User talk:Magnus Manske/Archive 2

url2commons
- what's wrong? Why I can't use it? ShinePhantom (talk) 16:50, 11 September 2013 (UTC)
 * I am trying to open that file in my browser, but it takes forever to load (over a minute and counting). My guess would be timeout error. If it's just that one image, try uploading manually. If it's many more, try copying them somewhere else (ftp, flickr). Ah, now it finished loading, but took way too long; probably worse from the US where the Labs servers are. --Magnus Manske (talk) 17:22, 11 September 2013 (UTC)

missing topics tool
Hi Magnus Manske,

would it be possible to enable the MissingTopics tool for wiktionary? Especially for the German one? Best regards --Yoursmile (talk) 07:38, 27 August 2013 (UTC)
 * It does not work for lv wikipedia anymore. It was used for several projects. We need it back, pretty please :) --Papuass (talk) 14:49, 19 September 2013 (UTC)
 * "Wontfix:upstream". Both projects are not available on Labs at the moment. I try to get this resolved. en.wiktionary works fine, for example. --Magnus Manske (talk) 15:35, 19 September 2013 (UTC)

Nomination of Astral spirit for deletion
A discussion is taking place as to whether the article Astral spirit is suitable for inclusion in Wikipedia according to Wikipedia's policies and guidelines or whether it should be deleted.

The article will be discussed at Articles for deletion/Astral spirit until a consensus is reached, and anyone is welcome to contribute to the discussion. The nomination will explain the policies and guidelines which are of concern. The discussion focuses on high-quality evidence and our policies and guidelines.

Users may edit the article during the discussion, including to improve the article to address concerns raised in the discussion. However, do not remove the article-for-deletion notice from the top of the article. Brainy J ~ ✿ ~ ( talk ) 19:39, 13 September 2013 (UTC)

Creation date in CatScan 2
It is possible to filter the articles on creation date, but is there a way to sort them on creation date? Apart from that, it would be useful to have creation date in the result table.

That, and the ability to find article/talk category intersections ("find articles in category X where their talk pages are in category Y"), are pretty much the only things that keep it from being perfect. :-) (IIRC, article/talk intersections tended to create a performance problem; not sure if that's still the case.) GregorB (talk) 10:15, 17 September 2013 (UTC)

BaGLAMa
Hi Magnus,

I was wondering if you could add commons:Category:Images_from_Livrustkammaren_och_Skoklosters_slott_med_Stiftelsen_Hallwylska_museet to the BaGLAMa tool? There are only roughly 200 images in the category at this point but a further ~13000 are expected within the month. Many thanks, Lokal_Profil 14:22, 17 September 2013 (UTC)
 * Done. Update soon. --Magnus Manske (talk) 11:50, 20 September 2013 (UTC)

Blank reports from GLAMorous
Hi Magnus- any idea why some Commons categories with large numbers of files show as zero files in GLAMorous, e.g. ? Thanks for this and all your brilliant tools, by the way. MartinPoulter (talk) 15:43, 20 September 2013 (UTC)

BaGLAMa (bis) - Rijksmuseum
Hi Magnus,

IMHO it could be very intresting and usefull to add commons:Category:Rijksmuseum_Amsterdam to the BaGLAMa tool. (I tried but didn't succeed to have the the pageviews with GLAMorous). It could be a very good example to show the success of open access for a museum. Thanks Shonagon WPfr 08:14, 20 September 2013 (UTC)
 * Done. Update soon. --Magnus Manske (talk) 11:51, 20 September 2013 (UTC)
 * Thanks! So there was 13M pageviews in July 2013 ad 9M in August 2013 for images in commons:Category:Rijksmuseum_Amsterdam. Shonagon WPfr 01:26, 22 September 2013 (UTC)
 * Yup. Much of the difference seems to be due to singular entities, e.g. for Rembrandt in July. The raw data seems to be missing a few data points, but hints at a massive "view event" on the 15th, probably a Main Page mention. --Magnus Manske (talk) 11:31, 22 September 2013 (UTC)

BaGLAMa request
Greetings, Magnus,

Could you add commons:Category: Images from the Chemical Heritage Foundation to the BaGLAMa tool? We are just starting, but hope to release more images on an ongoing basis. Thank you so much for creating tools which we can use to track impact over time. Mary Mark Ockerbloom (talk) 17:56, 23 September 2013 (UTC)
 * Done. --Magnus Manske (talk) 21:19, 23 September 2013 (UTC)

Yet another BaGLAMa request
Hi Msgnus, I've a similar request for the national (royal) library of the Netherlands. We are kicking off out Wikipedian-in-Residence project (together with the National Archives) in two week time, and would like to track the impact of the image donations we plan to do.

Hence my request: could you pls add commons:Category:Royal Library, The Hague to the baglama-tool? Many thanks --OlafJanssen (talk) 09:17, 26 September 2013 (UTC)
 * Done. --Magnus Manske (talk) 09:51, 26 September 2013 (UTC)
 * Many thanks! This will help convince our management of the potential of Wiki — Preceding unsigned comment added by OlafJanssen (talk • contribs) 12:36, 26 September 2013 (UTC)

And another - Tate
Hi Magnus, commons:Category:Tate; they are keen to see their figures. Thanks a bunch as always! Johnbod (talk) 15:00, 27 September 2013 (UTC)
 * Added, will update later. --Magnus Manske (talk) 17:47, 27 September 2013 (UTC)
 * Thanks again Johnbod (talk) 17:16, 28 September 2013 (UTC)

Catfood creates invalid feed
Hello. Sorry to not using JIRA, but I don't have account there. I tried to add these two feeds ( and ) to my InoReader, but it refuses to accept them with invalid feed error. Feed Validator also display errors. --Peter — Preceding unsigned comment added by 194.228.68.191 (talk) 20:26, 27 September 2013 (UTC)
 * Should work now. --Magnus Manske (talk) 11:47, 28 September 2013 (UTC)

Perfect, thank you very much. :) Peter --194.228.68.109 (talk) 21:03, 30 September 2013 (UTC)

Mongolism listed at Redirects for discussion
An editor has asked for a discussion to address the redirect Mongolism. Since you had some involvement with the Mongolism redirect, you might want to participate in the redirect discussion (if you have not already done so). Ego White Tray (talk) 03:49, 2 October 2013 (UTC)

BaGLAMa request - State Library of New South Wales
Hi Magnus,

I'm working at the State Library of New South Wales on a Wikimedia project. Would it be possible to add commons:Category:Images by the State Library of New South Wales and commons:Category:Images from the State Library of NSW please?

