#devuan-dev

04:13:10 <gnu_srs4> fsmithred: How long delay did you need to fix the eudev problems in #483?

04:14:38 <gnu_srs4> Problem with eudev and earlier versions of udev+sysvinit is that we don't have --notify-await option in start-stop-daemon.

04:30:53 <Evilham> fsmithred: someone rec'd a walkthrough of refracta installer: https://www.youtube.com/watch?v=FqrIgyZhtUk

04:36:39 <LeePen> Just finished the migration of git.d.o/devuan-packages to gitea/devuan.

04:37:36 <LeePen> I have tried to weed cruft. I will present the list at the meet tomorrow, but if you think I have missed something that should be migrated, do say.

04:41:58 <LeePen> GitLab repositories which have been migrated have gitea url added to description and are archived, so read-only.

05:34:31 <onefang> I'm not understanding how to set the gitlab permissions on my old projects. rrq suggested "check out devuan-packages/dbus as example", it says public, there is no source code or anything else. I can't get mine to do that just by messing with the permissions.

05:38:27 <bgstack15> Gitlab preserved issues, presentation attempt #2 is ready now! https://beta.devuan.org/gitlab-issues/

05:40:11 <onefang> The gitea version of my repos is done now at least.

05:42:22 <onefang> And all the links to the old gitlab repos in the code are now also updated. That helped to test pushing to the new repos. B-)

05:49:45 <LeePen> onefang: I have just discovered the GitLab archive setting. I am now setting that on all the devuan-packages/* that are migrated.

05:50:03 <LeePen> Just updated https://git.devuan.org/devuan-packages/dbus to that if you want to look.

05:50:17 <onefang> I was looking at that, but it wasn't what rrq told me to do, so I didn't.

05:51:07 <LeePen> I realise that. I only recently found the archive option and suggested that to rrq as a better approach.

05:52:52 <onefang> https://gitlab.devuan.org/onefang/apt-panopticon and https://gitlab.devuan.org/onefang/apt-panopticon_cgp are now migrated and archived.

05:53:37 * onefang logs out of gitlab.d.o for the last time, unless we need to do something else to the old repos.

05:56:47 <onefang> Tomorrow will be another red letter day, if my new headset solves the "onefang's voice is too noisy in jitsi" problem.

05:57:17 * onefang wonders if jitsi has an echo service for testing that sort of thing?

06:13:21 <gnu_srs4> LeePen: If you want I could build eudev 3.2.7-7 fixing a ln bug and adding a short sleep (how long?) for start/restart in beowulf as well as build 3.2.9-7~beowulf1 for beowulf-proposed-updates.

06:14:55 <LeePen> gnu_srs4: Will 3.2.9-7 fix the short sleep issue?

06:29:00 <fsmithred> gnu_srs4, I used 10 seconds, the guy who filed the bug used 1 sec, but he's on netboot and I'm on optical media.

06:29:45 <fsmithred> I was thinking of trying a shorter time, like maybe 2 seconds, but I'm also waiting to see if I learn anything useful before I burn another dvd that will get put in the trash.

06:40:13 <onefang> Ah, opening up the same jitsi meeting in different tabs gives you a poor mans echo service. lol

06:40:25 <fsmithred> onefang, I do isohybrid and also multi-boot live-usb that's on a fat32 partition. I can send you email showing the file layouts I use, later today or tonight.

06:41:18 <onefang> Sure, thanks. This is what you want EFI SYslinux / Isolinux to do?

06:41:46 <fsmithred> yeah, right now the isos use isolinux for bios and grub for uefi

06:42:01 <fsmithred> would be nice to have them both use the same bootloader if possible

06:42:37 <onefang> Send me the details, I'll see if I can get it to work.

06:43:13 <gnu_srs4> LeePen: Yes, the plan was to introduce a sleep, dunno yet how long is needed!

06:43:37 <onefang> In case you didn't know, Grub invaders is a space invaders game built to be booted directly from Grub.

06:43:40 <fsmithred> gnu_srs4, I can try a shorter time later today and let you know what happens

06:44:03 <fsmithred> onefang, yes, I know that, and I was surprised to see that you got it to work with syslinux.

06:45:13 <onefang> The only problem with Grub invaders, and it didn't matter if I booted it from grub or syslinux, is it doesn't like my KVM. I have to plug a keyboard in directly. ror

06:45:31 <fsmithred> gnu_srs4, if you have any educated guesses about how many second I should use, please speak up.

