Commons talk:WikiProject Time

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

This is a rough draft of guidelines and discussion of proposed policies and practices for the Wikiproject Time.

Translation or creation of discussion in other languages is encouraged.

Years[edit]

Years are numbered according to the Gregorian calendar. Individual years are to have categories (eg, Category:1911) for media from that year. Years also can have articles (eg, 1901), the purpose of which is yet to be decided.

Year categories[edit]

Year categories are in supercategories (eg, Category:1730s; Category:Years in the 18th century). Year categories have sub categories. People who were born or died in specific years should be put in birth or death subcategories (eg, Category:1911 births, Category:1911 deaths). A common sub-category is paintings by year (eg, Category:1911 paintings).

Less common sub-categories may be made where such can be uesful for organizing multiple images within a year category, but need not be applied to other years if there are few or no similar images in those other years (eg, Category:1911 baseball cards, Category:New Orleans Mardi Gras 2005).

Year articles[edit]

There is currently disagreement over what should and should not be included in year articles. Suggestions and proposed guidelines welcome.

  • User:Emijrp and I have come into conflict over the year pages, specifically Emijrp's putting images not from that specific year on the page. I would prefer individual year pages to have images from that year (or images of objects created in that year) only, perhaps as a gallery of the more important or interesting images within the year category. If Emijrp has a different conception I hope Emijrp will elaborate on the alternative suggestion. -- Infrogmation 19:10, 3 August 2006 (UTC)
I can see putting an image representative of a year when others are not available. (Eg a flower power painted Van which was actually taken in 1974, but include it in the section on summer of love for 1968. Other than that, I don't see why you'd put in an image from a wrong year.
Speaking more largely, my general take on articles that the same split of coverage between articles and categories. Regarding articles- Years ago, Life Magazine published the Year in Photos edition. I think a Year article would have that kind of feel. Categories on the other hand would be totally inclusive- they exhaustively have everything for that year. I think the only criteria for category inclusion should be an analogue of determining allowable uploads to Commons. If someone searching on a date could concievably want to see a particular image in response to that query, then it should use the year category corresponding to the image.
Image Gallery articles try to present the best of the best and correspondingly, I beleive year articles should maintain editorial standards for high quality. For example, gallery articles exclude particular images due to :
  • repetition, ( the umpteenth bowshot of BB Iowa need not be included)
  • poor technical quality (low res, out of focus, motion blur, frame or photographer watermark not cropped
  • poor composition (the object in the image is only a minor portion of the composition)
  • Not representative image- the item has been modified so that gives a false impression of the thing that is the subject of the gallery article.
None of this list needs to be codified. But I only itemize to give the drift of the kinds of criteria people would be using with article inclusion decisions. The Year articles should focus on presenting the most representative slice of the times for that year. Just what is representative may be the subject for controversial years, Eg 1968 in Paris, years with events of great religious significance, etc. But that sort of thing can be fought over on the talk pages for the particular years.
  • One thing I anticipate is that perception of the importance of events is highly relative to geographic location. An image of a politician's or athelete's upset victory might produce images of immense significance in the particular country where they occured, but be virtually unknown elsewhere. So possibly the Year articles could be sliced along very general geographic boundaries- eg. Africa, Europe, Central Asia, Western Pacific rim countries, etc. Just a thought, but some sort of slicing like that might reduce a lot of the "that's an irrelevant/noise image" type of argument that wastes everyone's time.-Mak 19:36, 3 August 2006 (UTC)
    • I did a rough reworking of 1919 as an experiment. Thoughts on what works, what doesn't, how to improve? -- Infrogmation 21:22, 3 August 2006 (UTC)
      • This is going to be so cool. One question is how it scales up. I guess you have to be super concise with the images selected for each year, and you could link to a "more detail" category like "fashion in 1919" in the case of styles. I was chatting with Deusey about how they use combination categories to do such things on WP. It really is scary when you get into highly detailed taxonomies- you could have things like German destroyers of the 1920s and even worse. This is not a problem specific to time but it comes up here because you are navigating by time, and users are very likely to be interested in subsets- Like Stan is likely to get goofey about the stamps and the ships but not much else. So it would be nice to move not just horizontally in time from year to year, but move down into more detail as will be necessary since you probably won't have pages big enough to handle all the very good images available in one year- eg 1945.
      • Anyway, long story short- There are two approaches- one is to make a list which is readable by a template. That's not hard to do, and you may have guessed- I would put such a thing in an info page. However, we could as easily do it in a bot script. So long as the bots work off of some commands that users can control, and that someone actually fire off the bots, that is not a problem. In a bot, you basically do set logic to take the intersection of German navy, Destroyers, 1919, and you have a "more detail" page to jump to. When people have enough time to hand maintain particular ages that could be on such a potentially deep tree of pages, the bot need not auto generate that particular one any longer.
      • But I suppose that is looking way down the road. For now, I just look and see that there will be possible scale/ editorial pressures for inclusion on the very dense years.-Mak 02:28, 4 August 2006 (UTC)