Thanks, Rubicon49bce (talk) 03:45, 4 October 2013 (UTC)
 * Done. --Magnus Manske (talk) 08:08, 4 October 2013 (UTC)
 * Thank you very much! Rubicon49bce (talk) 08:37, 4 October 2013 (UTC)

Terminator-Tool / Filter
Hi Magnus, regarding your Terminator-Tool it would be useful to have some kind of filter mechanism for it. E.g., the first 50 or more of the entries for "missing label" in German appear to be Chinese localities, with only one Wikipedia link to Chinese Wikipedia. Therefore, they are very difficult and for most German users impossible to label. Perhaps one could have a filter e.g. for "missing label in German", with the filter being e.g. "and at least 5 available wikipedia links", that means including probably one wikipedia link from a wikipedia with latin script. The same issue exists with regard to Russian localities. This is just a proposal, perhaps something like such a filter would be possible for you to implement. Best regards, Longbow4u (talk) 20:03, 7 October 2013 (UTC)
 * As an alternative, it would also work if the result table would show the language codes of the available wikipedia links of the item. Then it would be easy e.g. to skip the items with only "zh" and "ru" wikipedia links. Longbow4u (talk) 20:11, 7 October 2013 (UTC)
 * Thank you! Longbow4u (talk) 21:00, 11 October 2013 (UTC)

GLAMorous bug?
Hi Magnus

Was just using your GLAMorous tool and noticed that it wasn't possible to toggle Commons on/off. This is probably linked to Commons showing up as commons.wikipedia rather than commons.wikimedia (it also responds to the wikipedia toggle rather than the wikimedia/commons one).

Cheers /André Costa (WMSE) (talk) 13:04, 11 October 2013 (UTC)

Cécile Chaminade quote
Hallo Magnus! I was startled tonight by a program note identifying Ambroise Thomas as a writer. Do you remember if the Etude article you used for the Cecile Chaminade article elaborated on this? Did he go further than occasional journalism? Sparafucil (talk) 09:01, 12 October 2013 (UTC)
 * Hi Sparafucil, I have no information about this other than the original text I copied, which I presume is all they had on the topic. --Magnus Manske (talk) 16:31, 13 October 2013 (UTC)
 * Update: Found the source. --Magnus Manske (talk) 16:36, 13 October 2013 (UTC)
 * Never guessed I would interrupt work on Lasso to try to become an expert on Thomas! I've failed to come up with more than a couple of letters that he wrote, so the claim of famous writer seems exaggerated at the least. Thanks for the link! I'll stick it in the article. Sparafucil (talk) 02:47, 15 October 2013 (UTC)

Thrilled to see & possible bug?
Hi Magnus, I was thrilled to see the page counts for "Images from the Chemical Heritage Foundation" starting to appear. It is very exciting. Thank you so much! One question, though; 2013-09 did not seem to appear for this or for several other categories, although it is there for most categories. A possible bug? Thank you! Mary Mark Ockerbloom (talk) 18:13, 18 October 2013 (UTC)

BaGLAMa question
Hi Magnus, a question about your wonderful BaGLAMa tool.

Consider the image File:Jean-Baptiste_Isabey_003.jpg which is used in en:Napoleon and counted with Commons:Category:Cultural heritage monuments in Austria with known IDs in the BaGLAMa tool.

It took me some time to find out why it is counted. The image itself is not marked with Commons:Category:Cultural heritage monuments in Austria with known IDs, but it is in the category Commons:Category:Portraits of Marie-Louise of Austria, Duchess of Parma -> Commons:Category:Marie-Louise of Austria, Duchess of Parma -> Commons:Category:Burials at Imperial Crypt -> Commons:Category:Imperial Crypt -> Commons:Category:Capuchin Church, Vienna, which is a protected object in Austria being marked with Commons:Category:Cultural heritage monuments in Austria with known IDs. So the image is counted as protected, because it is the portrait of a noble person being buried in a crypt underneath a protected church. The portrait of the lady is not protected at all (at least not in the scope considered).

It looks as if you are stepping down the category tree and count everything underneath the root category. The same situation arises also for images in a museum, when the museum is protected, or for events taking place in a protected building. While it is natural to add works of art or events to the location / building, it is not always clear that the property of cultural protection is inherited down the category tree.

My question: Is there a way to exclude certain subcats from being considered by BaGLAMa? Is it possible to restrict BaGLAMa matches to only files and not inherit them down the category tree at all? Or do we have to remove the category considered by BaGLAMa from object categories like Commons:Category:Capuchin Church, Vienna or alternatively, restructure the corresponding category trees?

What is your recommendation? (As the category Commons:Category:Cultural heritage monuments in Austria with known IDs is generated via a template, it would be (technically) easy for me to not generate the marker category on object categories at all.)

This is a question in the tension between truth and marketing, it could well be possible that huge figures are intended. I'm not the initiator for counting that category and I'm not the primary contact to the external organization. While a change in counting should not be a problem when truth is the point, it might be a problem in the marketing case. regards --Herzi Pinki (talk) 15:27, 21 October 2013 (UTC)
 * Hi Herzi Pinki, the problem here is clearly the wrong use of the category system. What is true of a "high-level" category should also be true of all items within the sub-tree. Rather than adding some crude hack to the Baglama data generator to circumvent the broken category tree, I suggest that the category tree is fixed. If all your images are using a template, you could also add a new category that is actually correct, if you don't want to meddle with the existing tree. --Magnus Manske (talk) 15:49, 21 October 2013 (UTC)

Du erhältst einen Orden!