06:56:39 <LeePen> gnu_srs4: Then since you wanted 3.2.9 in beowulf, I would roll it and the sleep together in beowulf-proposed-updates.

06:57:14 <LeePen> Since beowulf is now released I think that is the right way in for package updates.

10:19:56 <amesser> LeePen: Regarding the gitea, i have migrated my personal repos. When updating a package, should I still push to gitlab or only to gitea?

10:20:24 <amesser> What is about the devuan-packages, should i still push changes to gitlab or just wait for gitea to be ready?

10:33:37 <golinux> amesser: After rrq migrated www, I could not even view the old source on gitlab.

10:36:27 <amesser> Hmm, ok. I see, devuan packages are already "archieved" in gitlab.

10:36:53 <amesser> So I guess, no need to push the updates there anymore.

10:52:02 <LeePen> I doubt it will be long before we will swap the DNS over so it will be back to git.d.o, but with the new path.

10:53:27 <onefang> I'll have to reconfigure my local repos to push to git.d.o and revert the link changes in my code then. lol

10:54:16 <LeePen> That is why I think we will change quite quickly: to minimise the number of changes that need to be made.

10:55:50 <LeePen> Now the official repos have been migrated, I would have thought the DNS could be tomorrow.

11:09:08 <amesser> Fine. I'll just wait then until all has settled. I'm too busy with other stuff at the moment anyways.

11:11:45 <onefang> fsmithred: which version of Syslinux are you using?

11:15:25 <fsmithred> oh, I missed the -1 at the end. ceres has -2

11:15:29 <onefang> 3:6.04~git20190206.bf6db5b4+dfsg1-2 (from sid / ceres) is the one that works the best for EFI. Which is repeated over and over again in the Syslinux support forums.

11:15:51 <fsmithred> I would guess that they are the same

11:16:15 <onefang> -2 also works better than -1 according to my tests.

11:16:31 <fsmithred> huh. ok. you running it in beowulf?

11:17:18 <onefang> My experiments involve a script I wrote where I can install from the various versions by changing a variable at the top of the script.

11:18:33 <onefang> That includes stretch, stretch-backports, buster, sid, and the official Syslinux releases 6.03, 6.04-pre1, 6.04-pre2, 6.04-pre3, and self compiled from git head.

11:19:07 <onefang> Nid and bullseye have the same version, so I skipped bullseye.

11:19:44 <fsmithred> same version is backported to ascii, too

11:20:34 <onefang> I did start running my tests from ASCII though. I've not installed beowulf anywhere yet. That'll be the next thing after I have tamed Syslinux. B-)

11:27:46 * onefang goes to download Refracta and notices a typo, you misspelled "current" in "***current stable***".

11:32:37 <fsmithred> you want a beowulf they are in the testing folder

11:32:49 <fsmithred> get beta5 if it's there, otherwise beta2

11:33:55 <onefang> refracta10-beta5_xfce_amd64-20200518_0033.iso but no i386 version.

11:35:22 <onefang> Magic Pixie Dust generally prefers 32 bit versions of stuff, but for testing, a 64 bit version of Refracta is OK for now.

11:40:42 <fsmithred> I think you could use beta2. Shouldn't be much different.

11:41:06 <onefang> Which is why you said "otherwise beta2". B-)

11:43:28 <onefang> I have unlimited bandwidth these days, I'll grab the lot.

11:47:21 <fsmithred> note: login:passwords are root:root user:user

11:47:42 <fsmithred> no root login, but user login with password

11:48:00 <onefang> Well, I saw the passwords, not the bit about ssh.

11:48:25 <fsmithred> makes it useful but a little more dangerous

11:48:52 <onefang> Getting them to boot under EFI using Syslinux is the important bit for me. Logging in is entirely optional. B-)

12:02:35 <fsmithred> gnu_srs4, it appears that sleep for 1 second is enough when booting from dvd.

12:53:10 <onefang> I copied the live directory from beta5 to my test USB stick, and the relevant syslinux menu entires. Boots to desktop, didn't even need to log in. Syslinux booted it fine from BIOS and EFI.

12:53:56 <onefang> Next thing to do - isohybrid and isolinux. Those might take a while, this was just the quick test.