Module talk:Messtischblatt
Jump to navigation
Jump to search
Categorizing into redirects
[edit]Category:Historical maps of Dortmund redirects to Category:Old maps of Dortmund so it ends up in Category:Non-empty category redirects. The same happens for Category:Historical maps of Duisburg, Category:Historical maps of Essen, Category:Historical maps of Mülheim an der Ruhr.
Maybe
- a check for redirects can be added (similar to {{Resolve category redirect}})
- or the logic inverted (add Old maps if present, not historical maps).
Enhancing999 (talk) 14:34, 5 June 2024 (UTC)
- Special:Diff/881566476 seems to have fixed it. Enhancing999 (talk) 16:30, 5 June 2024 (UTC)
- You've 'fixed' generic code to deal with your special case only. There are categories where your findings above are not applicable and for these the former correct categorization is not done anymore. You suggest to check, if a category page is a redirect before categorizing to it, but this is not what you've done with your 'fix' in the code. --84.135.125.242 23:59, 5 June 2024 (UTC)
- Do you have a sample where this broke the former "correct" categorization"? Enhancing999 (talk) 06:23, 6 June 2024 (UTC)
- It broke any categorization where Category:Historical maps of <x> is a valid, non-redirecting category while a Category:Old maps of <x> exists as well. In addition you did not check for the case that Category:Old maps of <x> may be a redirect as well; for instance: let Category:Old maps of <x> be a redirect to Category:Historical maps of <x> - then, with your fix in place, the categorization is done to the former only while it should be vice versa.
- I have not checked for a live sample underlining this, but it is easy to picture that these constellations may be instantiated in the future, even if that may not happen within the geographical scope of duisburg/dortmund/rheinland.. --93.201.169.189 01:22, 7 June 2024 (UTC)
- Please don't get stuck on this - it's fixed now for sure and there is other fields to work on / contribute. — Preceding unsigned comment was added by 93.201.169.189 (talk) 01:25, 7 June 2024 (UTC)
- It still needs to be sorted out. I don't think this categorization is ever valid. Commons has its definition for "old maps" and "historical maps" and these are part of the first group. It's possible that the later category exists and hasn't been redirected, but this doesn't mean the template should add both if they happen to exist. If the old map category was a redirect, it would have shown up in the maintenance category: Category:Non-empty category redirects and it didn't. (I do check). Special:Diff/881566476 seems the correct solution for this issue. Enhancing999 (talk) 08:34, 7 June 2024 (UTC)
- People don't ever understand? "Please don't get stuck on this." - of course you're doing the inverse. It's natural but unfortunately not very sensible. The first paragraph extends more on the general problem doing module changes on commons, the second is specific to the scope dealt with here.
- "I do check" only tells me that you do not have much experience with wiki projects all together. Such promises are independable on the whole - people come and go, at random, at their will. Something else, the university or whatever gets more important and "I do check" mutates to "I did check in 1984". There are unoverseeable amounts of reasons that lead to shifting priorities. For whatever reason you feel that this is important now: This setting won't last. It starts with your assumptions you base your arguments on: Commons has its definition for .. - this is complete bull, commons is a heterogenous place with lots of authors; sometimes a consensus is found; sometimes people actually listen in discussions; sometimes a new star with lots of fuel pops up and quickly burns; sometimes people claim Commons has its definition for .. but they act on I have my definition for .. principle. Pick any. I repeat: "Please don't get stuck on this." There is plenty more you can invest your time more usefully on. And you can start your own, new modules and templates in places that are sparsely structured yet.
- More closely on the topic:
- historical maps does not as closely relate to old maps as you suggest, the link is much weaker; for instance, there is no strict necessity that both of them need to exist when combined with a select topic/toponym. This may or may not be the case. However, the code you suggest turns this (forms it) into a necessity for the code to work correct. (You'd impose upon the category data structure, using the code, taking some freedom away from human categorists; e.g. if they feel historical maps is sufficient in combination with hypothetical area Entenhausen and they deliberately want to omit old maps, your code variant would either hinder them to do so, or they'd simply skip using the template, voiding the unification efforts and possibilities we offer those templates/modules for in the first place.)
- This is not fully true, the break variant is not so strict as to uphold this argument, but it does have unwanted implication nonetheless: If old maps of x is created a long time after historical maps of x, the object to be categorized wanders from the latter to the former, regardless of historical maps of x being changed to a redirect or continuously used as a cat. --84.135.119.180 02:53, 8 June 2024 (UTC)
- the category tree is not static (better: tree views generated from the category graph). for example, people may create a historical maps for place x first, then 3 years later decide to subcategorize that into old maps (not existing in the mean time). they may also move historical maps for to old maps for or vice versa, with or without leaving a redirect. there is no dictator watching over a creation order or particular neatness. in general, people organize their media, building upon the category graph in a self-sufficient and efficient way on keep-it-simple-stupid paradigm. If module/template code works against that, then this is a reason for modules or templates either not to be used, or, if using such has become common practice / a convention a reason for loosing interest to contribute (entry level too high, because creating all the categories expected by the module/template seems too cumbersome to a contributor; this means you absolutely must accept and account for intermediary, simpler states of the category tree as well, as when it has not "bloomed" yet in certain parts of the graph; this is often the case when the amount of media to be categorized is small - the keep-it-simple-stupid principle means that an intermediary, even imperfect state to organize these may be sufficient, even for a long time, before an 'ideal' structure is about to replace or amend such stubs)
- in different contexts historical maps and old maps may not be in a subset but rather in a sibling relation. for example it's easy to picture that some publications may produce historical maps in the here and now, i. e. maps that are researched and published somewhere near to now but depict a state centuries ago. in this case there is argument enough to not subcategorize historical maps to old maps, but you'd still want to have the Messtischblatt collection of maps to appear in both of them (!), because they are both, historical and old (not produced somewhere near now). In addition, and this is not nail picking: there are old maps that are not historical, either because they have passed the frontier to being outdated, but are not old enough (yet) to perceive them as historical, or because they are too trivial to not earn the predicate historical.
- In any case, I hope you see that the code should not make too many assumptions about the structure of the tree in all the places it may be employed. --84.135.119.180 02:39, 8 June 2024 (UTC)
- Some of the above is direct speak from experience - no pun intended: truely. It wears to iterate, though. --84.135.119.180 02:39, 8 June 2024 (UTC)
- As I have referenced KISS principle above, i should add as simple as possible, but not simpler paradigm as well, as neither one of both are holy grails. --84.135.119.180 03:31, 8 June 2024 (UTC)
- No worries. Whatever the principles, generally one would look at testcases to check if a change is appropriate in practice. I gave 4 for mine and added a minimum of documentation to Module:Messtischblatt.
- It's an interesting module btw. Maybe we could update the layout on files a bit and possibly output some description on category pages. A while back I did File:1230 Guttannen.jpg and File:Karte des Deutschen Reiches (5820656c).jpg (some others of these series). Enhancing999 (talk) 06:54, 8 June 2024 (UTC)
- Do you have a sample where this broke the former "correct" categorization"? Enhancing999 (talk) 06:23, 6 June 2024 (UTC)
- A check for redirects has now been implemented as you've suggested. --84.135.125.242 00:48, 6 June 2024 (UTC)
- You've 'fixed' generic code to deal with your special case only. There are categories where your findings above are not applicable and for these the former correct categorization is not done anymore. You suggest to check, if a category page is a redirect before categorizing to it, but this is not what you've done with your 'fix' in the code. --84.135.125.242 23:59, 5 June 2024 (UTC)