Commons:Bots/Requests/Polbot

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

Polbot[edit]

Operator: Quadell

Automatic or Manually Assisted: Automatic, supervised

Programming Language(s): Perl

Edit period(s) (e.g. Continuous, daily, one time run): In small batches

Bot flag requested: (Y/N): Y

Functions: Multiple.

  1. Uploading multiple images at once from sites that allow GFDL permission. (Most of my recent uploads have been done using tools in this way.)
  2. Categorizing at the species level (e.g.), so that Wikipedia articles that link to categories will see the full panoply of images there (e.g.).
  3. Surrounding English descriptions with {{en|...}} (e.g.)

Discussion[edit]

Stop hand.svg If you came here to discuss this bot flag request, please keep in mind that this is not a vote or a support-oppose discussion. This is a discussion that should be formatted like discussions in the Village Pump.

Greetings. I'm an admin here, and on en.Wikipedia, and I run a bot on enwiki named polbot. I've been using tools to automate repetitive tasks here, and a few users recommend that I use a bot account and get a bot flag here. Quadell (talk) 13:39, 3 July 2007 (UTC)

  • Will task 3 be supervised? If not, how will the bot determine which language something is? -- Bryan (talk to me) 14:10, 3 July 2007 (UTC)
    • Task 3 will be supervised. Basically, I'll look for a common English word (like "with") and change a couple dozen; then I'll go back and review each change, reverting myself when appropriate. Is that a good idea? Quadell (talk) 15:22, 3 July 2007 (UTC)
      • That sounds at least very difficult to implement. In case you want to continue this semi-supervised insertion of language templates, I guess a test run will be nice. No further objections, I am confident that the operator is willing to follow consensus on categorization. -- Bryan (talk to me) 19:38, 3 July 2007 (UTC)
  • No objections. Recent changes already flooded with plant images categorizations :-) --EugeneZelenko 14:12, 3 July 2007 (UTC)
  • Support the categorisations with a user with a bot bit. Also interested in the answer to Bryan's question. Siebrand 14:38, 3 July 2007 (UTC)
  • No objections. Was surprised to see my watchlist bursting with my butterfly images today (although a bot flag won't fix that!) Cary Bass demandez 15:42, 3 July 2007 (UTC)
  • oppose because of function 2. Even after a lot of discussions there is by far no agreement about this. -- aka 18:07, 3 July 2007 (UTC)
  • strictly oppose with the same reason as aka -- Necrophorus 18:23, 3 July 2007 (UTC)
  • oppose because of function 2. There is no agreement on policy concerning the category vs. gallery pages issue. Should not be any support for making policy by bot. --Franz Xaver 18:45, 3 July 2007 (UTC)
    • For those voting oppose about lack of consensus, it's true that there is no consensus about whether you should browse list images in category pages or articles. Many people feel different ways about this, and I am not trying to enforce one view over another. This bot will not, for example, remove images from articles. (That would be "making policy by bot".) But what this function would do is simply add category information, so that (until consensus is determined) both ways will work. On the English Wikipedia, for example, many pages link to commons articles, while others link to commons categories. Certainly it would be better if we had consensus on this, but for the time being we don't. Does that mean we shouldn't categorize individual images? Or that we shouldn't put images in articles? No. It means we shouldn't take images out of articles (or categories) until consensus is established. This function will certainly make Commons more usable for those who prefer categories, but it will not remove any functionality for those who prefer articles. I promise not to use this bot to make a point or force consensus or anything like that. All the best, Quadell (talk) 19:05, 3 July 2007 (UTC)
      And it will create much redundancy, i.e. for every article a category with the same name will be created. I can give a very simple example of the recent works of Quaddel with Category:Kopsia - two entries instead of one, both containing the same images. However, there are genus and family categories with much more articles included. This bot will double the number of entries in these categories. Most names will be included in both the pages and the subcategories list, but some only in one or the other. So both lists have to be checked, if one is browsing the category tree. This bot action will also increase the number of categories with more than 200 entries, i.e. with more than one page to check. In my opinion, doubling the structure does not make Commons more useable - on the contrary, it does make it more chaotic. Of course, one could work aroung this double entry problem, by hiding the species articles in the species categories. And that's the way, how policy is made by this bot: In order to keep Commons usable, species pages will be forced to get invisible from the supercategory pages. In the future, species articles will be visible only from the species categories as humble blue links swamped by an unorderd heap of images. --Franz Xaver 19:38, 3 July 2007 (UTC)
      I commented below, de-indenting. Quadell (talk) 20:00, 3 July 2007 (UTC)

