[01:49:53] --- DominusExult is now known as Dominus
[15:55:03] <Dominus> Marzo, it fails to compile https://pastebin.com/r8TRWpyE
[15:56:46] <Marzo> Can you change this:
[15:56:47] <Marzo> MIDISend(mOutPort, mDest, packetList);
[15:56:49] <Marzo> to this:
[15:56:53] <Marzo> MIDISend(mOutPort, mDest, &packetList);
[15:56:57] <Marzo> and retry?
[16:14:55] <Dominus> that compiled
[16:15:26] --- ilkdntithgssekuh is now known as xveswilxxmrgkimk
[16:15:30] --- xveswilxxmrgkimk is now known as Marzo
[16:18:29] <azeem> Marzo: did you see that: 18:19 <@Dominus> that compiled
[16:18:40] <azeem> cause you bailed out pretty quickly afterwards
[16:18:51] <Marzo> Connectivity issue
[16:19:01] <Marzo> Strange, I did not see it in exultbot logs
[16:19:06] <Dominus> Azeem: thanks for taking care of us ;)
[16:19:30] <Marzo> Oh, it is there but I did not see it before
[16:19:32] <Marzo> Huh
[16:19:43] <Marzo> Anyway: does audio work fine with this?
[16:20:02] <Marzo> It should, but since I cannot test myself, I have to ask :)
[16:20:13] <Dominus> I need to test ios compiling and then actually if it works but kids need evening attention now. Sorry, will take a bit
[16:20:26] <Marzo> No worries
[16:20:45] <Marzo> I can just leave the changes stashed locally until you have time to test
[16:46:24] <Dominus> hmm, I need to see how I route my mt32 again, it's playing something but I don't hear it.
[16:46:32] <Dominus> BUT Marzo, it crashes with your changes
[16:46:46] <Dominus> Assertion failed: (sizeof(packet.data) >= length + 2), function send_sysex, file CoreMidiDriver.cpp, line 162.
[16:47:09] <Marzo> Hm
[16:47:26] <Marzo> So the added space from the old buffer was neccessary after all
[16:48:02] <Marzo> Which is bad, because it overwrites parts of the structure that are used for other purposes
[16:48:22] <Marzo> And is only working because Apple's code is not checking things it should be
[16:49:39] <Marzo> I will have to rething that
[16:54:42] <Dominus> (on the bright side I figured out again how to playback my mt32 music)
[16:58:19] <Dominus> Marzo: maybe someone else thought of how to do it properly already. E.g. ScummVM (where this code was lifted from)
[16:58:57] <Marzo> I found some decent documentation now
[16:59:19] <Marzo> As it turns out, I had the order of fields wrong, so my analysis above is a bit spotty
[17:00:49] <Marzo> Try this patch instead of the old one: https://pastebin.com/MTe5drE3
[17:02:08] <Dominus> (Again kids need attention ;))
[17:02:16] <Marzo> :)
[17:32:13] <Dominus> Marzo: compile error https://pastebin.com/zeD4qRkb
[17:34:17] <Marzo> Ack, it should have been aligned_storage_t instead of aligned_storage
[17:34:43] <Marzo> Why you should never program while tired
[17:44:27] <Dominus> I changed std::aligned_storage to std::aligned_storage_t but the error remains the same
[17:45:04] <Dominus> we can do this another time :)
[17:48:05] <Marzo> Yeah, another time
[17:49:13] <Dominus> (I'm very tired myself today, for a walk I wanted to change from one jeans to another, took one off, went to use the toilet, put the same jeans on again and only noticed an hour later...)
[20:23:46] <Marzo> @Dominus: for what is worth, the "MIDIPacketList *packetList" line needs to be this: "MIDIPacketList *packetList = new (&buf) MIDIPacketList;" (note the & before "buf")
[20:25:43] <Marzo> But yeah, it also needs to be aligned_storage_t
[20:26:40] <Marzo> I am thinking of maybe setting up a false OSX dev environment using some headers from here: https://github.com/phracker/MacOSX-SDKs
[20:27:06] <Marzo> At least good enough that I can use clangd on my editor, so it can catch silly errors
[20:27:45] <Marzo> Would you recommend any specific version of the SDK?