I strongly feel that the many images from other years (or whose year is not known) which User:Emijrp added to year articles at present do more harm than good. People accessing those pages, I think, tend to assume that the images are from the year of the page article-- a reasonable assumption unless the images are clearly labeled otherwise! I appreciate much of the good work Emijrp has done, but I think those year articles at present are a mess. I strongly want any images included in a year page different from the actual year to either be removed or clearly labeled with the actual year and an explanation of why they are being used to illustrate a year other than their own. (If we cannot get that, I think we would be much better just blanking all those image galleries on the pages and starting fresh with year appropriate images.) -- Infrogmation 16:34, 11 August 2006 (UTC)

As I have made such requests for weeks now without response, and I spot numerous images on year pages which are unlabeled, mislabeled, &/or misplaced by years or even decades, I have started re-researching and fixing some year page galleries, and simply removing others when I have no reason to trust them. -- Infrogmation 03:11, 19 August 2006 (UTC)
I have created a new version of 1901. Suggestions for further improvements welcome. -- Infrogmation 14:41, 19 August 2006 (UTC)

Proposed year article guidelines[edit]

While there are details about the year articles yet to be discussed, I would like to submit the following as a preliminary guideline:

Year articles should feature highlights or representive samples from within the category for that year. Selected images or media should have a short caption. As a rule, images should be from the specific year of the article, or be images of objects created in that year.

I really don't see the need to include images from other years. If others feel they are necessary to illustrate something, I would suggest the additional guideline "Any images from other years included in a year article should be clearly labeled with their actual year and an explanation for their inclusion in the year article." Other suggestions? -- Infrogmation 14:38, 19 August 2006 (UTC)

Paintings centuries: Suggested name modification[edit]

At present, paintings by year are in supercategories with names like "Category:XVII Painting". I don't like those names-- there is no need for "Painting" to be capitalized, and I think the names should be in the format "17th century paintings" or "XVII century paintings" -- a slight preference for the former, as that fits with existing Category:17th century. I suggest those paintings century categories be changed accordingly, after first also asking at the paintings talk page and presuming there is no serious objection. Thoughts? -- Infrogmation 19:10, 3 August 2006 (UTC)

It is an archaic cataloging convention that has continuity stretching back to latin speaking monks in the middle ages. These conventions tend to propagate themselves in perpetuity. The cats on fr:wikipedia are named that way, but I agree- I think it is completely unnecessary to propagate roman numerals especially since this would force use on users who have no cultural link to Roman numerals. What are we going to do- label paintings from the Han dynasty this way too? It is silly and everyone understands arabic decimal numerals you propose. -Mak 19:50, 3 August 2006 (UTC)
I have added a note about suggested modification at Category talk:Painting by century, and on the talk page of User:Semnoz who created the paintings by century categories in their current format. -- Infrogmation 16:45, 11 August 2006 (UTC)
If there is no objection after another week, I suggest we go ahead and move the painting century categories. -- Infrogmation 14:29, 19 August 2006 (UTC)

Navigation bar[edit]

Mak has suggested a navigation bar. Possible prototypes:


< <17th century < −18th century 19th century 20th century−> 21th century> >
< <1840’s < −1850’s 1860’s 1870’s−> 1880’s> >


< <16th century < −17th century 18th century 19th century−> 20th century> >
< <1774 < −1775 1776 1777−> 1778> >

Notes on the templates[edit]

  • By the way, these have an option for color coding so they can for example be sepia toned for the 19th century. Lots of other formating options in that template.
  • Regarding the decade century navigation template above, I proposed a modification for individual year where you would have an entire row of individual years with the decades row being shorter. I suppose you could have a century row above that. I can whip one up if anyone wants that sort of navigation on their pages- eg. Clothing styles might be cool that way. -Mak 20:03, 3 August 2006 (UTC)

Why not Timelines?[edit]

First question folks might have is- Why not use Timelines? Actually, that is the first thing I tried. The answer is that timelines are incompatible with templates. So this makes them only manipulable with folks that know how to run bots. If people are ok with that, I can run a bot to generate real timelines, but it would be a huge pain for anyone else to make global changes. So I built a template that folks could manipulate much more easily than bot generated timelines. If folks prefer timelines by bot, I will generate them, but I am not going to take time to run the bot everytime someone has a new formatting idea.