 * Danke! --Magnus Manske (talk) 16:14, 27 October 2013 (UTC)

Joanjoc tools' "Articles without interwikis" problem
Hi. Joanjoc tools' [http://tools.wmflabs.org/joanjoc/artniw.php? "Articles without interwikis" tool's] "Namespace" parameter doesn't seem to work (i.e., works only for articles). Is there anything you can do about it? 85.194.253.105 (talk) 16:34, 27 October 2013 (UTC)

Cross-site scripting
Hi, some of your tools are vulnerable to cross-site scripting. Here is an example :

http://tools.wmflabs.org/isin/?language=de&isin=%3Cscript%20type=%22text/javascript%22%3Ealert%28%27hello%27%29%3C/script%3E

194.254.53.103 (talk) 18:03, 28 October 2013 (UTC)

Category filler
Hello Magnus! I am a French admin that salutes you.

I have just seen your numerous tools on wmflabs and I now wonder if you could help me by building the one I need. I have asked people on the French Wikipedia but they whether don't know how to do it or maybe don't care. I think you may be able to bring a different answer.

This would be a category filler. As a French Wikipedia user, I would go to your external server and fill the name of a French category. The tool would then go to all the equivalent categories that exist in other Wikipedias (using the links provided by Wikidata) and it would look there for pages with a French equivalent (again through Wikidata). The tool would then select only those that aren't yet in the French category selected by the user at the beginning. And then it would eventually list all those French pages in one column and their equivalents in another (under their language code, maybe).

This would be very useful, especially if the URL is modified by the request. We could then put a link in any French category through our template fr:Modèle:Boîte des catégories that would basically, once clicked, provide a list of suggestions of articles that we already have and that could be in the category. Wonderful!

For the moment, I am using a less sophisticated version, which is already great and useful, but not exactly what I imagine. It is available here. It deals with one foreign language at a time and it provides a full list of French articles, not one shortened to the pages that are absent from the French category the user chose. Plus the user actually choses a category in a foreign language, not his or her own.

And eventually there is no possibility to have depth selection. I would like the user to be able to tell the tool what he or she wants regarding this. Of course the depth would be reached in the foreign versions, not the French one (for the French category could actually be created and linked to others but still empty).

Can you help me? Thierry Caro (talk) 07:18, 20 October 2013 (UTC)
 * OK, first attempt here, prefilled example here. Like that? --Magnus Manske (talk) 20:42, 20 October 2013 (UTC)
 * Note: The depth is used for the source (e.g. "fr") category tree. Using depth for the foreign categories is possible, but might lead to significantly longer runtime. Not implemented. --Magnus Manske (talk) 21:05, 20 October 2013 (UTC)
 * Update: Added "external depth" so you can set a tree depth for all external ("non-fr") categories found. External depth 1 takes about 5x as long as ED0, and doesn't really get better results AFAICT. Your choice ;-) --Magnus Manske (talk) 08:59, 21 October 2013 (UTC)
 * Thank you very much! You were fast and the tool seems very powerful. So much in fact that I am still a bit troubled, ah ah! I think it's great and will be very useful, at least to me, for I like to create new categories! This will really help to fill them! Thierry Caro (talk) 14:34, 23 October 2013 (UTC)
 * Mmmh. In fact, it seems that there is a misunderstanding. I believe that your tool first goes into the original language category tree and then moves to the foreign languages. While it would be much better for it to move straight to the foreign languages from the original category and then into their category tree. For example, I have just created fr:Catégorie:Entreprise ayant son siège à Toronto and linked it to its equivalents such as Category:Companies based in Toronto. I would like the tool to go to the latter first and then look there for articles with French links that are not in the new French category. I believe your tool first look to the new French category content and then to equivalent. But this category being new is void and thus there are no results. While we know that there should be some. Can you have a look? Plus there seems to be a problem with categories that have a ' in ther name. Thierry Caro (talk) 14:59, 23 October 2013 (UTC)
 * The was a small bug, unrelated to the "order" issue you described (which is not an issue IMO). Try it now. --Magnus Manske (talk) 10:55, 24 October 2013 (UTC)
 * Thank you! Thierry Caro (talk) 00:51, 31 October 2013 (UTC)

BaGLAMa issue
Hi Magnus! Thanks as always for your GLAM tools. I noticed something odd, a huge spike, in recent NARA data, so looked closer. In August, one of the NARA pictures was the featured image on enwp's main page, and that resulted in the main page getting counted (with 300 million views), when it was only there for a day. It probably makes sense to exclude any project main pages when counting the data, since content is only featured for a short time and they will greatly skew the data. It would be nice if we could count the view for the day it was featured (for enwp, that could be 10 million views!), but I'm sure that's pretty complicated. Dominic·t 18:50, 4 November 2013 (UTC)

Herding Sheep source
Hey Magnus, i really like the Herding Sheep tool for tracking what my students are doing after i gave a Wikipedia workshop. However, i'd really like to add an option for giving a list of individual users instead of a category. Unfortunately i can't find the source on Bitbucket, so i wonder if it might be in another place? I could do a pull request after i'm ready so you could incorporate it into the tool. Thanks. Husky (talk page) 11:11, 5 November 2013 (UTC)
 * There you go! --Magnus Manske (talk) 12:58, 5 November 2013 (UTC)
 * Wow! Thank you so much! Still wondering about the source though, next time i'll give you a pull request :) [[File:Huskyoog.jpg]] Husky (talk page) 11:04, 7 November 2013 (UTC)
 * Now here. --Magnus Manske (talk) 11:37, 7 November 2013 (UTC)

BaGLAMa problem
Hi Magnus! There seems to be a serious problem with these stats recently. Look at "Tate" - this is one of a number of institutions which has shown massive decline in recent months, with October total figures well under half July, losing over 4.5 m views. But if you look at the en:wiki figures, just comparing the top 10 images, and then look at the articles and their views, it is clear that most of the July top 10 have just disappeared for no reason I can work out. Look at the Hogarth 006 in Pug and Aesthetics for example, or the Blake in William Shakespeare. And 6/9 of the top language figures declined by over 50% between July and October - link. Several unchanged images seem no longer to be being picked up. Might it be due to changes in the categories? Eek! Johnbod (talk) 17:38, 6 November 2013 (UTC)
 * I am doing the best I can with the data I can get. I suspect the problem is the latter, and that will not change until the WMF gets their act together on view stats. --Magnus Manske (talk) 11:39, 7 November 2013 (UTC)

url2commons
I am trying to upload some books with https://tools.wmflabs.org/url2commons/, got error Proxy Error

The proxy server received an invalid response from an upstream server. The proxy server could not handle the request POST /url2commons/.

