[12:38] <cjwatson> wgrant: builder processor sorting> do you mean on id or (contrary to the current UI) alphabetically by processor name?
[12:39] <wgrant> cjwatson: Anything that's deterministic, but name might be sensible.
[12:42] <cjwatson> The overall sort key should probably remain id, so that x86 builders land at the top.
[12:42] <wgrant> Indeed.
[12:42] <cjwatson> Although I guess that would still be the case since amd64 is lexicofirst
[12:42] <cjwatson> But might not always be, who knows.
[12:42] <wgrant> Lucky we don't have aarch64.
[12:47] <infinity> Aardvarch.
[12:47] <infinity> I guess we have no concept of sort-of-related processor families?
[12:48] <infinity> Cause it would be nice for amd64/i386 to sort together, arm64/armhf/armel, ppc64el/ppc64/powerpc
[12:48] <wgrant> Not at the moment.
[12:48] <wgrant> But in practise they will eventually be grouped.
[12:48] <infinity> Yeah, fair.
[12:48] <infinity> Once they're all grouped, I guess that'll sort (hah!) itself.
[12:49] <cjwatson> wgrant: Actually, this *is* the overall sort key.
[12:49] <cjwatson> So if we're doing this in that method it might as well just be sorted(p.name for p in builder.processors)
[12:50] <wgrant> cjwatson: Oh, is builder.processors sorted already?
[12:50] <cjwatson> By id
[12:51] <cjwatson>         return list(Store.of(self).find(
[12:51] <cjwatson>             Processor,
[12:51] <cjwatson>             BuilderProcessor.processor_id == Processor.id,
[12:51] <cjwatson>             BuilderProcessor.builder == self).order_by(Processor.id))
[12:56] <cjwatson> wgrant: Does that address your MP comment, then?  I thought you knew about that and wanted it explicit in the browser code too.
[12:56] <cjwatson> I could add a comment indicating the sorting.
[12:57] <wgrant> cjwatson: I think it's fine as it is. I hadn't realised the property was explicitly sorted already.
[12:58] <cjwatson> OK