Matthew Draper

Hmmm... I think this feels like a rather arbitrary flail, and a substantial violation of layering: the adapter is a utility class doing work for the server, not something that...

I don't think the linked test matches/describes quite this behaviour -- that is showing a conversion between stored and reported time zones, which shifts the result back a few hours...

> They changed the SQL statements here which suggests to me this is intentional: 6444ec0d307b4c0e599bb7b26dc29aa300097191 @paulreece I don't see how that change is related, let alone implying intention? > A...

> What about when it's not possible to change the name of the field on the client side, or you need to support field names with and without the []...

What does the input stream look like? Does it meet the HTTP specification? Is one version of Rack departing from specified behaviour, or are they both just making different guesses...

IMO actual merge conflicts just don't happen often enough to worry about. But if we really want this, we should just use @homu.

@durandom without actually checking the failures, I suspect the majority, at least, are failures due to changes in dependencies. Strictly speaking, insisting on running the tests immediately before the merge...

I meant what I said. :unamused: You can't merge a PR if it has _detected_ conflicts, due to physically overlapping changes. The theory behind a "re-run the tests with the...

Confirm. See the first green builds after the red ones in your report... I doubt you'll find many that are untangling a conflict between recent merges: I believe they'll either...

@chrisarcand no, that's the sort of thing @durandom's proposal _would_ prevent, that I'm arguing ~never actually happens