Reason: Error reading from remote server

- Jayanta Nath (Talk|Contrb) 08:24, 12 November 2013 (UTC)
 * Example URL? --Magnus Manske (talk) 14:06, 12 November 2013 (UTC)


 * Here it is https://archive.org/details/holybibleinbenga00brit .- Jayanta Nath (Talk|Contrb) 13:00, 13 November 2013 (UTC)

GeoHack language links
Hi! Please see Module talk:Coordinates. --81.182.250.189 (talk) 15:40, 14 November 2013 (UTC)

Oh category guru
Got a tricky one for you. I'd like to determine the following:
 * Recursively find all articles that are under but which aren't in any categories under  - excepting any sub-categories that are under ! The tree is nested together, so I'm not sure if there's an easy way besides enumerating all of the categories I'm interested in to calculate this intersection.--Obi-Wan Kenobi (talk) 00:56, 16 November 2013 (UTC)

Geograph2commons
Not working again - error pops up "Could not get photo data - check photo ID" :-(   Cnbrb (talk) 18:21, 29 November 2013 (UTC)
 * If it's "again", you should have learned to give me an actual example to work with by now. --Magnus Manske (talk) 23:49, 29 November 2013 (UTC)

Catscan2 finds irrelevant articles.
Hello, it looks like catscan2 works with mistakes or may be this is some api bug or may be something else, I can't get what's wrong but catscan2 returns articles which have no relation to the category I'm searching in - searching in the category ru:Категория:Право (law) returns articles: It works so as if it finds some random articles. So, some part of articles on result page are relevant, some are not. Some of them are from one parent category (for example several articles from ru:Категория:Полки армии Конфедерации. I have no idea what's wrong and what to do. Sorry, I don't have an example of this in English or German. This is the only one category where I observe this problem, it works fine with others. ~ Nirvanchik ~ ⊤άλҟ 09:48, 25 November 2013 (UTC)
 * ru:Лебедев, Никифор Дмитриевич - this article shouldn't be searchable, it's not categorized well
 * ru:Любимов, Семён Иванович - a teacher in university
 * ru:Лебединский, Валерий Ефимович - writer, poet
 * ru:49-й вирджинский пехотный полк - infantry regiment
 * ru:Рейнхарт, Иоганн Кристиан - painter
 * OOps sorry I missed the link "please use the JIRA system" on top of talk page. It would be nice to have it at the tool html page. ~ Nirvanchik ~ ⊤άλҟ 09:52, 25 November 2013 (UTC)
 * hi, it looks like you're using a depth of 6- when you go that deep you will often find lots of articles that don't seem related to the topic. If you trace back all of the categories you can usually figure out why - I've done detective work on catscan results before and there's always a reason - for example, I was looking at Jews and a bunch of African American artists showed up - it turns out under jew is Jewish families, under Jewish families is Guggenheim, under guggenheim are Guggenheim fellowship award winners - thus, lots of non-Jews. The English category tree is full of this kind of thing.--Obi-Wan Kenobi (talk) 13:20, 25 November 2013 (UTC)
 * Hi, you are absolutely right. All of these articles are "correctly" included through category network. I just couldn't find some chains for long time, so I started to suspect the tool. Another reason is that, from my experience, irrelevant articles start to appear in large scale on depth 8 (russian category Biology), sometimes on depth 7 (russian category Computer networking). I didn't believe that could be on depth 6, but for this particular category "Право" this is true. ~ Nirvanchik ~ ⊤άλҟ 13:04, 30 November 2013 (UTC)

Adding Hebrew wikipedia to a tool
I am from the Hebrew wikipedia and we need you to help us adding hebrew to the options in the tool that shows "Top 1000 wikipedia-linked items without an article in he.wikipedia". Thanks a lot.Uziel302 (talk) 23:30, 29 November 2013 (UTC)
 * Done, should show up within 24h. --Magnus Manske (talk) 23:58, 29 November 2013 (UTC)
 * Still don't see a change. Maybe you added the wrong language? Uziel302 (talk) 09:17, 1 December 2013 (UTC)
 * I see it. --Magnus Manske (talk) 20:32, 1 December 2013 (UTC)
 * I see that too, thank you, I actually meant we want to add he.wikipedia here, that's why I thought there was a mistake. Sorry for the misunderstanding. And thanks in advance for adding us to the tool.Uziel302 (talk) 21:18, 1 December 2013 (UTC)
 * Added there too now. Link already shows up, but it may be up to 24h before there is data. --Magnus Manske (talk) 22:50, 1 December 2013 (UTC)
 * Thank you very much. Uziel302 (talk) 15:36, 3 December 2013 (UTC)
 * Maybe there is some problem now, it says:"There was an error running the query [You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near ')' at line 1]". Uziel302 (talk) 14:30, 5 December 2013 (UTC)

Yeah, Labs had a bad hair day. On it. --Magnus Manske (talk) 16:49, 5 December 2013 (UTC)
 * Now it works, partly, but not the part that I need...Uziel302 (talk) 21:23, 6 December 2013 (UTC)
 * I thought you were looking for this? --Magnus Manske (talk) 23:29, 6 December 2013 (UTC)
 * Yes, now it works. Thanks. Uziel302 (talk) 07:53, 7 December 2013 (UTC)

Trying to reconcile reports from different tools
Hullo, I'm just trying to wrap my head around the results from a couple of different tools.

http://tools.wmflabs.org/glamtools/glamorous.php reports that  "Images from the Chemical Heritage Foundation" has 172 images on 16 sites, 129 images on en.wikipedia, with 107 distinct images used. A previous glamorous report, 2013-10-29, showed 12 sites, with 145 images of which 90 distinct images were were used, so the number of images released and used increased during the last month.

http://tools.wmflabs.org/glamtools/baglama.php?group=Images+from+the+Chemical+Heritage+Foundation&date=201310 reports Page views of pages containing files from "Images from the Chemical Heritage Foundation" for 2013-10, and shows details for 39 pages on enwiki, 4 pages on dewiki and 1 on cswiki (no other sites are mentioned).

