Commons:Categories for discussion/2018/01/Category:Maps with unidentified projection

This map found by me in the category shows a piece of France ≈ 150 km from north (Gien) to south (Petite Creuse; Latin: Creuse minor), and less than it from west to east. On this distance, how much will different projections deviate from each other? For e.g. gnomonic vs orthographic (which distort surface geometry in opposite ways), both centered in the middle of the territory, the difference is about 25 metres – enough said. Cylindrical projection—with vertical meridians—would deviate more significantly (by about 3 km), but anyway… does anybody expect this precision from such a map?

Determination of projection is a sensible problem for larger maps (at least of entire continental France); and even then, for ancient maps, one can’t distinguish between many kinds. It isn’t an argument against the category, but it must be supplied with suggestions:

• Which maps should not be added and why;
• For maps where a slight hope exists to learn something about their geometry, what namely can be said about properties of the projection?

Possibly, rectification of this problem requires creating new categories for maps where projection is not (and arguably cannot be) determined precisely. Incnis Mrsi (talk) 19:51, 8 January 2018 (UTC)[reply]

Interesting point. I think unidentified-type categories are mainly for things that could be described by some kind of classification system but we currently are not sure where exactly in that system they actually belong. In terms of maps, that would easily apply for modern maps that are known to follow some kind of standard projection, but which one exactly that is has not been identified yet (for example File:1-12 Grey Map World.png). For historic (really old) maps, that whole concept probably does not make much sense: If the map maker did not use any standard projection (because at that time there weren't any standards), there's nothing to "identify" (for example File:Cotton world map.jpg). Those maps should probably go into a second "unknown projections" Category. --El Grafo (talk) 13:46, 9 January 2018 (UTC)[reply]

Maybe, Category:Maps with indeterminable projection? But flooding it with thousands of local maps will be pointless, because it would make any scrutiny (searching for maps where some hope exists to determine anything) virtually impossible. Incnis Mrsi (talk) 15:18, 9 January 2018 (UTC)[reply]

Stale discussion. @Incnis Mrsi and El Grafo: seems cartography-specific topic. I started discussion in enwiki en:Talk:Map_projection#Classification_of_old_maps_without_projection?--Estopedist1 (talk) 10:41, 29 November 2021 (UTC)[reply]

I don’t know much about the categorization schemes under discussion. I do know about projections. I point out that a lot of maps older than two hundred years had so little accuracy over large areas that the question of projection has no practical relevance: those inaccuracies swamp the contribution of errors from the projection. If the old map’s scope was the entire world, then the question of projection can often be answered, but not always. The 16th century, in particular, spawned a lot of one-off projections. Strebe (talk) 02:14, 30 November 2021 (UTC)[reply]