Huntster

starfighter

It got changed in a big job because it was a "fighter", which is an aircraft type. I'll remove that Vicarage (talk) 18:09, 17 January 2024 (UTC)

@Vicarage: Again, no worries, it happens. — Huntster (t @ c) 18:40, 17 January 2024 (UTC)

Reordering items

I was thinking for some time whenever to start automatically reordering items based on ranks (e.g. preferred->normal->deprecated), but you seem to have different idea. Could you please explain your reasoning? Ghuron (talk) 04:56, 12 June 2024 (UTC)

@Ghuron: Generally based on the date of publication, unless there's some abstract overriding reason in a particular case. — Huntster (t @ c) 04:59, 12 June 2024 (UTC)

Year, but mass (P2067) statements use the same article as a reference (but has different values, probably due to recalculation from M⊙ to Mj) Ghuron (talk) 05:31, 12 June 2024 (UTC)

@Ghuron: Oh! I'm sorry, I misinterpreted what was being asked. I tend to place "good" values at the bottom because they are typically later values, as again I try to order chronologically. In that example's particular case, it was simply a habitual change based on that. — Huntster (t @ c) 05:39, 12 June 2024 (UTC)

This sounds a little counter-intuitive. Don't we want the best (or latest) measurements ​​to be at the top and available at a glance? Ghuron (talk) 05:43, 12 June 2024 (UTC)

@Ghuron: It might seem that way on the surface, but it aids in future simplicity for non-bot operators. If values get ordered chronologically now, then it rarely needs to be addressed in the future as new publications would necessarily get added to the end of the pile. It also visually shows the progression of research as methods improve. — Huntster (t @ c) 05:48, 12 June 2024 (UTC)

I think this only applies to quantitative values, right? Ghuron (talk) 05:53, 12 June 2024 (UTC)

@Ghuron: Correct. — Huntster (t @ c) 06:07, 12 June 2024 (UTC)

rocket series

An instance of rocket series (Q111722634) is a series of rocket models, so an instance of an instance of rocket series (Q111722634) is a rocket model, which is not a physical object. An instance of group of artificial physical objects (Q66661745) is a group of physical objects, so an instance of an instance of group of artificial physical objects (Q66661745) is a physical object. Thus the two classes should not be in a subclass of (P279) relationships. I have reverted your change. Peter F. Patel-Schneider (talk) 10:06, 4 August 2024 (UTC)

has parts vs vessel class

For naval ships vessel class (P289) is preferred to has part(s) (P527) for that class (by a ratio of 14 to 1). I am only removing the latter when the ships concerned have the former. Vicarage (talk) 10:35, 28 December 2024 (UTC)

@Vicarage: So the answer is to erase the reciprocal relationship between ship and class? Until and unless a property is created to serve that function (ships in class, perhaps?), I would suggest leaving those associations in place. — Huntster (t @ c) 11:36, 28 December 2024 (UTC)

Reciprocal arrangements are deprecated on WD when a single property can serve both roles in a SPARQL triplet. I'd oppose any proposal for a ships_in_class property, and don't plan to revert my block of changes. Vicarage (talk) 11:45, 28 December 2024 (UTC)

@Vicarage: I see. And where it that written? — Huntster (t @ c) 18:49, 28 December 2024 (UTC)