Doing a comparison of 2013-10 to 2013-09 reports a drop in image use: itwiki, for example, is not counted in 2013-10 even though the glamorous reports indicates that an image is used there. http://tools.wmflabs.org/glamtools/baglama.php?group=Images+from+the+Chemical+Heritage+Foundation&date=201310&diff=201309

So I'm not sure what's happening. If I assume that the tools are looking at similar data, shouldn't more sites and pages be included in the Baglama report? Thanks, Mary Mark Ockerbloom (talk) 18:11, 9 December 2013 (UTC)
 * Hi Mary, the view data I have access to is slow to access, and sometimes has "hickups". It may well be that the same data, queried at different times, produces different results. There is not much I can do about it, except re-run something and hope for the best. Still waiting for WMF to make view stats available to us. --Magnus Manske (talk) 23:15, 9 December 2013 (UTC)

About FIST
Good day! Where to write about error in FIST?--GlaDooo (talk) 17:26, 10 December 2013 (UTC)

CommonsHelper and NoCommons
I think CommonsHelper mark incorrectly NoCommons template German wiki as no verificable good license. It's only reminder to do not transfer file without an individual review. In result, CommonsHelper marks invalid and example de:Datei:Microsoft Word-Logo.png OK file as no verificable good license. --Rezonansowy (talk &bull; contribs) 12:21, 10 December 2013 (UTC)
 * Now works OK, I've added param to it. --Rezonansowy (talk &bull; contribs) 13:02, 12 December 2013 (UTC)

CatScan2 tool doesn't work for Russian wikipedia
Hi. It stopped working some 2 days ago. When you try to scan a category (in Russian wiki) it says "Could not connect to ruwiki.labsdb : Access denied for user 'p50380g50536'@'10.4.1.120' (using password: YES)". It seems this is not a bug in the tool so I'm writing here. Can you handle this? ~ Nirvanchik ~ ⊤άλҟ 17:13, 14 December 2013 (UTC)
 * I don't know how, but it started to work again. ~ Nirvanchik ~ ⊤άλҟ 19:11, 15 December 2013 (UTC)

Kleinigkeit, aber irgendwie symptomatisch
Wenn man bei "Glamorous" auf den WM Labs den Link unter "Copy the URL of this link to return to this view" kopiert, hat der ein Format, das in MediaWiki nicht als Weblink funktioniert… (wegen der eckigen Klammern). Ist das WM Labs-spezifisch? Ich kann mich nicht erinnern, bei Toolserver-Tools jemals solche Probleme gehabt zu haben. --AndreasPraefcke (talk) 19:20, 23 December 2013 (UTC)

Catscan suggestion: talk page categories
Hi, sorry I don't have a JIRA account but I still wanted to leave a quick suggestion for catscan: could you add searching for talk page categories? (similar to how it is implemented for templates) Main use case I would want to use that for is searching for stub-class articles (talk page category) sorted by article size, but I would imagine it could be used for many other purposes as well. --WS (talk) 09:56, 24 December 2013 (UTC)

Wikidata Query: link/nolink seems broken
Hello Magnus, I am one of those who learned from you about Wikidata Query during your presentation at the recent CEE meeting in Slovakia. Now when I wanted to make a use of it, I realized that searching for data items with/without a link to a specific wiki (i.e. an article in that language) does probably not work. Your own example http://208.80.153.172/api?q=link%5Bdewiki%5D%20and%20nolink%5Bfrwiki%5D delivers only an empty set. Will this please be fixed so that I can use the tool to find articles on a certain topic worth translating from one language to another? (it's linked to my Czech-Slovak Wikipedia project, also presented at the CEE meeting). Thank you. Blahma (talk) 18:16, 20 December 2013 (UTC)
 * It's working at this moment, but now the labels do not show, and I also have problems with the new props argument in the API. Anyway, what would now help me the most is adding support for "link" and "nolink" to the graphical "wdq" interface, so that I could easily point editors to a list of articles that are worth translating. Do you think you could have a look at this, please? And is this a good place for bug reports for Wikidata Query or is there a better one? Blahma (talk) 16:10, 25 December 2013 (UTC)

Flickr2Commons
Merry Christmas Magnus

I've been using Flickr2Commons quite a lot recently and plan to use it more in the next week or two, I have two questions:

1. Is there a way of seeing what files I have uploaded using the bot (they don't show up in my uploads). I guess I've done about 3-4000ish and it would nice to see them to make sure I've got all the categories correct etc.

2. Is there a way of retrying failed uploads from the transfer form? I've had a couple of situations with many tens of files and I get "Transfer failed : 500 Internal Server Error" and when I click to the transfer files button again it seems to not try the failed uploads again.

Thank you for making your tools, I'd be stumped without them

--Mrjohncummings (talk) 23:36, 25 December 2013 (UTC)

WiDaR sorting different from AutoList
Hi Magnus, I tried out the [http://tools.wmflabs.org/widar/index.php? WiDaR] in connection with [http://tools.wmflabs.org/wikidata-todo/autolist.html? AutoList]. If you check a specific entry in AutoList and add a claim, then it might add the claim for another entry. I guess that the sorting of these two tools are differently. My example was to add P17 = Q143 in the lists http://tools.wmflabs.org/wikidata-todo/autolist.html?props=P132,P131&q=claim[132%3A184188 %20AND%20noclaim[17 ] ]. --Zuphilip (talk) 08:26, 8 January 2014 (UTC)
 * That's odd. WiDaR only takes a specific list of items from AutoLists, there is no sorting involved. This is the first I hear of adding claims to the wrong entries. Can you point to an example edit that went wrong? And, why are you adding "Country:Esperanto" (P17:Q143) to items? --Magnus Manske (talk) 13:37, 8 January 2014 (UTC)
 * Sorry for the typo, I meant Q142 = France. Example from above (url) and check "All" entries. This will add the statement on 50 entries but not on the ones shown on the first page. I guess these are the first 50 entries sorted by Q-number. Is this a bug or feature? --Zuphilip (talk) 15:51, 8 January 2014 (UTC)
 * Ah, I think I know what's going on! When you start WiDaR, the checked elements are removed from the AutoList, so you can process the next batch easily. When you go back from the WiDaR screen, the Qs are different. To try, check just one, remember its Q before you add the claim. --Magnus Manske (talk) 17:04, 8 January 2014 (UTC)
 * Hm.. maybe that was the reason. Thank you for the explanation. If I see another example, I will let you know. --Zuphilip (talk) 18:15, 8 January 2014 (UTC)

