00:33 <Skipp_OSX> hey x512 can you test my menu patch pls

00:34 <x512[m]> Link to patch?

00:38 <Skipp_OSX> https://review.haiku-os.org/c/haiku/+/10946 <= you'll need the whole stack but especially obscured

00:39 <Skipp_OSX> you can test the feature if you wish but all I really need to know is that it doesn't crash horribly or not activate menu items when it should

00:42 <Skipp_OSX> To test feature you must know what is wrong in current build. If you don't, keyboard navigation works as long as mouse cursor is outside the menu, but not when it is inside the menu. You should see that keyboard navigation still works even when your mouse is over the menu even a supermenu with menu items. The mouse is entirely ignored when it is obscured.

00:42 <x512[m]> IsCursorObscured is immediate NAK. It is fundamentally wrong at least because of its asynchronous nature (cursor obscured state may change anytime after being queried, so queried value means nothing).

00:44 <Skipp_OSX> well test anyway

00:44 <x512[m]> As I said before, you should always call obscure cursor without attempting checks that cursor is obscured.

00:45 <Skipp_OSX> so obscure cursor in app server doesn't say what it says it does in the Be Book?

00:46 <Skipp_OSX> every time you key down a nav key it re-obscures the cursor already.

00:46 <x512[m]> > ObscureCursor() hides the cursor until the user moves the mouse.

00:46 <Skipp_OSX> right

00:46 <Skipp_OSX> so if the user moves the mouse the cursor is no longer obscured which is exactly what I want.

00:46 <x512[m]> This is how Haiku behaves, I see no problem. Calling ObscureCursor() if cursor is already obscured is fine.

00:47 <Skipp_OSX> I see no reason not to expose the obscure state from app_server, it works.

00:47 <x512[m]> There are absolutely no need of IsCursorObscured.

00:47 <Skipp_OSX> The menu feature depends on it.

00:48 <x512[m]> No, it doesn't. Just always call ObscureCursor().

00:48 <Skipp_OSX> I do

00:48 <x512[m]> So why use IsCursorObscured()?

00:48 <Skipp_OSX> so that the mouse will take us out of obscured and I track menu differently.

00:48 <Skipp_OSX> the mouse way

00:49 <Skipp_OSX> we're in a different menu tracking thread here

00:51 <x512[m]> I am not even see obscuring cursor in menu as good idea. Is it done be BeOS? Other OSes seems not do that.

00:51 <x512[m]> > _QuitTracking(be_app->IsCursorObscured());

00:51 <x512[m]> Why do this?

00:52 <x512[m]> ESC usually close current menu, not all menus in all cases.

00:57 <x512[m]> There are no separate "keyboard navigation mode". User can manipulate menu by both mouse and keyboard and switch between them anytime.

01:12 <Skipp_OSX> it is explained in the commit msg, to close all menus in mouse mode vs. the last menu in key nav mode

01:20 <Skipp_OSX> pls test patch

01:54 <x512[m]> Heh, I found menu bug in BeOS while exploring keyboard navigation behavior.

01:56 <x512[m]> I got menu frozen and not responding during exploring keyboard and mouse interaction.

02:42 <Skipp_OSX> https://www.haiku-os.org/legacy-docs/bebook/BApplication.html#BApplication_ObscureCursor

02:42 <Skipp_OSX> ok so you did test on BeOS haha

02:43 <Skipp_OSX> as I said it is badly broken and not to go off of, they should have made obscure cursor public back then and did it my way.

03:27 <x512[m]> "did it my way", "vision" is wrong approach and quick way to trigger controversy an d change rejection. Just fix bugs without altering behavior any you will be loved by everyone.

03:29 <x512[m]> I see meaning on differencing Esc behavior depending on obscured cursor. There are separate shortcut for closing whole menu chain: Alt+Esc.

03:29 <x512[m]> meaning -> no meaning

04:16 <Skipp_OSX> I don't want to be loved by everyone I want the menu to work right

04:29 <AlienSoldier> i want to be loved by the menu

06:16 <Begasus[m]> g'morning peeps

06:28 <nekobot> [haikuports] Begasus pushed 1 commit to branch master: https://github.com/haikuports/haikuports/compare/9af348fd6ae5...ccbbe643a0cc

06:28 <nekobot> • Begasus (ccbbe643): ktextaddons_kf6, add KF6 version (#14112)

06:45 <phschafft> morning.

06:47 <phschafft> 'Just fix bugs without altering behavior' sounds like an oxymoron to me.

07:03 <Begasus[m]> moin [phschafft](https://matrix.to/#/@_oftc_phschafft:matrix.org)

07:03 <phschafft> all good?

07:07 <Begasus[m]> yeah, at the dog school, so afk in between :)

07:07 <Begasus[m]> bumping KF6 locally

08:14 <nekobot> [haikuports] Begasus pushed 1 commit to branch master: https://github.com/haikuports/haikuports/compare/ccbbe643a0cc...7963f591bb36

08:14 <nekobot> • Begasus (7963f591): dav1d, bump to version 1.5.3 (#14109)

08:17 <x512[m]> By fixing bugs without altering behavior I mean fix bugs in order to make intended behavior to work instead of crash or do something else.

08:18 <x512[m]> But do not introduce not features or alter intended behavior.

08:18 <x512[m]> not features -> new features

08:20 <x512[m]> Splitting bug fixes and new features/altering behavior to different commits will make everyone happy and reduce controversy.

09:17 <nekobot> [haikuports] Begasus pushed 1 commit to branch master: https://github.com/haikuports/haikuports/compare/7963f591bb36...8372e1673c89

09:17 <nekobot> • Begasus (8372e167): libavif1.0, bump to version 1.4.1 (#14110)

09:52 <nilFinx> gm Begasus[m] and phschafft!

09:53 <phschafft> :)

09:53 <phschafft> the difference between 'gm' and 'gn', very subtile!

10:24 <nekobot> [haiku/haiku] aplgithub pushed 1 commit to master [hrev59701] - https://git.haiku-os.org/haiku.git/log/?qt=range&q=bd2396297e9a+%5Ed37ebc55c001

10:24 <nekobot> [haiku/haiku] bd2396297e9a - HaikuDepot: package selection in model

11:05 <nekobot> [haikuports] Begasus pushed 1 commit to branch master: https://github.com/haikuports/haikuports/compare/8372e1673c89...e40b5f7652b0

11:05 <nekobot> • Begasus (e40b5f76): libheif, bump to version 1.21.2 (#14111)

11:06 <Begasus[m]> packing gear, bbl

11:06 <phschafft> :)

11:13 <nekobot> [haikuports] jmairboeck pushed 1 commit to branch master: https://github.com/haikuports/haikuports/compare/e40b5f7652b0...a8bf76626c0e

11:13 <nekobot> • jmairboeck (a8bf7662): guile: bump to version 3.0.11