91, what picture do you mean? this picture has no RudolphousBot in de edit history. Rudolphous (talk) 23:42, 2 January 2018 (UTC)

Up till now I can see 2 cases when the bot went to the level of a ghat in categorisationː Ganges, Varanasi (8748085710).jpg (Category:Prayag_Ghat) and File:Ganges_flower_ceremony_(5105390746).jpg (Category:Sankatha_Ghat). Is is hard to say how many files have been categorised this way. Regarding this picture - it is an example of inaccurate localisation of the original file from Flickr by the author that can result with wrong category based on geocoridnates only. --Oo91 (talk) 00:00, 4 January 2018 (UTC)

Thanks for your message. I checked last 25.000 edits of RudophousBot and don't see any images going to ghats anymore. Let me know if you see more problems. Rudolphous (talk) 20:58, 4 January 2018 (UTC)
the checking algorithm was not well calibrated, apparently - I've noticed yet another example of ghat-level categorisationː File:Ganges, Varanasi (8748086774).jpg - RudolphousBot (replaced category on name from Category:Varanasi to Category:Bhonsale Ghat).
Hi Oo91, this one is from 1 januari and before the last 25k that I checked manually. OSM gives the ghat as attraction, see also [10]. Are you sure this is wrong? Then I will skip mappings to ghats from now on all together. Rudolphous (talk) 16:03, 7 January 2018 (UTC)

Inept Category ReplacementsEdit

Hello, it seems this bot recently recategorized a significant number of pictures by the place they were taken from, and at the same time took these pictures out of categories matching the location of the photographed object. Please revert these changes as soon as possible. --Abderitestatos (talk) 02:49, 3 January 2018 (UTC)

It were only a few pictures. Some of the changes were right and others (especially panarama photo's of mountains) I did revert. Let me know if you see something that is not right. Rudolphous (talk) 08:17, 3 January 2018 (UTC)

Adding inappropriate (but similarly spelled) Category tagsEdit

A possibly related problem. Spotting two inappropriate inclusions to Hue, I attributed them to some shortage in description of the category and did my best to indicate it is not about any kind of location. Apparently the bot doesn’t care. Please, double-check your category-related software. Incnis Mrsi (talk) 09:24, 5 January 2018 (UTC)

Will have a look into this. Rudolphous (talk) 19:05, 5 January 2018 (UTC)
I get the following keys for this gps locations: {administrative_area_level_1=Thua Thien Hue, administrative_area_level_2=Hương Thủy, administrative_area_level_3=Thủy Bằng, city=Hue, country=Vietnam, country_code=vn, county=Hue City, locality=Hương Thủy, political=Hương Thọ, postcode=54000, route=Minh Mạng, state=Thừa Thiên-Huế Province}. This image is mapped to Hue based on city, but yesterday I already changed that mapping to cities can only be done when the gps is known and within close region. Currently this images is mapped to Vietnam (because provincial categories are missing yet). Thanks for your report. Let me know if your see something else that is wrong. I will splitup Vietnam soon. Rudolphous (talk) 20:36, 5 January 2018 (UTC)
Category:Vietnam is now recategorized. Rudolphous (talk) 07:37, 6 January 2018 (UTC)
Do you understand what am I speaking about? To which category do these files pertain rightfully, can you write down himself? Ok, a hint: which location category does Perfume River belong to? Stop saying polite fluff (“will have a look”, “let me know”…) and fix your algorithms now. Yet one obviously wrong category added by your bot and I file a request for de-flagging. Incnis Mrsi (talk) 07:56, 6 January 2018 (UTC)
I did not understand your point. I see now we have [11] and [12]. Rudolphous (talk) 08:15, 6 January 2018 (UTC)
@Incnis Mrsi: please be civil, Rudolphous isn't writing "polite fluff", but is just being polite. Be clear in your reports, give clear examples where things went wrong and I'm sure Rudolphous will fix it. Multichill (talk) 11:13, 6 January 2018 (UTC)
  Thank you for cleanup. Have a good day. Incnis Mrsi (talk) 11:47, 6 January 2018 (UTC)
  Thank you for your patience. Rudolphous (talk) 13:25, 6 January 2018 (UTC)

One more one-month-old diff: Revision of File:A_band_playing_in_Eastwood_Square_-_panoramio.jpg. It’s silly not to built some sanity checks into the bot. Has the category {{disambig}}? Has it interwiki? If it’s Wikidata, then which properties has the element? I don’t expect this to be very complex. Incnis Mrsi (talk) 09:58, 9 January 2018 (UTC)

This is indeed already fixed a few weeks. The bot does not place images in {{disambig}}. Also it checks in wikidata for gps information. Rudolphous (talk) 17:35, 9 January 2018 (UTC)
Seriously does it check? Why does it choose Commons categories via text-matching and ignores Commons category then? We wouldn’t have troubles with the Vietnamese city. Incnis Mrsi (talk) 17:56, 9 January 2018 (UTC)
Yes seriously but differently then you propose :-) I map commons categories to gps information via P301 and P625. Then I calculate the distance from the picture location and the gps location in wikidata. The found gps location must be close, how close depends if it is a village, city, suburb, country, etc which is found in via google maps and open street map. The P373 is probably a good one as cross check (thanks for this) and will look tomorrow into it. Rudolphous (talk) 18:05, 9 January 2018 (UTC)

To come back on this one. The current process is:

  • 1 commons picture -> gps
  • 2 query gps on google maps and open street map to get category proposals
  • 3 generate variants of category proposals, for example beside Hue als Hue, Vietnam is checked.
  • 4 check category proposals on category existance from most specific (neighbourhood) to widest (country).
  • 5 if step 4 finds something, convert the commons category to gps via wikidata
  • 6 calculate the distance between gps from step 1 and step 5, if to far away reject this category and proceed with a wider category.