Can I haz no depth?
Hello, my name is rp279cat. I haz fright of deep trees and long sheets... When people use Property tree tool to upper classes, me frighted to death. Me want go away from this pot, but me can't now because me fright it may happen again. =~_~V Me understand you need show upper classes too... I seem can make happy if you just not haz it so deeply. Can I haz some help? =^_^V

4th-otaku (talk) 17:36, 22 December 2013 (UTC) 
 * U can haz this --Magnus Manske (talk) 20:28, 8 January 2014 (UTC)

FIST
The fist tool on WMFLabs is giving out an error message.

I know that you have a link on the top of this page redirecting to jira.toolserver.org but since this is not on the toolserver, I didn't know where to post it. (t) Josve05a  (c)  17:38, 12 January 2014 (UTC)

Adding author to Wikidata Stat
I wanted to help by adding the author to some books' Statement where there are none. But it seems that you are allowed to add only those names it exists a corresponding wiki page for. Is that so? Carlotm (talk) 03:07, 17 January 2014 (UTC)
 * More precisely, a Wikidata item for the author must exist, so you can link to it. Most items do correspond to Wikipedia pages, but you are allowed to create new items (see sidebar on Wikidata) for "notable" or "structurally needed" entities. Check out d:Wikidata:Notability for details. --Magnus Manske (talk) 12:59, 17 January 2014 (UTC)

Nested Queries in WDQ
Hi Magnus, I am too fascinated by your neat work on the Wikidataquery and I want to build something on top of it. Problem is: I am as dumb as a rock and I cannot understand how to make nested queries using your API. So, your example of German bridges is easy: first you select everyhing that looks sort of a bridge and then you ask only for those in Germany or one of its parts. What if I want to obtain the list of all mayors of a capital? Here the result is a list of people, but the constraint is on the city they are mayor of. Do I need to get first the list of capitals and then get their mayors? If so, how? If not, is there a better strategy or is it not currently possible with the APIs? Thanks!

Michele.coscia (talk) 21:03, 15 January 2014 (UTC)
 * Wikidata is not yet set up well to do this on a large scale, but in principle, you can do:


 * 1) Pick a city (or use a subquery to get the cities you like), e.g., Toronto (Q172)
 * 2) Build a query to get the item for the position "Mayor of Toronto", like this
 * 3) Now use that query as a subquery to find items that have it as "position held", like this
 * If you use more than one city (via a subquery instead of a specific item), you will lose the person-to-role-to-city association; the current query language only goes so far. --Magnus Manske (talk) 13:15, 17 January 2014 (UTC)

Reset TUSC pwd?
Hi Magnus, is the procedure to reset my TUSC password still to contact you? I forgot my pwd and can't login :-(

Best, Ter-burg (talk) 10:00, 19 January 2014 (UTC)

Terminator languages
Hi. Was wondering if it would be possible to add swedish to the top 1000 lists available through your terminator tool. Would be a good way of getting the translator community hooked on wikidata :) /Lokal_Profil (logged out) 11:22, 22 December 2013 (UTC) — Preceding unsigned comment added by 94.234.170.44 (talk)
 * Done, should show up within 24h. --Magnus Manske (talk) 07:51, 8 January 2014 (UTC)
 * Many thanks. Will tie this into our translation project. /Lokal_Profil 08:04, 20 January 2014 (UTC)

More content, please
When you created Alexander Thomas Glenny you didn't give us a clue as to why the man was notable, beyond being an FRS. Please remember to give a bit more information even in a short stub. I've expanded this one a bit now. Thanks. Pam D  20:33, 20 January 2014 (UTC)

Happy Magnus Manske day!
How does one wish that to you? Happy "you/r" day? Anyway, thanks :) --Elitre (WMF) (talk) 10:42, 25 January 2014 (UTC)
 * Thank you! --Magnus Manske (talk) 16:56, 26 January 2014 (UTC)

+1 :) --Nemo 18:43, 26 January 2014 (UTC)

Last change time in CatScan 2
See e.g. this example - although the parameter says "last changed before 20110101", the "Last change" column in the Results table is odd, with seemingly all dates in 2013 or 2014. That doesn't seem right, as the articles in the list indeed were changed in 2010 or before. GregorB (talk) 00:17, 3 February 2014 (UTC)

CrossCats
Hello! First of all, I want to thank you again for this tool. I use it daily and it has been added as a default feature in fr:Modèle:Boîte des catégories, a template that appears in all the categories of the French Wikipedia. The thing is I sometimes get an error message. For instance, this happens when I use it in fr:Catégorie:Sport à Milan. Do you know where it comes from? Can it be fixed? Thierry Caro (talk) 16:18, 3 February 2014 (UTC)

Wikidata project stats
Suggest: Create a list like this taking only items into account that exist in five/ten (or variable number) Wikipedias. These would be more likely targets in any other language edition. I am really curious how many items exist in Wikidata with articles in five/ten Wikipedias, but not in English/Dutch/German/. For Bulgaria I found many items that existed in 4+ languages but not in English. Androoox (talk) 06:52, 17 February 2014 (UTC)

Wikimania 2014 London
Hi Magnus. I'm organising Wikimania this year, and it's a little different to previous years; bigger and more publicised, with a big focus on trying to recruit technical contributors. I hope you'll be attending, and would be keen to present on various topics. I'd be interested to chat to you; I think you'll be able to tell me about many interesting projects I've not heard of, and I may be able to help you by publicising the correct things. Let me know; ed@undefinedwikimanialondon.org EdSaperia (talk) 19:01, 11 March 2014 (UTC)

