Is there any consensus on when and when not to add full etymologies? For instance, I think it makes sense to add the full etymologies to Frankish entries, back to PIE, that contain a PGm. root with no entry page, but if they do have a page, I'm inclined to not duplicate the work, and keep that info on there.
I think it's standard practice to include the entire "chain" of inheritance back to PIE, with all intermediate steps (although Dutch entries normally have Old Dutch rather than Frankish). As for cognates, the consensus seems to be to include sister languages that were spoken around the same time. So an Old English entry would not have a modern German cognate, or vice versa.
Okay, I'll just keep to what I've been doing than. Wouldn't it be nice if we had some auto-mechanism for this?
(Late to the game, but there you have it...)
I dimly recall some time back discussing the possibility of selectively transcluding etymologies, since these all build on one another. Japanese スプーン (supūn) comes from English spoon comes from Middle English comes from Old English comes from Proto-Germanic etc. etc. It'd be handy if we could create some sort of chaining mechanism, but ultimately, I don't think the Wiki software is robust enough for that, and our template "coding" language, as it were, frankly sucks donkey balls (apologies for the rudeness). If/when we ever get Lua, that might change, but I've long since stopped holding my breath. :-| -- Eiríkr Útlendi │ Tala við mig 06:55, 12 February 2013 (UTC)
My donkey balls are offended by your rudeness. Actually, it's quite possible with our current infrastructure (i.e. transcluding subpages), just not worth the effort.
I think I'd be worth the effort in the time it saves afterwards.