Requests for -current (14.2-->15.0)

camerabambai · 01-24-2022, 11:19 AM

Quote:

Originally Posted by marav

keyservers are apparently hard coded (cf. dirmngr/server.c in source code)

But you can make your own gpg.conf

Code:

$ echo "keyserver hkps://keyserver.ubuntu.com:443" > ~/.gnupg/gpg.conf

Thanks
I already did this of course, only suggest a patch because the default gpg keyserver is not working anymore

marav · 01-24-2022, 11:28 AM

Quote:

Originally Posted by camerabambai

Thanks
I already did this of course, only suggest a patch because the default gpg keyserver is not working anymore

the part of the server.c file:

Code:

  /* We used to have DNS CNAME redirection from the URLs below to
   * sks-keyserver. pools.  The idea was to allow for a quick way to
   * switch to a different set of pools.  The problem with that
   * approach is that TLS needs to verify the hostname and - because
   * DNS is not secured - it can only check the user supplied hostname
   * and not a hostname from a CNAME RR.  Thus the final server all
   * need to have certificates with the actual pool name as well as
   * for keys.gnupg.net - that would render the advantage of
   * keys.gnupg.net useless and so we better give up on this.  Because
   * the keys.gnupg.net URL are still in widespread use we do a static
   * mapping here.
   */
  if (!strcmp (uri, "hkps://keys.gnupg.net")
      || !strcmp (uri, "keys.gnupg.net"))
    uri = "hkps://keyserver.ubuntu.com";
  else if (!strcmp (uri, "https://keys.gnupg.net"))
    uri = "hkps://keyserver.ubuntu.com";
  else if (!strcmp (uri, "hkp://keys.gnupg.net"))
    uri = "hkp://pgp.surf.nl";
  else if (!strcmp (uri, "http://keys.gnupg.net"))
    uri = "hkp://pgp.surf.nl:80";
  else if (!strcmp (uri, "hkps://http-keys.gnupg.net")
           || !strcmp (uri, "http-keys.gnupg.net"))
    uri = "hkps://keyserver.ubuntu.com";
  else if (!strcmp (uri, "https://http-keys.gnupg.net"))
    uri = "hkps://keyserver.ubuntu.com";
  else if (!strcmp (uri, "hkp://http-keys.gnupg.net"))
    uri = "hkp://pgp.surf.nl";
  else if (!strcmp (uri, "http://http-keys.gnupg.net"))
    uri = "hkp://pgp.surf.nl:80";

  item = xtrymalloc (sizeof *item + strlen (uri));
  if (!item)
    return gpg_error_from_syserror ();

  item->next = NULL;
  item->parsed_uri = NULL;
  strcpy (item->uri, uri);

franzen · 01-24-2022, 12:00 PM

Quote:

Originally Posted by Sl4ck3ver

A start/stop script for the "dhcp" package (rc.dhcpd), please :-)

In the meantime you might use /etc/rc.d/rc.dnsmasq

RadicalDreamer · 01-24-2022, 01:18 PM

Quote:

Originally Posted by chrisretusn

Before that message is displayed, this one is shown.

Code:

Your kernel image was updated, and your /etc/lilo.conf indicates
the use of an initrd for at least one of your kernels. Be sure to
regenerate the initrd for the new kernel and handle any needed
updates to your bootloader.

Press the "Enter" key to continue...

True, but that is a lot of words and they could just "Enter" their way to the short and sweet debug message that tells them to reboot and focus on that. The Debug should say something short and sweet for kernel updates like, "Update your initrd and bootloader, and then reboot for changes to take effect" for kernel updates. There should be a way to parse the updates to see if kernels are updated and split them off from the rest of that message. The guy from Linus Tech Tips uninstalled his DE despite the warnings.