No, no, I certainly don't want to force one position over another. I will not make species articles "invisible from the supercategory pages". If you notice, the categories I create (e.g. Category:Kopsia fruticosa) link to the articles directly -- they do not hide the articles. If my bot can do other things to make it clear that the articles exist and are useful, I will be happy to include such functionality.

Yes, it is redundant, and that's unfortunate. I too wish that we had consensus on whether to use articles or categories. You said that because of the bot, "both lists [categories and articles] have to be checked", but that's true either way. Since some people only add categories to their new uploads, and others only add their images to pages, both lists have to be checked, with or without this bot. This bot function will simply make it more useful for those who prefer categories, but it will not make it less useful for those who prefer articles.

Of course, those who prefer articles will say "We should keep it simple and not categorize images", while those who prefer categories say "We should keep it simple and not have articles at all". That debate has been going on for years. But in the meantime, what should we do? I think giving the users the choice of which system to use is best. Quadell (talk) 20:00, 3 July 2007 (UTC)

Yes, you do not make species articles invisible, but you create a situation, where someone else will feel forced to do this - simply in order to reduce redundancy. --Franz Xaver 20:32, 3 July 2007 (UTC)

P.S. It has also come to my attention that a discussion on the German Wikipedia is ongoing, wherein several users have banded together to try to prevent this nomination. (Users on this page are the several "no" votes who came at once.) I think all discussion should be out in the open, and I don't think such factionalism will help the project. Quadell (talk) 20:02, 3 July 2007 (UTC)

  • oppose also because of function 2. Biological names are not stable entities and can and do change rapidly due to taxonomical reasons on their generic and specific level. Thus categorizing species on its names is a ridiculous idea, which will cause tons of additional maintainance work. Compared to that titanic work of maintenance, moving a gallery page is quite easy. By that system, people will simply arrive on categories then whose name is simply utterly wrong or, even worse, will be directed to a category whose name confusingly differs, because of a correct article in the Wikipedia they came from, but a wrong/unmaintained entry in the Commons. Or they might never arrive there, because the article's editor only searched at the "correct" place. Denis Barthel 20:16, 3 July 2007 (UTC)
P.S.: Discussions in the German Wikipedia are open too. And as probably no Wikipedia has such an active community in zoology and botany as the German, you might excuse our special interest in this matter. Of even lesser use to the project it is, to start making drastical changes without having discussed them before with those who are concerned by this. Denis Barthel 20:16, 3 July 2007 (UTC)
With or without this bot, species information will have to be kept up to date. If the ICZN reclassifies a species, we will have to change the article (while keeping a redirect), and change all articles that point to this article. Also, we will have to change all applicable images that are in that species' category. This is true with or without the bot. I believe my bot will make it easier to keep this up to date, because if I am informed that (for instance) "Felis silvestris" is changed to "Felis domesticus", I can run a very simple script to point everything to the new name. Quadell (talk) 20:32, 3 July 2007 (UTC)
Ah. And so everybody who intends to maintain an article has to ask for you to run your bot, making dozens or even more of changes, instead only to move the article to its correct place in a second? By the way: Neither the ICZN nor the ICBN (did you forgot this institution? Or do you even know it?)classifies species. Botanists and zoologists do. Do you intend to keep up to date with systematic literature in the future? Are you the right guy to keep such a bot running? And will you actualize my list, that I can see new entrys in the category, where I am an expert? Denis Barthel 20:42, 3 July 2007 (UTC)
Don't forget the ICNB and the ICSP! ;-) Not that we have a lot of images of prokaryotes on Commons, but still. Look, let's keep this civil, okay? I'm not insisting anyone ask me for anything. I'm just offering to help move categories and pages, that's all. As it stands now, if someone simply moves an article, he will miss the images that are categorized by species and not included in pages. These already exist, regardless of Polbot, and have to be dealt with anyway. Quadell (talk) 20:59, 3 July 2007 (UTC)
In a well-maintained gallery pictures have no categories. Denis Barthel 21:00, 3 July 2007 (UTC)
Yes, with or without the bot, information has to be kept up to date. Anyway, it is less work, if there is only an article to be changed, compared to the article and a category. By the way, in order to change the name of an article, you only have to move the page and all images are moved at once - no admin and no bot needed. If you change the name of a category, you have to change every single image - you need a bot or a lot of time. That's one of the reasons, why I dislike species categories at all. (Changing links pointing to a page or to a category is the same effort.) --Franz Xaver 21:09, 3 July 2007 (UTC)
  • strictly oppose because of function 2. --Olei 20:46, 3 July 2007 (UTC)
  • Comment: I can see that those who prefer articles to categories have come out en masse to comment. That's fine. But please keep in mind: I will never do anything to make it less useful to use articles. There are those who prefer articles, and I respect that. Please respect those who prefer to navigate by categories, and don't vote against a bot just because it will make it easier for these people. Quadell (talk) 20:59, 3 July 2007 (UTC)