Wiki-ToDo
Wiki-ToDo is a really nice tool. Is it possible to feed it with an article rather than just picking one on random? My hope is that it could be possible to construct a user script where i can use something like 'https://tools.wmflabs.org/wiki-todo/index.php?lang=sv&page=' + wgPageName Is it possible in some way already? If not, would it be easy to add? Jan Ainali (WMSE) (talk) 15:04, 2 April 2014 (UTC)
 * Already in there. --Magnus Manske (talk) 16:59, 2 April 2014 (UTC)

MakeRef
I am hoping you can spare a few moments to rename "MakeRef/mr" to "MakeRef/Mr" and undelete "MakeRef/Fr" some time. Thanks! -- 79.67.246.210(talk) 21:40, 30 March 2014 (UTC)


 * Please also consider these changes. -- 79.67.241.252 (talk) 21:51, 8 April 2014 (UTC)

File:Jean Sylvain Bailly1.png listed for deletion
A file that you uploaded or altered, File:Jean Sylvain Bailly1.png, has been listed at Files for deletion. Please see the to see why it has been listed (you may have to search for the title of the image to find its entry). Feel free to add your opinion on the matter below the nomination. Thank you. Calliopejen1 (talk) 20:04, 9 April 2014 (UTC)

Anregung zum Tool add-information
Beim [http://tools.wmflabs.org/add-information/? add-information] wäre es sinnvoll eine Leerzeile zwischen der Lizenz und den folgenen Kategorien einzufügen. Der Quelltext liest sich besser. Das Tool nimmt irgendwie zwei Zeilen heraus. --Atamari (talk) 08:27, 16 April 2014 (UTC)
 * Aus

oder

wird regelmäßig == und == --Atamari (talk) 10:22, 16 April 2014 (UTC)

Move-to-commons-Assistent
Gibt es Probleme mit dem Move-to-commons-Assistent? der Aufruf funktioniert schon den zweiten Tag nicht. Kann man die Probleme beheben? --Atamari (talk) 07:44, 16 April 2014 (UTC)
 * Webserver klemmte, neu gestartet, läuft. --Magnus Manske (talk) 08:42, 16 April 2014 (UTC)
 * Danke --Atamari (talk) 10:22, 16 April 2014 (UTC)
 * kann sein, das es wieder klemmt. Als ich es ausprobieren wollte funktioniert es nicht. --Atamari (talk) 10:37, 16 April 2014 (UTC)

commonshelper
Hallo Magnus, seit einigen Tagen läuft der commonshelper nicht mehr. Kannst Du mal schauen woran es liegt. Danke und frohes Osterfest --Knochen ﱢﻝﱢ‎ 20:13, 19 April 2014 (UTC)
 * Warten auf Godot. --Magnus Manske (talk) 21:47, 19 April 2014 (UTC)

Catscan2 Outage notifcation
Just a notification to say seems to be offline. Wasn't worth raising a full JIRA issue about.Sfan00 IMG (talk) 10:26, 9 May 2014 (UTC)

Wdsearch
I'm wondering if you can help me solve an encoding issue in MediaWiki:Wdsearch.js. Could you please take a look at MediaWiki talk:Wdsearch.js? Whym (talk) 13:16, 14 May 2014 (UTC)
 * Please disregard above. ‎TheDJ and Mr. Stradivarius were working on it, and it now seem working fine. Whym (talk) 10:55, 15 May 2014 (UTC)

Novice question about WikiData queries
I'm new to queries like this so would be glad for some help. I tried this query: BETWEEN[569,1850,1900] AND NOCLAIM[570] hoping to find people with birth dates between 1850 and 1900 who have no death date but it doesn't seem to work right. That is, I get results for people who have death dates. This query on the lists page: BETWEEN[569,,1900] AND NOCLAIM[570] doesn't return any results. What am I doing wrong? Thanks very much! Mfbjr (talk) 19:20, 14 May 2014 (UTC)
 * Hmm, using your query I get >13K people born between 1850 and 1900 without death date, which is what you wanted, right? Maybe you accidentally ran CLAIM instead of NOCLAIM? The second query should work, but doesn't; I'll look into it. Meanwhile, try something like BETWEEN[569,-99999,1900], that should catch most of them ;-) --Magnus Manske (talk) 15:05, 15 May 2014 (UTC)
 * Thanks very much - I got the same number of results BUT those those people have death dates in the articles (or a good sampling, I didn't review all of them). Do you know how 570 is populated (that is, where does that data come from)? Is it coming out of persondata and/or infoboxes? If you're not the right person, my apologies - feel free to send me somewhere else. Mfbjr (talk) 20:17, 15 May 2014 (UTC)

Revert
this doesn't tell me much Magnus. What is broken and how am i supposed to know an en.wp script is being used by reasonator ? Could have put that in the header at least. —Th e DJ (talk • contribs) 20:26, 15 May 2014 (UTC)
 * I'm not sure which part of 's edit was incompatible with Reasonator, but is there any way to avoid using escape? Using escape causes broken links such as 千代田区 in Wdsearch results in Japanese. Whym (talk) 04:01, 16 May 2014 (UTC)
 * The script comes from Tools Labs, and is still used there; the reason it's on Wikipedia is that it's used in a JavaScript tool of mine on many Wikipedias, and Wikipedia scales better. Tools on Labs are not running MediaWiki, and, thus, have no mw object. Since you used that object without properly checking for its existence first, it broke some of my tools, including Reasonator. --Magnus Manske (talk) 08:12, 16 May 2014 (UTC)
 * But where EXACTLY is it used ? what URLs ? That helps keeping stuff NOT broken. —Th e DJ (talk • contribs) 12:17, 16 May 2014 (UTC)
 * Here, for example. --Magnus Manske (talk) 12:53, 16 May 2014 (UTC)