The message is about lilo and people with newer systems probably don't use it. A path to a document to explain this stuff would be helpful to users that are confused. I think it should instill fear in people before they "Enter" it away,

Code:

"*WARNING KERNEL UPDATED!* 
Please immediately check "/usr/../../" for more information. 
"*FAILURE TO UPDATE BOOTLOADER BEFORE RESTARTING WILL BREAK SYSTEM!*"

Press the "Enter" key to continue...

That is my reasoning and my opinion.

marav · 01-24-2022, 01:35 PM

Quote:

Originally Posted by RadicalDreamer

True, but that is a lot of words and they could just "Enter" their way to the short and sweet debug message that tells them to reboot and focus on that. The Debug should say something short and sweet for kernel updates like, "Update your initrd and bootloader, and then reboot for changes to take effect" for kernel updates. There should be a way to parse the updates to see if kernels are updated and split them off from the rest of that message. The guy from Linus Tech Tips uninstalled his DE despite the warnings.

The message is about lilo and people with newer systems probably don't use it. A path to a document to explain this stuff would be helpful to users that are confused. I think it should instill fear in people before they "Enter" it away,

Code:

"*WARNING KERNEL UPDATED!* 
Please immediately check "/usr/../../" for more information. 
"*FAILURE TO UPDATE BOOTLOADER BEFORE RESTARTING WILL BREAK SYSTEM!*"

Press the "Enter" key to continue...

That is my reasoning and my opinion.

Honestly, I clearly don’t know why this initrd.gz is not auto generated after each kernel updates ( ditto with lilo)
Even if you have a custom one, at least it doesn’t fail on reboot

SCerovec · 01-24-2022, 01:39 PM

Quote:

Originally Posted by LuckyCyborg

How about to abandon those ancient technologies used by Tutankhamen when he built piramides and just to use the modern Grub2 ?

Grub2 is smart enough to not need additional liloconfig/eliloconfig runs, assuming that an eventual initrd is properly generated in place.

I remember that years ago here was an user who said many times: do not complicate your life in the name of simplicity!

By the time You persuaded them into modernism, rEFInd has become the more modern, working and simpler to indefinitely maintain alternative.

Join me into the future of effortless booting with rEFInd

LuckyCyborg · 01-24-2022, 02:24 PM

Quote:

Originally Posted by SCerovec

By the time You persuaded them into modernism, rEFInd has become the more modern, working and simpler to indefinitely maintain alternative.

Join me into the future of effortless booting with rEFInd

I known about rEFInd. And I use it often - in some boxes, even I setup an UEFI DUET (which is an UEFI layer for BIOS computers) then I use rEFInd.

The reason is not only the easiness to select the boot devices, but also to make bootable the USB 3.0 PCIE cards which I use in many of my boxes, because they have no native USB 3.0 support.

In the attached screenshots is rEFInd running on top of UEFI DUET, running on top of BIOS from a RS780 motherboard with Athlon x4 605e. And the selected item is one of the USB 3.0 hard drives which I have with a full installation of Slackware-current (and a local Slackware mirror) on them, and which I use as recovery tools.

HOWEVER, this rEFInd is UEFI only and many times is not feasible or is way very difficult to setup an UEFI DUET layer just for it. I for one, I do not managed to setup UEFI DUET on motherboards with DDR2, for example. Just on some with DDR3.

So, Grub2 is the single modern bootloader which support both the (legacy) BIOS and UEFI.

ZhaoLin1457 · 01-24-2022, 04:32 PM

Quote:

Originally Posted by LuckyCyborg

I known about rEFInd. And I use it often - in some boxes, even I setup an UEFI DUET (which is an UEFI layer for BIOS computers) then I use rEFInd.

The reason is not only the easiness to select the boot devices, but also to make bootable the USB 3.0 PCIE cards which I use in many of my boxes, because they have no native USB 3.0 support.

In the attached screenshots is rEFInd running on top of UEFI DUET, running on top of BIOS from a RS780 motherboard with Athlon x4 605e. And the selected item is one of the USB 3.0 hard drives which I have with a full installation of Slackware-current (and a local Slackware mirror) on them, and which I use as recovery tools.