I would also like to bring everyone's attention to our relevant policy on the topic:

Some users are of the opinion that when an image is included in a gallery article, it is sufficient to categorize that article. Others believe that each individual image should be categorized too. The consensus on this issue, as determined by vote, is that both systems are equally valid and should be used concurrently.

If our consensus is to use both, why should I be prevented from making edits which use both? Quadell (talk) 21:53, 3 July 2007 (UTC)

As someone will most probably soonish point out, there is even no consensus on the quoted consensus :(. I fully Symbol support vote.svg Support all the proposed functions for this bot, and I am a strong believer in both preserving information and not removing any information that was once added to a media file, other than by refining present information (i.e. not removing a media file from a category because it was added to a gallery page, but only change the category because there is a more specific category the media file may belong to), and the possibilities of categories and galleries being able to co-exist. In all discussions I have seen, the most extreme points of view have come from those not wanting categories, while the other half (it appears to be about 50-50) just thinks that both should be possible. Here we have someone offering to automatically provide us the best of both worlds, or better: the best MediaWiki can offer at this point in time as far as it is possible to maintain a file repository in the software, and he is being bashed. He *is* offering to put in the work, he has made clear that he *will not* move articles from the categories where they are now. Should the other half simply navigate as the other other half prescribes? Let us do both and flourish. Cheers! Siebrand 23:05, 3 July 2007 (UTC)
  • strongly support because of function 2. Commons is very far behind in categorization of images and this will hopefully fix that. If you don't like viewing categories, then don't view categories. It seems silly to deny those that prefer browsing categories the opportunity to do so. Commons:Categories states: The category structure is the primary way to organize and find files on the Commons. Cacophony 22:54, 3 July 2007 (UTC)
... and it continues: It is essential that every file can be found by browsing the category structure. To allow this, each file must be put into a category directly, and/or put on a gallery page which is categorized.. Please do not cite in a manner, which distorts the intended meaning of a Commons-guideline. Denis Barthel 07:49, 4 July 2007 (UTC)
Ehhm .. let me guess: you both have a firm knowledge on the special aspects of scientific names in biology? Uh? Ah, okay. Just wanted to know. Thanks. Denis Barthel 23:15, 3 July 2007 (UTC)
Please remember that targeting the user instead of the topic does not help you to make a stronger case. Please remain civil and stay on-topic. Thank you. Siebrand 06:09, 4 July 2007 (UTC)
As in this topic the special aspects of scientific names make an important point, discussing ones skills on the topic is on-topic. Simple rule-fetishism which completely misses the point and is obviously solely based on dogmatic principles instead well-thought, is in fact a reason to ask for the qualification of a contributor. I'd never give a statement on bridges or photography. I am clever enough to take a back-seat there. Denis Barthel 07:49, 4 July 2007 (UTC)
  • support sounds good to me. anything that helps in cleaning up gets my support. Gryffindor 00:18, 4 July 2007 (UTC)
  • sounds reasonable. Doronenko 05:44, 4 July 2007 (UTC)
  • support Undecided on 1. Support 2. Support 3 if supervised. /Daniel78 07:13, 4 July 2007 (UTC)
  • Strictly oppose due to function 2. The reasons are alredy explained above. As experience shows Quadell is nothing cleaning up but producing lots of messes. Category copies of species articles don't give any further information but if pruduced in masses are confusing and soon annoying. Stop that nonsense! Ies 18:19, 4 July 2007 (UTC)
  • Strictly oppose I'm not exactly informed about the rights and duties of this "Bot-version".