I'm fine with either suggestion. Templates seem to be used in more contexts at present. -- Infrogmation 22:27, 5 August 2006 (UTC)

Tours and Time[edit]

I have implemented a Temporal Tour, Tour:USS Nevada- Pearl Harbor attack which is a time sequence of images taken over several hours on December 7, 1941 at Pearl Harbor. The interface is a simple language neutral 1234 etc. Bar. If you click on one of the numbers of the tour, you will be taken to a particular image, where you will see the same navigation bar (since it has been transcluded there as a template). The index only appears on the title page, but can be transcluded on each image if desired. This page could be used as a template for other such temporal tours. -Mak 20:07, 20 September 2006 (UTC)

Decade articles[edit]

I think decade categories are useful. Decades in general (eg, Category:1930s) have been in existance for a while. Some subcats (eg, Category:1920s automobiles, Category:1910s fashion have been in use perhaps nearly as long. I started a couple of other decade subcats, for example Category:1920s paintings seemed useful for paintings where the decade is known but not the exact year. So, questions:

1)Does anyone object to decade cateories? Presuming we decide to continue using them, additional questions:

2)How general should we make them?

3)How should they be categoriezed and should individual years in decades have the decade category?

My initial thoughts are to extend them back as far as we have any significant material that would be useful -- I'm not sure how far that would be, but off hand I'd say at least back to the 1400s for both years and paintings. I'd like to see individual years within the category of the decade they are in, for example Category:1925 paintings is in Category:1920s paintings. Other thoughts? -- Infrogmation 22:27, 5 August 2006 (UTC)

Decade categories are absolutely essential to the Ships scheme. From the period 1830 to 1960 there have been exceptionally rapid changes in technology. Grouping by year is too fine grained, and grouping by period (eg victorian era) is far too coarse since you have square riggers and coal burning screw driven iron battleships in the same period. Decades is just about right for ships.
I can see that during very narrow periods, a year category might be useful- eg during WWII. I would expect to list a decade cat in the year cat for ease of navigation (scoping the data to a coarser level). It makes no sense to jump out to the century level.
I expect that I would see Fashions making heavy use of this fine a granularity. 1966 to 1967 was and immense shift, as were other seminal years when bold new looks came out. But even for more ordinary years there are all sorts of fads. EG the year in the 70's when all the guys were wearing white shoes and white belts- Maybe 1972? Scoping out to the decade makes sense- so a decade supercat makes sense in all the single years. Putting them in a century cat totally makes no sense- bundling starched high button collars with hippie tie dye is a little silly.
I can see that whether individual years are allowed could be decided on a case by case basis, but certianly if individual years are in place, certainly I would want the decade as the next level up. -Mak 05:47, 6 August 2006 (UTC)

I don't mind to sort years, births, deaths, automobiles, and paintings categories by decade, and then by century.

Category:1916 (births|deaths|automobiles|paintings)
|_Category:1916
|_Category:1910s (births|deaths|automobiles|paintings)
|_Category:1910s
|_Category:XX (Painting|Births?|Deaths?|...) Someone proposed to rename this categories
|_Category:(Painting|Births?|Deaths?|...) by century
Category:1916
|_Category:1910s
|_Category:Decades in the 20th century
|_Category:Decades
|_Category:20th century
|_Category:Years in the 20th century
|_Category:20th century
|_Category:Years

This is the scheme by now, isn't it?. --Emijrp 06:33, 6 August 2006 (UTC)

Limits to usefulness of some individual year topic categories[edit]

User:Haabet and I have been having a long-running dispute over the sub-categorization of Category:19th century fashion. It's my position that as a purely practical matter, people looking for illustrations of historical clothing styles are likely to be more often interested in a certain period of consecutive years (e.g. a 3-year "window" or 5-year window or 10-year window or 15-year window or whatever), rather than in one and only one specific calendar year. In such cases, having decade categories makes it easy to find intriguing images by only searching through one or a few categories (instead of having to slowly poke through a lot of tiny separate individual year categories). Also, many images don't have available exact year information, but are often relatively easy to assign to a decade (due to the nature of Victorian fashion changes). And separating things out by decade correlates with what's being done on English Wikipedia under en:History of Western Fashion.