Catscan v2
Hi Magnus, I'm using Catscan v2 to analyse all the files in a certain category. The fact that it allows you to filter upload date and includes details on usage is very useful. There seems to be a limit to the size of category it can handle, is there a way round this? Richard Nevell (WMUK) (talk) 12:13, 6 May 2014 (UTC)
 * Example? --Magnus Manske (talk) 17:27, 6 May 2014 (UTC)
 * So for example if I run CatScan v2 on Category:Content media by years - Supported by Wikimedia UK - 2014 (nearly 38,000 files as of writing) to produce a list of the files in that category, it produces the message "Maximum potential result objects exceeded, aborting". Richard Nevell (WMUK) (talk) 15:11, 16 May 2014 (UTC)
 * I have upped the thresholds to include your query. If you need something simple, maybe try out quick_intersection. --Magnus Manske (talk) 15:33, 16 May 2014 (UTC)
 * Thank you, that's tremendous! Richard Nevell (WMUK) (talk) 16:10, 21 May 2014 (UTC)

Hello
The WIkidata-game did not report d:Q11691684 for deletion after merging it with d:Q3005481. (t) Josve05a  (c) 15:04, 21 May 2014 (UTC)
 * Same thing with d:Q16525260. It doesn't report around 1/20 for deletion of what is being merged. (t) Josve05a  (c) 11:15, 22 May 2014 (UTC)
 * And d:Q16069530 & d:Q15872588. Now it has stopped reporting for me... (t) Josve05a  (c) 11:19, 22 May 2014 (UTC)

ORCID & Wikidata user page
You can now use Authority control to put your ORCID (and other) identifiers on your Wikidata user page. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 17:06, 27 May 2014 (UTC)

Recent changes in category
Hi! ;) I wrote to you on jira. Can you tell me if you (or any other technician) could create tool like the one I described or not? Just want to know. Thanks, Tufor (talk) 13:13, 4 June 2014 (UTC)
 * Out-of-the-box, I can offer you this example: German scientist articles changed in the last 10 hours on en.wp. Is that sufficient? --Magnus Manske (talk) 15:56, 4 June 2014 (UTC)
 * Unfortunetaly, it's not. It lacks many features like time of the change, size change, author, summary, links to list of versions (history) and difference made in the change. The first catscan had it all and I kind of miss it (in fact not only me, couple of other users on pl wiki also commented on how the new catscan lacks these information). I also know that there is Deep out of Sight tool but still it's not it. I was thinking about sometking like Recent Changes but limited to one category (+subcategories) so that you can review what is up in the area of your regular contributions. Best, Tufor (talk) 19:32, 4 June 2014 (UTC)

Wikidata game - Suggestions for the "No date" game.
Hi Magnus,

I have been playing the "No Date" wikidata game for a while now, and i must say its one of the most addictive applications i came across in some time. I have a couple of suggestions though- three are related to date detection in specific languages and one is related to "generic" date detection. The content on the left of the arrow is the article's text, the content on the right a Google translation of the content.


 * GENERIC: (1896– 1968 ) -> 1896-1968
 * POLISH: "ur. 21 grudnia 1981 " -> "Born. 21 December 2981"
 * POLISH: "ur. 18 kwietnia 1921 w Zabłociu, zm. 30 lipca 1999 w Krakowie" -> born. April 18, 1921 in Zabłocie, d. July 30th, 1999 in Krakow
 * POLISH: "ur. 18 marca 1898, zm. w kwietniu 1940 " -> "born March 18, 1898., in April 1940"
 * POLISH: "ur. 21 grudnia 1894 w Stęgoszy, zm. 25 kwietnia 1981 w Poznaniu" -> "born. December 21, 1894 in Stęgoszy, d. April 25, 1981 in Poznan"
 * CZECH: "7. února 1852, České Budějovice – 26. července 1924 " -> 7 February 1852, Czech Budejovice - July 26, 1924
 * CZECH: "28. ledna 1896, Plzeň - 12. března 1962 " -> January 28, 1896, Plzeň - 12 March 1962
 * CZECH: "11. května 1909, Fort William − 14. srpna 1985 , Lac du Flambeau" -> May 11, 1909, Fort William - August 14, 1985, Lac du Flambeau
 * FINNISH: "o.s. Demys, s. 16. syyskuuta 1943 Orsay, Essonne, Ranska" -> nee Demys, p 16 September 1943 Orsay, Essonne, France
 * FINNISH: "22. marraskuuta 1967 " -> November 22, 1967

The format marked as "GENERIC" fails to highlight the starting year if there is no space in front and after the dash (somewhat common issue). The lines marked with a language display specific-language dates where the month and day format does not seem to be recognized. Naturally this occasionally occurs in every language, but for these three languages the above date annotations seem to be the default format (Or at least a common variety). It would be rather convenient if some detection routine for the above could be added to this game. Thanks in advance for considering it! Excirial ( Contact me, Contribs ) 23:59, 10 June 2014 (UTC)

Thanks, but more info needed
I'm using processing (programming language) got get information (e.g. on nato countries) as JSON. I still find the API-documentation very hard to follow. I get a JSON list of the numbers of the 22 NATO members all right, but how can I ask for name and description and wikipedia-url of the countries (when I have their ID)? Something like .....api?q=claim[22] ? Do have entries like name and description also an ID? Where can I find a list of those? It would be great if there could be an example with a comnplete url and not only bits and pieces - also a complete processing program would be great (I will submit mine here when I am done ;-)   ) Thanks a ton! Following entry is not by me: http://forum.processing.org/two/discussion/5714/wikidata-wikipedia-api-access-big-data-general-question#Item_1  — Preceding unsigned comment added by 62.91.26.166 (talk) 11:30, 12 June 2014 (UTC)


 * WDQ doesn't store names, descriptions, or site names (though you can query if a site link exists). Its purpose is to get you query results, not to be a complete Wikidata mirror. Once you have the numbers from WDQ, you can get such additional data from the Wikidata API (search the page for "action=wbgetentities"); just remember to prefix the numbers with a "Q" :-) --Magnus Manske (talk) 17:17, 12 June 2014 (UTC)


 * Thanks a lot... It still would be nice to have an example program of some sort...


 * done. http://forum.processing.org/two/discussion/5714/wikidata-wikipedia-api-access-big-data-general-question  — Preceding unsigned comment added by 93.196.87.167 (talk) 06:54, 14 June 2014 (UTC)