Due to the experience of the last days I think, nobody should obtain a roboter - tool or be an admin, who without any sensibility corrupt the sensefull and usefull performances of other users. One view to the project: "Tree of Life " would have showed, that there is a not finally ended discussion in question of categories or articles in rank of species with good arguments of both sides. Thousends of hours in botanical and zoological areas are now impossible to recover or repair.

An now a question for my personal handling to the experts: In my special-watchlist are more than thousand articles. What ist the way to get knowledge of new and correct or not correct classified pictures in a species category? I thank for information. Orchi 19:05, 4 July 2007 (UTC)

That's a good point. The watchlist function enables to see imediately, if someone has added an image to an article. You can check instantly, if its identification is correct. (Anyone who has done some maintainance work should know that a quite high percentage of photos uploaded here is misclassified.) This is not possible with categories, as the category file is not changed when an image is categorised in it. Having images by default ónly in categories, makes quality assurance much more difficult. --Franz Xaver 20:32, 4 July 2007 (UTC)
Strange things happen. I was just typing something similar. That´s exactly what I was thinking about this evening. For instance, if someone adds a picture to the Category:Kölner Dom, this person surely knows she/he actually took a picture of this building. However, with plants this is a totally different matter. There are too many species that look very much alike but belong to different species. This means every species category has to be checked manually to look for misidentifed plants. I must say I oppose point 2. --BerndH 21:01, 4 July 2007 (UTC)


