Do phabricator ghosts dream of mustard trucks

more of a mustard, really.

The last exception in daemons.log:

06-May-2016 20:36:27 America/New_York] [2016-05-06 20:36:27] EXCEPTION: (FilesystemException) Failed to remove file '/var/local/phacility/phabricator/filestore/1a/3e/94c5ea8fadb9f60811517d0eaca9'! at [<phutil>/src/filesystem/Filesystem.php:325]

OK so I've been keeping an on my icons this week to see when it would happen. The ghosts indeed came back and I grabbed logs during the shortest timeframe I could narrow it down to - roughly between 1pm and 8pm today ("narrow", sorry).

A very specific example of my confusion is the decision to place the link to the revision first in the mockup on T10283, and how it contrasts with the stated problem of triage efficiency:

(I think this is broadly similar to the original discussion on T10283, which also framed problems in terms of efficiency but also appeared to enable every inefficient/low-signal option available.)

As a thought experiment, if we sent one email per inline comment like GitHub does (and just repeated the author information several times in the body or something), I think that would technically satisfy all the points you raise:

This feedback is also sort of hard for me to understand.

The "epriestley added a comment" and/or "epriestly added inline comments" while technically having the appropriate information, is pretty much visual noise that as a reader of a huge number of these emails you just sift over and want to jump straight to the meat of the matter. These changes will make said meat of the matter a lot more prominent and trivial to jump to, but it means you want the identity to be *right there* as well.

Is metamta.can-send-as-user not configured in your environment? My expectation is that the mail comes "From" the user writing all the text in the mail, which makes it clear who they are. I expect that this information is normally prominent in the UI of most mail clients.

Overall it looks a *lot* better now :)

That actually works! Thanks :)

Haha, alright. You might be able to add it to conf/local/local.json manually as a workaround, but I'm not sure if that'll work or not.

It involves dirty solutions to idiosyncratic problems, maybe we're both better off not talking about it.

After that action phabricator redirects to the main repo page. Is that expected?

AphrontMySQLDatabaseConnectionBase was renamed but I plan to remove this option anyway.

Thanks for that quick fix. :)

To get @kornrunner’s suggestion to work on Ubuntu 16.04 with PHP7:

D15865 fixed some issues, but if you go to manage repo, and click reload there is still a problem: After that action phabricator redirects to the main repo page. Is that expected? The disadvantage is, that this causes long webpage load, if you have a normal repo. Is there a solution for that?

We iterated on the design a bit and ended up here:

Ahh! It's right there on the bloody page. Sorry.

@jcarrillo7, something in that vein will split out of T1460 eventually but we don't have a specific plan for it yet. I don't intend to address it as part of this task.

Thanks! Maybe this can save some time for some people...

I already thought that while writing the ticket ;-) That's the reason i added the last sentence to adjust the documentation. Maybe it would be good to know that this feature is using the git mirror function and may overwrite data of the mirror.

This is expected, and consistent with git's own definition of "mirror". The exact command we use is:

@epriestley This is kind of random but this looks like it will look really cool in email. Is there any chance of bringing these snippets into the discussion view on a review instead of just the links to the inline comments? Kind of like how github does this?

What information specifically in the sidebar should be read first?

I've been hoping for the actions and information to swap places. It seems like the information is what should be read first, before getting to the actions...

Inline comments are fixed after D15850.

I think I fixed this in D15847:

D15856 appears to have fixed this

Ok, thanks.

As a work-around, you can set the repo to clone from the GitHub mirror:
git remote set-url origin https://github.com/phacility/arcanist.git (and libphutil for libphutil).

I filed T10927.

Looks like anonymous clone over https stopped working? T10927

(not opposed to reordering the sidebar, either)

This isn't the most thrilling UI but seems to get the job done. T10912 covers one minor followup issue.

This is getting somewhere functionally; we're going to iterate on the design a bit more internally.

Spacing/padding is definitely pretty iffy in Mail.app, at least.

My early-morning logic went something like "it's still there on disk so...".

I nuked the new controller in that diff

Actions - will get reduced in size and sidebar content will move up as a result. We're heavily prioritized right now so no current ETA. It is our intention to reduce these globally. This will raise content underneath significantly, we expect to remove about 1/3 to 1/2 the actions globally if possible.

I'm happy with the new UI but only because I happen to work with a monitor in portrait mode. :)

That seems to be doing something fairly reasonable in my client, yell if it looks funky somewhere else:

awesome, thanks for the report!