[[monddood]]
Personally I don't think MewBot should be putting the pronunciation above the etymology in single-etymology entries. This is contrary to the de-facto consensus that's emerged. If you don't like it, we should discuss it instead of unilaterally making a change like this, because it produces inconsistency in the Wiktionary structure.
It's more inconsistent when pronunciation is after etymology if there is one etymology, but before it if there are multiple etymologies. It's also annoying as hell when I have to swap them around each time I add a second etymology section.
(Wouldn't a proper database back-end be nice -- enter the Pronunciation data in a Pronunciation field, and let a layout engine worry about where to put it. Ah, well.)
(... actually, does anyone know if Semantic Wiki or any similar extension might allow just that?)
I completely agree. Although I wonder if that wouldn't make contributions even harder -- currently someone can just hit Edit and make a simple change like fix a typo, and see the whole structure. A database backend would have to be quite complex to handle all the variations that are currently encoded in the textual interface.
The point here is that consensus exists for a reason, which is to keep the dictionary broadly consistent. You can't (or shouldn't) just arbitrarily change things like this, even if it seems better.
I'm not changing anything, I'm making it consistent. By having the pronunciation in the same position whether there is one etymology section or many. To do it any other way would be inconsistent.
You're making the new entries inconsistent with the millions of existing entries that do it the other way.
BTW if there are multiple etymologies with different pronunciations, the pronunciation has to go under the etymology in any case, so it's not like you can always have the pronunciation first.
I already mentioned that above. But there's a big difference: nesting. In the case here, there is just one pronunciation for the whole entry, and I am arguing that it should be located above the etymology in all cases where it applies to the whole entry, just like we already do with Alternative forms.
FWIW, what CodeCat describes here is what we do for certain Japanese entries, such as せんし (senshi, with multiple pronunciations) or せいたん (seitan, with just one pronunciation for all terms with this reading). I know there are examples too where we have a top-level Pronunciation that applies to multiple terms with distinct and explicit Etymology sections (as opposed to implicit etymologies for the two examples here), but I cannot think of any at the moment.