HOWEVER, this rEFInd is UEFI only and many times is not feasible or is way very difficult to setup an UEFI DUET layer just for it. I for one, I do not managed to setup UEFI DUET on motherboards with DDR2, for example. Just on some with DDR3.

So, Grub2 is the single modern bootloader which support both the (legacy) BIOS and UEFI.

I agree that an UEFI boot manager could not be considered as a default solution for an operating system supposed to boot also from legacy hardware using classical BIOS.

As an aside, I wonder how this UEFI DUET works. It's like a kernel started by a boot loader?

LuckyCyborg · 01-24-2022, 05:05 PM

Quote:

Originally Posted by ZhaoLin1457

As an aside, I wonder how this UEFI DUET works. It's like a kernel started by a boot loader?

Well, technically you can consider the UEFI DUET as being a kernel launched by a dedicated bootloader.

The UEFI DUET "kernel" itself is a file named "Efildr20" which lives in the root of EFI partition, and which is loaded by a special partition MBR.

Essentially, in a way or other (a disk MBR looking for the active partition or even LILO) is started the boot right from EFI partition, then the partition MBR looks for EFI kernel file, loads and starts it.

Honestly, kinda remembers me of how {MS,Free}DOS is loaded, BUT this time the "kernel" is 64bit.

After this file "Efildr20" is launched, you are in 64bit EFI mode, with the usual basic features - the UEFI system looks for /EFI/Boot/Bootx64.efi and execute it, etc...

The really nice part is that beyond EFI (and its framebuffer, as some video cards requires EFI) with a convenient setup, you can load various EFI drivers - aka DXEs, and initialize and power up things unavailable directly from BIOS, like USB 3.0 PCIE cards or NVME hard drives.

bassmadrigal · 01-24-2022, 05:33 PM

Quote:

Originally Posted by ZhaoLin1457

I agree that an UEFI boot manager could not be considered as a default solution for an operating system supposed to boot also from legacy hardware using classical BIOS.

This could be handled the same way as elilo... it only prompts the user if the installer detects it is running in EFI mode, otherwise it defaults to the legacy option (currently lilo).

marav · 01-24-2022, 07:17 PM

Thunderbird 91.5.1

Release notes:
https://www.thunderbird.net/en-US/th.../releasenotes/

perrin4869 · 01-24-2022, 08:07 PM

libadwaita 1.0 won't build under gtk4-4.4 that is available in -current right now, this might limit usability of applications that rely on libadwaita.
Even if libadwaita is not included, without upgrading gtk4 to 4.5 we will not be able to add a SlackBuild for it...

volkerdi · 01-24-2022, 08:53 PM

Quote:

Originally Posted by perrin4869

libadwaita 1.0 won't build under gtk4-4.4 that is available in -current right now, this might limit usability of applications that rely on libadwaita.
Even if libadwaita is not included, without upgrading gtk4 to 4.5 we will not be able to add a SlackBuild for it...

No, sorry.

EDIT: I was almost willing to entertain this request, but it also requires (or pulls from git) a newer version of pango, just as 4.6.x does. For the record, since IMHO gtk4 was barely ready to be included at all, I have no problem whatsoever with SBo providing a SlackBuild for a newer version. But it's not happening in-tree on my end for 15.0.

perrin4869 · 01-24-2022, 09:05 PM

Ah, I missed the discussion in previous pages, sorry ><;;
What about the fix to mkinitrd_command_generator.sh that I've been pushing since last year though :P
https://www.linuxquestions.org/quest...es-4175688858/

volkerdi · 01-24-2022, 09:10 PM

Quote:

Originally Posted by perrin4869

Ah, I missed the discussion in previous pages, sorry ><;;
What about the fix to mkinitrd_command_generator.sh that I've been pushing since last year though :P
https://www.linuxquestions.org/quest...es-4175688858/

Use -R.