ooooor this, feel free to do whatever bro

The patch is already approved but maybe it's nice to bring more attention to extension developers

I'd love to have Phorge skip common words and short words though.
So maybe keeping these functions could even make sense so future improvements could call them?

Hi, thanks a lot for sharing this! I assume you are running 2024.35 and not yet 2025.18.
You shared seven issues above.

• The first issue is unresolved T15824.
• The second issue has not been reported yet about https://we.phorge.it/source/phorge/browse/master/src/applications/oauthserver/controller/PhabricatorOAuthServerTokenController.php$26, I may write/propose a patch later once finishing some other work.
• The last five issues were fixed by rP372316c998362e3530c0e591fe0869e6a7e62af6.

Fingers crossed.

In D26000#27480, @valerio.bozzolan wrote:

Is getDisplayIconColor() related to these? (spoiler: probably it's not)

remove unused CSS as it's also unhelpful CSS when actually used

These selectors currently do not get applied at all in Conpherence rooms when opening or creating URIs with anchors.

git rebase master

Not really what we expected, but nice try

Not really what we expected, but nice try

@param loves $name

Can't wait to enjoy the production Grafana CPU graph of Wikimedia Phabricator after this change

In T15903#18783, @avivey wrote:

pht() also does some escaping, formatting, and other safety related stuff

In D26032#27637, @valerio.bozzolan wrote:

Probably one screenshot is missing (?)

In D26032#27621, @aklapper wrote:

Feed before applying the patch (T3922) and after applying the patch (T3923):

add more and more PHPDoc, and fix typos - thanks

If I guess correctly, isNewObject is automatically guessed correctly, so, the root problem is probably here:

Can you discover what's happening here? In theory the editor should detect isNewObject.

Some small nitpicks only. Makes sense to me, didn't spot any logic flaws, keeps close to already existing code in that very directory

Feed before applying the patch (T3922) and after applying the patch (T3923):

getIsCreateTransaction() should return true here but it doesn't.
I believe that this underlying problem can also trigger an exception, see https://phabricator.wikimedia.org/T361459

git rebase master

Agreed.

Now you do.

git rebase master

no no I rock like a grapstar

You grep like a rockstar

Uh right, thanks! Let's do as mainframe98 says

Unlocking patch

In D26007#26996, @aklapper wrote:

@valerio.bozzolan core/remarkup.css has five more unused selectors so I'd put them all together into one patch.

You're right, thank you (as usual)!

git rebase master

Flagging as "andre is merging stuff https://we.phorge.it/D26007#26996"