[10:13] <LocutusOfBorg> Laney, FYI ghc 8.6.5 no change rebuild in focal was fast, and 8.8.4 rebuilt with that 8.6.5 was already slower but not that much (took 2 days)
[10:14] <LocutusOfBorg> so my wild guess is that 8.6.5 is really faster and 8.8.x became slower
[10:14] <LocutusOfBorg> if I no-change reupload another 8.8.1 in my ppa to be built on top of 8.8.4 it will probably take weeks
[11:55] <ogra> waveform, i have a hard time trying to get one of my UC16 Pi3's (which i cant easily physically reach to put UC18 on) to work with BT ... it only gives me "Controller AA:AA:AA:AA:AA:AA Pairable: yes" ... any hint what i could try to get this device to work without having to replace the OS ?
[11:57] <Laney> LocutusOfBorg: what is fast?
[12:08] <waveform> ogra, I vaguely recall that's from the hciattach command failing to specify the MAC address for the adapter - just a mo
[12:09] <ogra> i got pi-bluetooth installed (and tried various versions of bluez) ... feels like UC16 is missing something (firmware ?? kernel bits ???)
[12:10] <ogra> the bootloader firmware seems to be from sept. 2017 (its an older install)
[12:16] <waveform> wouldn't surprise me - I don't think uc16 has been updated in a long time
[12:19] <ogra> hrm 😕
[12:23] <ogra> aha
[12:23] <ogra> Sep 18 14:22:43 stream kernel: bluetooth hci0: Direct firmware load for brcm/BCM.hcd failed with error -2
[12:23] <ogra> Sep 18 14:22:43 stream kernel: Bluetooth: hci0: BCM: Patch brcm/BCM.hcd not found
[12:24] <ogra> waveform, ^^ looks like we're mssing firmware in the pi2-kernel snap
[12:24] <ogra> *missing
[12:24] <waveform> yes, that sounds relevant
[12:24] <ogra> yep ... and the file is indeed not there
[12:25] <waveform> you'll find the relevant .hcd files in the linux-firmware-raspi2 package
[12:25] <ogra> well, this is core
[12:26] <ogra> you cant modify the /lib/firmware content easily
[13:25] <LocutusOfBorg> Laney, fast is building 8.8.4 with 8.6.5 in focal
[13:26] <LocutusOfBorg> now I'm building in focal 8.8.4 with 8.8.4, to see if the slow stuff is something else in groovy (e.g. libc/gcc/binutils) or gcc itself
[13:29] <Laney> LocutusOfBorg: I mean how long, like a few hours?
[13:29] <Laney> and s/or gcc/or ghc/ in that last line I assume?
[14:07] <LocutusOfBorg> sure
[14:07] <LocutusOfBorg> Laney, the sed is correct
[14:07]  * Laney is a sed wizard
[14:07] <LocutusOfBorg> well, https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+build/19929163
[14:08] <LocutusOfBorg> the no change rebuild of 8.6.5 too 19 hours
[14:08] <LocutusOfBorg> and the one in the archive for focal took 14 h
[14:08] <LocutusOfBorg> the 8.8.4 took instead 2 days 2 h
[14:08] <LocutusOfBorg> everything on focal
[14:08] <LocutusOfBorg> https://launchpad.net/~costamagnagianfranco/+archive/ubuntu/locutusofborg-ppa/+build/19935040
[14:19] <Laney> but debian doesn't see this, right?
[14:19] <Laney> I'm not sure what we're learning :\
[15:32] <LocutusOfBorg> Laney, I'm not sure too, sorry
[15:32] <LocutusOfBorg> I want to understand how long will take the no change rebuild against itself (8.8.4->8.8.4)
[15:39] <seb128> LocutusOfBorg, hey, do you intend to upload the gstreamer packages this week?
[15:59] <bdmurray> Adding something to "big_packages" in autopkgtest cloud also means they end up as a part of "long" i.e. have a greater timeout. Is that correct?
[16:17] <Laney> bdmurray: https://git.launchpad.net/autopkgtest-cloud/tree/worker/worker#n485
[16:19] <bdmurray> Laney: but the default timeout-test is 10000 correct?
[16:19] <bdmurray> Laney: If so it'd nice to add '--timeout--test=10000' in the last else so that it appears in the test logs.
[16:21] <Laney> If we think that it's appropriate to potentially diverge from a changed default, sure
[16:53] <LocutusOfBorg> seb128, they still meson-fail to configure
[16:53] <LocutusOfBorg> see my ppa locutusofborg-ppa, help is appreciated