I think it's probably reasonable for us to look for a valid address anywhere in the recipient list. CC'ing bugs@ makes a lot of sense to me, in particular.

Is this the actual use case (CC'ing a revision)? Specifically, not CC'ing something like bugs@phabricator.company.com? Is there a reason the revision address is in the CC line instead of the "To" line? Presumably, the author had to move it there explicitly, which seems a little unusual.

We could possibly do something like:

We are very unlikely to pursue this change, specifically. Although the referenced library happens to define head() and last() in the same way, there's no guarantee that existing head() and last() functions are compatible.

We should probably also set this prior to accepting pushes via HTTP and SSH, so those writes happen under a favorable umask.

We could also implement a very sketchy version of this by maintaining an SVN working copy on disk and running normal grep against it. This is terrible, but maybe only moderately terrible.

Oh, and that also wouldn't solve the problem that central auth was aimed at solving, which is:

In theory, we could:

• Primarily, their username may collide with a Phacility username, and this seems hard to resolve in a way that makes any sense to users.
• We don't want them to lose access to their Phacility account if they leave the company.
• We don't have any account synchronization code in the upstream, and don't really want to add it since it would only serve Phacility use cases.

I'm just going to close this since there's nothing specifically actionable here and T2009 covers the same ground in a more measurable way.

Working fine for me with or without a mouse in Chrome + Safari...

Oh, let me try without a mouse.

Oh, looks like you did get the scrollbar working (at least, works for me).

One thing we could do is confirm the commit after we create or amend it, so you'd get this:

When you answer affirmatively to this prompt:

I'm having trouble following your reproduction instructions. I hit various errors around step (2) and I'm not sure exactly what you're doing. Can you give me a sequence of commands, or a transcript which reproduces the issue where you add --trace to each command you run?

I'm going to push all that stuff into HEAD. Unified diffs are still significantly buggy, but I believe they've come far enough that the mobile experience is in a better place overall, and anticipate fixing most of the remaining issues shortly.

That is, the git diff output is this:

If you do this:

but but but

Yeah, that repro'd things:

Oh, I think this is related to the magic "fill in missing context" stuff we do, since downloading and then re-creating the diff doesn't reproduce the issue:

Yeah, that'd be perfect.

Haha -- what did you mean for the "=" menu to do? @btrahan and I both assumed that was widgets.

(Unless you managed to fix that.)

I'll see if I can figure out the scrollbar thing once I get to a good breakpoint on T2009.

Do you have a way to reproduce this? I'm not totally sure that D11970 fixes this, it just fixes every case I was able to identify.

Using a mouse wheel on Yosemite with Chrome 41, I can't reproduce.

Definitely desirable, although some of them (releeph, oauth server, arcanist project edit) might be tricky to test.

I think we can put the frame logic in buildApplicationPage(), and then switch to it. I can tackle that if you aren't sure how to not break stuff.

buildApplicationPage() calls setDeviceReady(true) on the $page, buildStandardPage() does not.

Removing $is_framed will break the DarkConsole "XHProf" tab, which loads the profile in an <iframe />.

In the general case, there are approximately one hundred trillion PHP files out there which do this:

You might be able to adjust phpmailer.smtp-encoding to fix this, but I'm not sure if that will work.

(And I think it's pretty hard to guess lengths.)

(I probably load the page like 35 million times a day, and on 50K of those the times happen to line up enough that it feels a little off.)

Mine has a placeholder, the loading times just are out of sync often enough that I see a bunch of commits and then tasks don't pop in for a bit. The panel that pops in is like 800px tall or something, and the loading indicator is like 100px, so if I'd started reading anything in the panel below it jumps 700px away.

Some discussion of the "other repositories" issue in T6516.

T6030 might be one way forward here, since I think there are (well, maybe) reasonable use cases for wanting a full URL or for wanting only a reference.