Yeah, totally reasonable feature request.
Do you think you can implement?
This might involve the Ferret engine (like this thing, or maybe there's a simpler approach for the query (title LIKE %text% in the Query class?).

IDK - this was code was never explicitly about git am - I think it was about patch, which started in the stone age and predates things like "formal specification".

I think phutil_nonempty_string fails faster then strlen (strlen allowing automatic conversion of string-like things to string and phutil_nonempty_string doesn't), but I'm a little confused right now...

phutil_nonempty_string is basically the exact same thing, but with stricter type control

I don't understand why the pht was there - it's a "path", why would it be translated?

One day, we'd like pht() to return an object (non-string), but we'll handle that when we get there I guess.

In D25382#11307, @bob wrote:

Do you mean ?
...

how about something like:

thank you @ncoker, but:

1. This is not a task to discuss specific errors when trying to run on php 8, and
2. We don't expect arcanist to work on php 8 yet

You could also try deleting the files from db using bin/remove, which will at least remove them from limbo... assuming you know which files are missing their data.

D25386 and D25387 provide the big "do not crash on deprecation warning", and are probably enough to declare "experimental support for php 8.2" on their own.

lint + a typofix

There are, in total, 3 call-sites to setErrorListener - in for the daemons, one for dark-console, and one (PhutilErrorLog) that lives in arcanist but is only used by the SSHD scripts.

I've replaced it with DifferentialRevertPlanCommitMessageField, anyway - that one already has all the required stuff.

Unfortunately, that fix adds the word "Test:" to the title of any new diff. I'm trying to find an existing field that can make the test work, or else find another creative solution.

The test-field gets picked up automatically because we pick all fields. It's missing FIELDKEY, so arc-diff doesn't work at all, and if we add FIELDKEY it shows up as the first field in the arc-diff form.

tl;dr: I like phutil_nonempty_string($map['query']) better if we know $map['query'] is defined, and phutil_nonempty_string(idx($map, 'query')) otherwise.

looks like arc patch tries to patch D25358 first, because this diff "Depends on" that one - but that one is in arcanist, not phorge.

In D25346#10708, @aklapper wrote:

In D25346#10087, @avivey wrote:

I think this should be implemented using the getHead() method that's already defined.

Could you elaborate a bit, please? I'm not sure what exactly not to do here, and what to do instead. Thanks in advance!

Implicitly, Phorge generally works fine for browsers where Javascript is completely disabled.

4. Why these nice URLs are set in these links, if nobody visit them?

If this is a feature request, then add all the things needed for a feature request - as in, "why would this feature be needed" and "what problem you're trying to solve".

Why would it need to be a that?
Is this a bug or a feature request?

LGTM, except for the comment text.

In T15540#11723, @valerio.bozzolan wrote:

Partially related? Sometime I put my pointer hover the "Authored By <foo>" or similar fields, but they have no Hovercards. Probably easy-peasy?

Small issue in the query, but otherwise looks good.

Created repo rQR QR Stickers and project QR Codes Extension

Created repo rQR QR Stickers and project QR Codes Extension

$edit_mode has a true value for 'create' or 'edit', and false (null) for read. The strlen() is superfluous.

maybe add that as a comment in there...

I thought we had "2 spaces" and the rest encoded as lints... I'll try to take a look.

I was thinking of adding a script to ./bin to manage (mostly add/remove from load-libraries, maybe also clone) extensions.

I was thinking "out out", but only visible to admins - and then, only as a yellow notice at the top of the page: