Core v14.0beta1

[aus9]

Hi Rich
if adwaita-icon-theme pulls in gtk-update-icon-cache then it would also pull in hicolor-icon-theme?

[CNK]

I just tested TC 14.0beta1 on an old PC with a Cx486DX4 CPU and booting fails when starting init with an "invalid opcode" error in ld-linux.so.2.

My previous installation of TC 13.0beta1, done the same way, still boots
successfully. In fact I'm posting with it in elinks now.

Unfortunately I won't have time to test on other systems such as my Pentium 1 for a few days, and I probably won't have time/energy to reply on the forum for about a day after now either.

[Juanito]

Hmm - I compiled both with a uname_hack forcing to i486 and -march=i486 - let me double-check

[jazzbiker]

Unfortunately TC14 release_candidates/distribution_files fail to load on my old Pentium MMX box reporting "Attempted to kill init! exit code 0x00000004". TC12 works fine on this box.

[Juanito]

I rebuilt glibc and created rootfs_i486.gz - maybe you could test

http://tinycorelinux.net/14.x/x86/release_candidates/distribution_files/rootfs_i486.gz

[Juanito]

There are some extensions that call  gtk-update-icon-cache  for their own icons.
Those extensions have been tracked down and will be fixed.

Hopefully done

[jazzbiker]

I rebuilt glibc and created rootfs_i486.gz - maybe you could test

Thanks, Juanito! Core14 loads fine.

Code: [Select]

              total        used        free      shared  buff/cache   available
Mem:          54076        6372       21452       19484       26252       24616
Swap:        262076           0      262076

I may have missed something in the discussion on rebuildfstab, but on my ancient box I have the feeling that rebuildfstab is much faster. Is it new version by @Rich?

Thank You very much!

Maybe dmesg will be of interest.

Best regards!

[Juanito]

It’s a newer version of rebuildfstab than 13.x, but not the very latest version - I’ll update it with the next version of 14.x

[Rich]

Hi aus9

if adwaita-icon-theme pulls in gtk-update-icon-cache then it would also pull in hicolor-icon-theme?

Yes.

[aus9]

Hi Rich
thanks

Hi Juanito
Thanks heaps for taking pity on me....again....and taking time out to update meson when I know you have other things to do.
Some of my sway related packages complains about old meson and others about python3.6
I was planning to try out a ~/local/ install showing up at mesonbuild.com but my coding skills are still poor.

No need to reply to my bodge attempt sent by pm thanks. That did actually work for seatd and borked for wlroots

once again thanks and sorry for drifting off TCB topic

[aus9]

Hi GNUser

I am confused again after Apps update for submitqc...I have

Code: [Select]

submitqc: icewm.tcz file and directory ownership/permissions follow wiki recommendations?
squashfs root dir ownership not tc:staff. Will rebuild squashfs to fix
I thought we were going with root:staff?

when I look at the last output I see
/tmp/submitqc/basicfileperms
      
and I think this was once something like tce.install changed...while basicfileperms is a bit vague for me

sorry if I have not read something...again

[jazzbiker]

It would be nice to adjust http://tinycorelinux.net/14.x/x86/tcz/tcc.tcz.dep to match an actual kernel version:

Code: [Select]

sed -i 's/5.15/6.1/' tcc.tcz.dep

Regards!

[GNUser]

Hi aus9. root:staff is for startup script and tce.installed directory, tc:staff on squashfs's root dir, root:root for everything else.

I got root:staff and root:root from wiki recommendations. tc:staff on squashfs's root dir is not specifically mentioned but is seen on vast  majority of repo's extensions, including all of Juanito's extensions.

[GNUser]

Hi aus9. Just two more thoughts for completeness's sake:

1. Inspecting extensions
To inspect ownerships and permissions inside an extension, don't extract it. When contents are extracted and written to your drive (or to your RAM), ownerships and permissions can change. To inspect an extension, look inside it without extracting it, like this:

Code: [Select]

unsquashfs -lls foo.tczTo see ownership and permissions of root directory of all your loaded extensions, do:

Code: [Select]

ls -l /tmp/tcloop

2. Flexibility of the extension system
TCL is very flexible and forgiving when it comes to ownerships and permissions in extensions. For example, wiki recommends root:staff 775 for startup scripts, but submitqc was forcing tc:staff 755 for a long time without any problems. Also, wiki recommends root:root for most directories and files, but I've seen extensions where everything is owned by tc:staff (or 1000:staff in the case of some of my old extensions) and they work just fine. Furthermore, most extensions have tc:staff ownership of squashfs-root directory, but you'll also see root:root (or 1000:staff in the case of some of my old extensions) and it makes little difference.

IMHO the job of submitqc should be twofold: 1. correct obvious errors (most important) and 2. ensure conventions and wiki recommendations are followed (less important, but still a nice service and cosmetically desirable).

    [Edit]: Fixed typo, changed  1001:staff  to  1000:staff.  Rich

[Rich]

Hi jazzbiker

It would be nice to adjust http://tinycorelinux.net/14.x/x86/tcz/tcc.tcz.dep to match an actual kernel version: ...

Fixed.