By contrast, Haabet seems to believe that decade categories are an evil blurring of the pure clarity of individual year categories, and Haabet would prefer that Category:19th century fashion consist of 100 separate individual year sub-categories, with no internal structuring. The current truce situation is that no individual decade categories are linked before 1858, so that all 1800-1857 images are placed in decade categories. For 1858-1899, Haabet has uploaded many drawings of corsets, or crudely-colorized illustrations from patent applications, which I regard as being of rather marginal value for fashion history, so that I'm perfectly content that they should stay in relatively inaccessible year categories -- but I try to put images from this period which do have real value into decade categories, and sometimes we argue about this.

Having both decade and individual year categories might be a partial (if rather redundant) solution to part of these problems, but only if it would work automatically -- because relying on the manual efforts of individual editors to bring clarity to the situation by hand-editing hasn't really done too much so far... Churchh 07:05, 16 August 2006 (UTC)

I recognize this class of dispute. It has to do with categories as a navigation device, or categories as a librarian organizational tool. The two conflict severely at times, depending on how inflexible the organizational structure/ advocate is. The way I have bypassed this is to work on navigational links that are distinct from categories. The timelines interface mentioned earlier is one instance, the 1234 interface on page Image:Browning M1919A4 Soldier 1949.jpg is another. The trouble is that the little bastards have to be tediously updated.
  • The general choices seem to be that you either to create huge numbers of transcluded "webring"/ trails list templates which must be hand updated, or you do a bot automated solution where you periodically regenerate the lists.
  • I have opted for bot generation, since you wind up with static lists that folks can manipulate. I know I will not be around forever to run the bot, but I could create a script on Pywikipedia that would support such a thing and people could pick up the gauntlet on that day (I suppose is probably inevitable) when I tire of contributing so much time to Commons.
  • The lists will be generated by Boolean combination of categories. EG cat1 and cat2 would generate an intersection of two categories- useful for something like generating a navigation list from a category that doesn't exist- eg. WWI ships and battleships. For the case of combining a bunch of cats like your case of fashions pigeonholed into a bunch of single years, you would Cat1 OR Cat2 OR Cat3.
  • Such lists would be available for each category and be indicated in a template associated with that category.
  • I have been mulling this over since last spring and think I shall get around to implementing it soon.

-Mak 02:56, 21 August 2006 (UTC)

Categorizing years by decade or century[edit]

Okay, we seem to agree that decade categories are useful. Do we need to choose either/or in categorizing years by century or decade? I suggest we do both; I suppose it could be considered technically redundant, but I think it improves navigation options while doing no harm (for example, Category:1965 in both Category:1960s and Category:Years in the 20th century).

They should both be there for ease of navigation. Good luck with the anal rententive crowd though.-Mak 23:45, 6 August 2006 (UTC)
Well, as we agree that both makes navigation easier, let's make that the Wikiproject Time standard if there is no objection. -- Infrogmation 16:46, 11 August 2006 (UTC)

One Example[edit]

It's impossible to put Image:Cocacola-5cents-1900.jpg into a specific "years in advertising" category, but if there were a Category:1890s advertisements category, it could be put there. Churchh 15:46, 20 August 2006 (UTC)

Date categories being speedied[edit]

Date categories (e.g. Category:15 September, Category:Days in August) are currently being emptied and speedied en masse. Comments? Man vyi 15:37, 15 September 2006 (UTC)

Have started restoring missing cats: Man vyi 08:27, 4 November 2006 (UTC)

Month categories within year categories[edit]

Some people have created month categories within year categories. If people find these useful, I have no objection. However I think if we use them, naming should be standardized-- for example, Category:January 2005 verses Category:2007-05. I lean towards using numbers rather than month names as that seems more language neutral. Other thoughts? -- Infrogmation 16:28, 28 April 2007 (UTC)

navbox for time-place[edit]

This should be self explanatory proposal. Consider the following:

1890s: Paris (Commons)" Paris 1895-06-15 mapsearch Paris //commons.wikimedia.org/wiki/Commons_talk:WikiProject_Time

It is produced by this:

{{Places by decade |Paris |1890
 |image    = Camille Pissarro 007.jpg
}}

Zero to 3 images in gallery. More info on Template talk:Places by decade

-J JMesserly (talk) 21:13, 14 January 2009 (UTC)

  • Thanks for your good work! One detail-- your new boxes are displaying parent categories in addition to specific categories. This is redundant in the category system (For example, Category:San Francisco in the 1890s doesn't need to display "Category:1890s", since it is already in the subcategory "California in the 1890s" which in turn is a subcategory of "United States in the 1890s".) Any way to fix this? Cheers, -- Infrogmation (talk) 05:23, 15 January 2009 (UTC)
Wow. I can't understand how such a glaring error slipped past. That error should disappear when I upgrade the stable version sometime tomorrow.

