Re: [gentoo-user] a day of PAIN.

Re: [gentoo-user] a day of PAIN.

[Ashley Dixon]

[-- Attachment #1: Type: text/plain, Size: 1543 bytes --]

On Tue, May 12, 2020 at 06:45:21PM -0400, Alan Grimes wrote:
> I don't feel comfortable manually yanking any of these packages... What
> is the cannonical solution to this CF?
> 
> ####################
> Total: 1701 packages (628 upgrades, 16 new, 2 in new slots, 1055
> reinstalls, 1 uninstall), Size of downloads: 10,521 KiB
> Conflict: 4 blocks (2 unsatisfied)
> 
>  * Error: The above package list contains packages which cannot be
>  * installed at the same time on the same system.
> 
>   (sys-apps/shadow-4.8.1-r2:0/0::gentoo, ebuild scheduled for merge)
> pulled in by
>     sys-apps/shadow required by @system
>     sys-apps/shadow required by (net-misc/openssh-8.2_p1-r6:0/0::gentoo,
> ebuild scheduled for merge) USE="X pam pie ssl -X509 -audit -bindist
> -debug -hpn -kerberos -ldns -libedit -libressl -livecd -sctp
> -security-key (-selinux) -static -test -xmss" ABI_X86="(64)"
> 
>   (sys-apps/util-linux-2.35.1-r2:0/0::gentoo, ebuild scheduled for
> merge) pulled in by
>     sys-apps/util-linux required by
> (app-text/build-docbook-catalog-1.21:0/0::gentoo, ebuild scheduled for
> merge) USE="" ABI_X86="(64)"
> ####################

From the shadow ebuild; emphasis on the `su? ( !sys-apps/util-linux[su-] )`:

        RDEPEND="${COMMON_DEPEND}
            pam? ( >=sys-auth/pambase-20150213 )
            su? ( !sys-apps/util-linux[su(-)] )
        "

What are your USE-flags on util-linux ?

-- 

Ashley Dixon
suugaku.co.uk

2A9A 4117
DA96 D18A
8A7B B0D2
A30E BF25
F290 A8AA


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[gentoo-user] a day of PAIN.

[Alan Grimes]

Ouch.....

Yesterday I made the dubious decision to go ahead and mobo swap my aging
1800x to [nondisclosed].

Ok... The new mobo is EFI only...

Back in the good old days, the BIOS would just load the first sector off
your floppy drive, check for a basic sanity check pattern to make sure
the floppy cable was connected, working, and not cross-circuited, and
then it would work.

DOS provided a single command for making a bootable floppy and it
**NEVER FAILED**.

This thing does not boot off the motorized cupholder, and setting it up
is like the twelve labors of Herculese and you have to get them all
exactly correct or it doesn't work at all and it doesn't tell you why it
doesn't work. =| I don't even know how it's working.

I had intended it to be a UEFI drive from the start but my old mobo was
very forgiving and I think the thing was actually booting in BIOS mode.

Turns out the missing ingredient was "efibootmgr"... =\

The designers of the UEFI standard cared about nothing but their own job
security.


So I got to see my GRUB menu.  YAY....    (lasted about 5 seconds...)

Machine appeared to stall at kernel loading.



 >>> YOU KNOW SOMETHING WAS DESIGNED BY PENGUINS WHEN IT MAKES THE WORST
POSSIBLE DESIGN DECISIONS AND THEN BLAMES THE USER FOR NOT MAKING IT
WORK... <<<

Case in point, EFI framebuffer.          I could end the post there....

Why is this not a forced-on setting for any machine with UEFI enabled? I
can't imagine that this would be unacceptable for more than 0.001% of
the install base.

So therefore it is disabled by default and you have to enable it in at
least three different deeply burried settings otherwise the kernel
appears to hang. 

In the good old days the BIOS did the following:

Is there something resembling a CPU in the socket? Good!
Where's the goddamn video card so I can post error messages?
Does this machine even have RAM?

So obviously linux is doing this profoundly wrong....


Ok, so it's time to emptytree world my system:

I don't feel comfortable manually yanking any of these packages... What
is the cannonical solution to this CF?

####################
Total: 1701 packages (628 upgrades, 16 new, 2 in new slots, 1055
reinstalls, 1 uninstall), Size of downloads: 10,521 KiB
Conflict: 4 blocks (2 unsatisfied)

 * Error: The above package list contains packages which cannot be
 * installed at the same time on the same system.

  (sys-apps/shadow-4.8.1-r2:0/0::gentoo, ebuild scheduled for merge)
pulled in by
    sys-apps/shadow required by @system
    sys-apps/shadow required by (net-misc/openssh-8.2_p1-r6:0/0::gentoo,
ebuild scheduled for merge) USE="X pam pie ssl -X509 -audit -bindist
-debug -hpn -kerberos -ldns -libedit -libressl -livecd -sctp
-security-key (-selinux) -static -test -xmss" ABI_X86="(64)"

  (sys-apps/util-linux-2.35.1-r2:0/0::gentoo, ebuild scheduled for
merge) pulled in by
    sys-apps/util-linux required by
(app-text/build-docbook-catalog-1.21:0/0::gentoo, ebuild scheduled for
merge) USE="" ABI_X86="(64)"
####################

-- 
The vaccine is a LIE. 

Powers are not rights.

Re: [gentoo-user] a day of PAIN.

[Andrew Udvare]

> On 2020-05-12, at 18:45, Alan Grimes <ALONZOTG@verizon.net> wrote:
> 
> Why is this not a forced-on setting for any machine with UEFI enabled? I
> can't imagine that this would be unacceptable for more than 0.001% of
> the install base.

Because not every user has UEFI and many just use the BIOS compatibility layer (CSM) if they don't care about UEFI and Secure Boot. The CSM is going to be there for a very long time.

If you don't need UEFI features, then there's nothing really wrong with using CSM. I am using UEFI because I like having a signed kernel and signed boot loader (systemd-boot).

--
Andrew

Re: [gentoo-user] a day of PAIN.

[madscientistatlarge]

I'm using a nearly 10 year old server, bought specifically so I can compile faster.  None of my machines is newer than that.  None of my machines support UEFI other than an older imac.  The server, which was inexpensive ($500) only has 48 cores, 64 as soon as I update the processors (2 generations newer and more cores!).
I love server grade hardware!  Servers lose value quickly, they are a great bargain if you don't mind the sound of a wind tunnel or have another room to put it in.  This machine was probably $20-30K new, no slouch, it was previously used for high frequency trading.  Sadly the seller does seem to have emptied their' warehouse.


Sent with ProtonMail Secure Email.

‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
On Thursday, May 14, 2020 11:50 AM, Andrew Udvare <audvare@gmail.com> wrote:

> > On 2020-05-12, at 18:45, Alan Grimes ALONZOTG@verizon.net wrote:
> > Why is this not a forced-on setting for any machine with UEFI enabled? I
> > can't imagine that this would be unacceptable for more than 0.001% of
> > the install base.
>
> Because not every user has UEFI and many just use the BIOS compatibility layer (CSM) if they don't care about UEFI and Secure Boot. The CSM is going to be there for a very long time.
>
> If you don't need UEFI features, then there's nothing really wrong with using CSM. I am using UEFI because I like having a signed kernel and signed boot loader (systemd-boot).
>
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Andrew