  • Comment. Keep in mind that the only thing this bot would do for function #2 is add a category to images that aren't categorized. I must admit, I'm mystified when people talk about "Thousends of hours in botanical and zoological areas are now impossible to recover or repair." It's just added categories. If consensus were that images of plants should not be categorized by species, I could (and would) run the bot to remove categories from these images. I only want to follow consensus. But the consensus is to use both articles and categorization, and that's what the bot does. Perhaps some users don't like the fact that our consensus calls for this, but it isn't helpful to take it out on me. Would it be problematic for a person to manually add categories to biological articles? No, of course not, which is why our policy says that such categories should not be removed. So why is it a problem for a bot to it, if it's not a problem to do this manually? The only conclusion I can come to is that there is a vocal minority who does not like our consensus policy to use categories for images as well as articles, and rather than trying to change policy, they simply want to prevent it from being acting on. I can't agree with that. Quadell (talk) 23:26, 4 July 2007 (UTC)
Can you give me a link to "the consensus is to use both articles and categorization" in botany- and zoology-images? Ah and btw: why don't you simply promote your idea in Commons:WikiProject Tree of Life? We can see then about minorities and all that. Denis Barthel 23:55, 4 July 2007 (UTC)
I quoted the Commons:Categories policy above, but I'll quote it here again for your convenience. "Some users are of the opinion that when an image is included in a gallery article, it is sufficient to categorize that article. Others believe that each individual image should be categorized too. The consensus on this issue, as determined by vote, is that both systems are equally valid and should be used concurrently. Do not remove categories just because an image is in a categorized gallery." There is no exception for photographs of plants and animals. I'm certainly open to discussing this at the ToL Wikiproject, but this is a Commons-wide policy which is not restricted to one particular project or another. Quadell (talk) 02:25, 5 July 2007 (UTC)
Quadell, even if you repeat it a thousand times: this is not a policy. It is not even a guideline. It is just a quotation of a help page. Help pages are neither rules nor dogmas nor decisions nor anything else of a binding character. They do not need a vote or a consensus to be changed, their only intention is to instruct the beginner. Neither all the above voters nor I are beginners. Many of us are part of the commons since a long time or even since the beginning and are having a solid overview on the effects of categories and galleries on their separate levels.
Furthermore, even if your quotation would be of a dogmatic character: I prefer understanding and sanity instead of rules, laws and dogmas. This has been the reason for the different "Ignore all rules"-pages in our sister-projects, which clearly state "If the rules prevent you from improving or maintaining Wikipedia, ignore them.". Though this are the Commons and not Wikipedia sensu stricto, it is not even lesser true here. Denis Barthel 06:38, 5 July 2007 (UTC)
Quadell, please, don't misrepresent the old vote on this issue. There was no consensus by vote in favor of the "mixed system", but by missing of a clear result of the poll the mixed system survived, as it was the system used before - obviously an unsatisfactory system as its existence had triggered this undecided vote. Actually the "mixed system" had significantly fewer votes than the pure "normal pages" and "categories" alternatives. Moreover, one should read carefully the description of the "mixed system" alternative: "Images can be both on normal pages and in categories; some can be both, but only images that are neither are considered 'defective'." This doesn't not call for creating double structures, but only allows it ("some can be both") under given constraints. In my opinion, every plant and animal image that is not cross-checked for its correct identification is 'defective'. And dumping it unchecked by default in species categories is a danger for Commons, as quality assurance is only possible with very big effort, given that watchlist function doesn't show images recently added to a category. Let me describe how the ToL stuff should work because of practical reasons of quality control: Images that are checked for correct identification should be only in species pages, not in higher categories and not at all in species categories. An uploader has the choice to add it to a gallery (and it can be checked by the person who has this page on the watchlist) or to a higher category - both would be OK. All images included in higher categories (genus, family) can be considered of being of uncertain identity, have to be checked and - after confirmation or correction of the identification - will be moved into a gallery page. This separation between images in normal pages and categories allows a straightforward procedure for quality check. Having every image only at a single place, either in a page - if correctly identified - or in a category - if unconfirmed - helps to keep an overview for the person who does category maintainance. And it is also easier for a person who is searching an images, as it doesn't force him to browse through redundant structures. In my opinion, a pure category system endangers the project out of reasons I described above, but a double system of mandatory species categories additionally to already existing species pages is even worse, as quality control gets lost in a jungle of redundant structures. I did a lot of maintanance work in the described manner, corrected many misidentifications, but I stopped this, when I was accused of doing harm to the project, because I had removed a correctly identified image from a family category, that was already included in its correctly categorised species page. I am not willing to restart any activity of checking species identifications, if I have to do this in an environment that systematically is impeding this work and if I take the risk to be accused of vandalism. --Franz Xaver 09:22, 5 July 2007 (UTC)
    • Thanks for your explanation. All I want to add is that I wasn't trying to misrepresent the result of the vote. I understood that the vote resulted in consensus to use both systems (since that's what the help file says the result was). Quadell (talk) 11:54, 5 July 2007 (UTC)
No objections (remember, this is objections or not, not Support/Oppose) as long as clear consensus is obtained for operation #2 before carrying it out. ++Lar: t/c 10:53, 5 July 2007 (UTC)

Important update: I can see that there will not be consensus to run Function #2. I have withdrawn this request and will not run this function. I am still requesting a bot flag, but only so that I can run Functions #1 and #3. I may wish to perform other bot functions in the future, but if I do I will request authorization here and see if there is consensus. If Polbot is given a bot flag, I give any admin permission to block this bot if any edit seems unhelpful or controversial. (For non-admins, just leave a message on my talk page. If anyone objects to something Polbot has done, I'll stop the bot and look for consensus before running again.) Thanks, Quadell (talk) 18:54, 5 July 2007 (UTC)

No more objections. Support Thanks for being insightful and not stubborn. Denis Barthel 19:06, 5 July 2007 (UTC)
No more objections after withdrawal from function 2. @Quadell: Thanks for you understanding. --Franz Xaver 08:31, 6 July 2007 (UTC)
No more objections after withdrawal from function 2. -- aka 09:16, 6 July 2007 (UTC)
No more objections. --Olei 21:08, 6 July 2007 (UTC)
Just in case it is not clear, now that #2 function has been withdrawn, I have no objections whatever. ++Lar: t/c 05:25, 7 July 2007 (UTC)
Looks like a good bot to me. I would actually support function #2 as well, but not until there's a more clear consensus on a consistent category scheme for the TOL-related files. --SB_Johnny|talk|books 10:34, 7 July 2007 (UTC)