BTW, this mutates into a different layout for File pages, going into a minimal collapsed mode, no galleries, just the timeline expanded. See example at File:Buffalo Bill's Wild West Show.jpg-J JMesserly (talk) 09:49, 15 January 2009 (UTC)

As described on its talk page, the template is emitting hCard and hCalendar microformats, neither of which are populated correctly, A better method of emitting such microformat is described here, having first been proposed in 2007. Andy Mabbett (talk) 21:48, 2 February 2009 (UTC)
Not many agree with your Mr. Mabbett's personal opinions of correctness in the microformats community, or in the geocoding community. What matters is if the microformats community has particular recommendations regarding how the foundation wikis go about exposing information. This is an effort that is currently ongoing, and progress is being made. -J JMesserly (talk) 06:26, 3 February 2009 (UTC)
The fact is that more people agree with me than disagree (and what disagreements do exist are about procedures and policies, not the semantics of mark-up); but if you don't, perhaps you could try to persuade me and others that I'm wrong by refuting the points I make, rather than by repeating your ad hominem attacks, like those for which you have already had to apologise for, here? Perhaps you might also find evidence of any support for, or even an example of, your habit taking a page with a date of (for example) 1877 and emitting it as a bogus metadata date of 1875-06-15; meaning "the 24 hour period between 1875-06-14 and 1875-06-16". Andy Mabbett (talk) 10:57, 3 February 2009 (UTC)
I am pointing out you are asserting controversial positions as if they were fact. The point is that very few people on Commons besides myself and a few others like User:Para understand the ramifications and the technical details of what you are asserting. In the case of dates, the elementary issue sighted above seems easy enough (my response to you is the same as before- precision is a detail implementation issue that is on the list of refinements). But consideration of dates if arbitrated on Commons would have to be complete, and arbiters on commons would need to consider issues such as the exclusive nature of a dtend property and what that implies. A lay audience sees that you are asserting authority on the subject matter. I am pointing out that this is in question, and suggest that the forum for arbitrating these sorts of propositions regarding correctness of microformat encodings is not Commons, but microformats.org. I choose to follow their guidance, not yours. But if you feel that you have the correct way of doing things, that may well be the case and you may be correct as you wrote on my talk page that "there are no authorities in that community" and that we should all just take your word for it. It is a surprising statement, and you will forgive me if I decline. Nothing however prevents you from doing as User:Swift suggested and creating templates that do things the way you think is correct, and persuading the community to adopt them. Fair enough? -J JMesserly (talk) 17:26, 3 February 2009 (UTC)
You may well think you are "pointing out [I am] asserting controversial positions as if they were fact", but you are doing so fallaciously, since the position is already documented in the links I gave in my previous response; work in existing parsers and are widely deployed, with no controversy on our sister project, Wikipedia. You also now appear to be trying to put words into my mouth; kindly do not. False precision is not merely "a detail implementation issue"; it is a problem for Commons and it is not supported by the specs you day you have read (do feel free to cite them, if you disagree). Your "list of refinements (where is this documented, please?) is of only hypothetical interest, while you are publishing bogus metadata on live pages. "dtend" properties are not required. You are ignoring advice from the microformat community. While I may be at liberty to do as Swift suggested, I am also free to expose the damage you are doing and to ask you to desist from doing more (and have offered you explanations and assistance, even though I also believe your efforts are redundant and misguided). Fair enough? Andy Mabbett (talk) 19:51, 3 February 2009 (UTC)

My response is the same. I don't think that Commons is the proper forum for this debate. If you can attract sufficient people interested in hearing and carefully considering your position here, then we can revisit this. Otherwise, I shall continue discussions at Microformats.org. -J JMesserly (talk) 19:59, 3 February 2009 (UTC)

You don't think Commons is the proper place to discuss the metadata emitted by Commons?!? Andy Mabbett (talk) 20:34, 3 February 2009 (UTC)