I don't see (yet) in which step I can use P373. Will think more about how to improve the bot and prevent (obvious) errors. Rudolphous (talk) 16:01, 10 January 2018 (UTC)

Republika Srpska vs Republic of SerbiaEdit

Can you please check this edit. Your bot categorize this image in Category:Republika Srpska, which is entity in Bosnia and Herzegovina, although this photo is georeferenced in Užice, Republic of Serbia, which is different country. Please see difference between Republika Srpska and Republic of Serbia --Smooth O (talk) 13:33, 10 January 2018 (UTC)

More examples: example and example. --Smooth O (talk) 13:35, 10 January 2018 (UTC)
Hi Smooth O, Thanks for your message for File:Zaovine, Serbia - panoramio (2).jpg (with geo location 43.836389, 19.382222) google maps returns. In shortadministrative_area_level_1=Republika Srpska, administrative_area_level_2=Zlatibor District, administrative_area_level_3=Municipality of Bajina Basta, country=Serbia, locality=Zaovine, route=Unnamed Road

That's reason why =Republika Srpsk used. The current version of the bot returns for the above examples currently "Užice", "Zlatibor District" and "Zlatibor District". I will recheck that category and make subcategories when needed. Rudolphous (talk) 15:09, 10 January 2018 (UTC)

@Smooth O: not an unexpected thing. Incnis Mrsi (talk) 14:10, 10 January 2018 (UTC)

@Smooth O: I see that category is already empty. Thanks for this! Rudolphous (talk) 16:11, 10 January 2018 (UTC)

Gazi Mustafa Kemal BulvarıEdit

Hey, bot, please don't add unrelated files to Category:Gazi Mustafa Kemal Bulvarı. :) --E4024 (talk) 12:12, 15 January 2018 (UTC)

Sure. 40.915685, 38.500793 is also Gazi Mustafa Kemal Bulvarı but is mapped to Keşap currently. Rudolphous (talk) 19:18, 15 January 2018 (UTC)

STOP adding inappropriate categories to Aircraft photos!Edit

During the recent weeks you have added geographical names as the only categories to hundreds, maybe thousands of aircraft files.

Instead of adding the proper type name, you have categorized them all in useless village names like "Le Grand-Saconnex" or "Meyrin" as their only cats.

Furthermore, you have removed their previous cat "Uncategorized", thus preventing the files being examined by experts and being put in their proper categories, namely aircraft type, operator, airport etc. - effectively hiding them away under tiny village names and preventing proper handling.

Please revert ALL those edits in order to make them available for expert handling.

You are causing a huge amount of totally unnecessary work to editors. Please stop this practice immediately !! --Uli Elch (talk) 13:47, 17 January 2018 (UTC)

You may have hidden many more aircraft files (hundreds, thousands?) somewhere else - nobody except you knows. I just stumbled about a mere 2 (two) cats by pure coincidence, to which you had removed those files from general access. Nobody searching for a particular Boeing 747 or Dassault Falcon would ever search for them in completely inappropriate categories like "Le Grand-Saconnex" or "Meyrin". I'm sorry, but as an admin you should have known better how to program and use a bot. Simply deleting this request within less than 15 minutes certainly was not a fair way to handle the problem. --Uli Elch (talk) 16:18, 17 January 2018 (UTC)
Don't worry I will recategorize them as you may have noticed in my contributions. More edits will follow soon. I will go through all the edits of RudolphousBot. Please ask a bit more polite next time because this annoys me. Rudolphous (talk) 16:31, 17 January 2018 (UTC)

File:Most probably Liris niger (31773344033).jpgEdit

  File:Most probably Liris niger (31773344033).jpg has been listed at Commons:Deletion requests so that the community can discuss whether it should be kept or not. We would appreciate it if you could go to voice your opinion about this at its entry.

If you created this file, please note that the fact that it has been proposed for deletion does not necessarily mean that we do not value your kind contribution. It simply means that one person believes that there is some specific problem with it, such as a copyright issue.
Please remember to respond to and – if appropriate – contradict the arguments supporting deletion. Arguments which focus on the nominator will not affect the result of the nomination. Thank you!

Afrikaans | العربية | Български | বাংলা | Беларуская (тарашкевіца)‎ | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | Zazaki | Ελληνικά | English | Esperanto | Español | Eesti | فارسی | Suomi | Français | Galego | עברית | Magyar | Bahasa Indonesia | Íslenska | Italiano | 日本語 | 한국어 | Македонски | മലയാളം | Plattdüütsch | Nederlands | Norsk nynorsk | Norsk bokmål | Occitan | Polski | Português | Português do Brasil | Română | Русский | Slovenčina | Slovenščina | Shqip | Српски / srpski | Svenska | Türkçe | українська | Tiếng Việt | 中文 | 中文(简体)‎ | 中文(繁體)‎ | +/−

Poecilotheria36 (talk) 13:13, 18 January 2018 (UTC)

Return to the user page of "Rudolphous".