Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.os.linux.misc > #73314 > unrolled thread

Re: Linux 32 bit support days are numbered

Back to article view | Back to comp.os.linux.misc

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: Linux 32 bit support days are numbered Fritz Wuehler <fritz@spamexpire-202509.rodent.frell.theremailer.net> - 2025-09-03 22:24 +0200
    Re: Linux 32 bit support days are numbered Marco Moock <mm@dorfdsl.de> - 2025-09-04 06:56 +0200
      Re: Linux 32 bit support days are numbered John McCue <jmclnx@gmail.com.invalid> - 2025-09-04 15:44 +0000
        Re: Linux 32 bit support days are numbered "Carlos E.R." <robin_listas@es.invalid> - 2025-09-04 18:13 +0200
          Re: Linux 32 bit support days are numbered John Ames <commodorejohn@gmail.com> - 2025-09-04 09:26 -0700
          Re: Linux 32 bit support days are numbered Richard Kettlewell <invalid@invalid.invalid> - 2025-09-04 17:47 +0100
            Re: Linux 32 bit support days are numbered Bobbie Sellers <bliss-sf4ever@dslextreme.com> - 2025-09-04 10:06 -0700
            Re: Linux 32 bit support days are numbered "Carlos E.R." <robin_listas@es.invalid> - 2025-09-04 21:47 +0200
    Re: Linux 32 bit support days are numbered Marc Haber <mh+usenetspam1118@zugschl.us> - 2025-09-04 09:39 +0200
      Re: Linux 32 bit support days are numbered The Natural Philosopher <tnp@invalid.invalid> - 2025-09-04 09:49 +0100
        Re: Linux 32 bit support days are numbered Marc Haber <mh+usenetspam1118@zugschl.us> - 2025-09-04 17:23 +0200
          Re: Linux 32 bit support days are numbered Lawrence D’Oliveiro <ldo@nz.invalid> - 2025-09-05 00:18 +0000
            Re: Linux 32 bit support days are numbered c186282 <c186282@nnada.net> - 2025-09-05 04:19 -0400

#73314 — Re: Linux 32 bit support days are numbered

Nuno Silva <nunojsi...@invalid.invalid> [NS]:
NS> Stuff like network news and email and file editing and coding on
NS> Emacs/vim, working with VCS repositories, typesetting in LaTeX,
NS> chat in well-designed plaforms with light clients, that should
NS> not be hitting singificant performance issues.

Besides personal computing, 10-20 year old computers are still
more than adequate even for (almost hidden) *enterprise*
applications: log servers, employee sign-in/sign-out systems,
IVR systems, smart building monitoring, access control.

Some of these applications have been relegated to the cloud in
the last few years, but thousands of enterprises still run them
on-premise.

If it ain't broke and all that...

[toc] | [next] | [standalone]

#73334

On 03.09.2025 22:24 Fritz Wuehler wrote:

> Nuno Silva <nunojsi...@invalid.invalid> [NS]:
> NS> Stuff like network news and email and file editing and coding on
> NS> Emacs/vim, working with VCS repositories, typesetting in LaTeX,
> NS> chat in well-designed plaforms with light clients, that should
> NS> not be hitting singificant performance issues.  
> 
> Besides personal computing, 10-20 year old computers are still
> more than adequate even for (almost hidden) *enterprise*
> applications: log servers, employee sign-in/sign-out systems,
> IVR systems, smart building monitoring, access control.

Indeed, but 10 year old computers are now retired because of W10 EOL,
so they can use them and install current Linux/BSD and don't mess with
the 25 year old stuff. Much cheaper, as time is money in that case.

[toc] | [prev] | [next] | [standalone]

#73380

Marco Moock <mm@dorfdsl.de> wrote:
<snip> 
> Indeed, but 10 year old computers are now retired because of W10 EOL,
> so they can use them and install current Linux/BSD and don't mess with
> the 25 year old stuff. Much cheaper, as time is money in that case.

I forgot about this, plus IIRC October 2025 is the drop dead
month for older hardware and Windows.  I will have to make a
note to watch used sites for good deals :)

-- 
[t]csh(1) - "An elegant shell, for a more... civilized age."
                        - Paraphrasing Star Wars

[toc] | [prev] | [next] | [standalone]

#73385

On 2025-09-04 17:44, John McCue wrote:
> Marco Moock <mm@dorfdsl.de> wrote:
> <snip>
>> Indeed, but 10 year old computers are now retired because of W10 EOL,
>> so they can use them and install current Linux/BSD and don't mess with
>> the 25 year old stuff. Much cheaper, as time is money in that case.
> 
> I forgot about this, plus IIRC October 2025 is the drop dead
> month for older hardware and Windows.  I will have to make a
> note to watch used sites for good deals :)
> 

June 2026, graphic cards die.

Spanish link - I don't have an English one:

<https://elchapuzasinformatico.com/2025/09/tarjetas-graficas-no-arrancaran-pc-junio-2026-secure-boot-microsoft-uefi-gop/>

DeepL translation:

In June 2026, your PC may not start up because of your graphics card's 
UEFI GOP

By Protector Indefinido  01/09/2025


In a few months, the world of PC hardware could face a silent but 
serious, very serious problem. The reason is not a physical failure or 
planned obsolescence, but something much more bureaucratic: the expiry 
of a certificate that until now signed the UEFI firmware of many GPUs, 
specifically the Microsoft UEFI CA 2011. If there is no timely response, 
those affected, millions of users, could find themselves with black 
screens from the moment they turn on their PCs, without getting past 
POST, and in some cases, with computers rendered virtually unusable due 
to graphics cards with an unsigned UEFI GOP by 2026.

Virtually no one had realised the significance of what we are about to 
see, but after collating the information and based on what Microsoft has 
revealed, it does appear to be true, as it all stems from an analysis on 
Reddit by user Gaseousgalaxy. This raises the key and most pertinent 
question: why is this issue so important and what does the so-called 
UEFI GOP have to do with all this?

(... continues on the link)

-- 
Cheers, Carlos.

[toc] | [prev] | [next] | [standalone]

#73386

On Thu, 4 Sep 2025 18:13:02 +0200
"Carlos E.R." <robin_listas@es.invalid> wrote:

> June 2026, graphic cards die.
> 
> Spanish link - I don't have an English one:
> 
> <https://elchapuzasinformatico.com/2025/09/tarjetas-graficas-no-arrancaran-pc-junio-2026-secure-boot-microsoft-uefi-gop/>

Gosh, it's almost like introducing code signing to firmware was a
terrible idea that was destined to result in exactly this scenario :/

[toc] | [prev] | [next] | [standalone]

#73387

"Carlos E.R." <robin_listas@es.invalid> writes:
> June 2026, graphic cards die.
>
> Spanish link - I don't have an English one:
>
> <https://elchapuzasinformatico.com/2025/09/tarjetas-graficas-no-arrancaran-pc-junio-2026-secure-boot-microsoft-uefi-gop/>
>
> DeepL translation:
>
> In June 2026, your PC may not start up because of your graphics card's
> UEFI GOP
>
> By Protector Indefinido  01/09/2025
>
>
> In a few months, the world of PC hardware could face a silent but
> serious, very serious problem. The reason is not a physical failure or
> planned obsolescence, but something much more bureaucratic: the expiry
> of a certificate that until now signed the UEFI firmware of many GPUs,
> specifically the Microsoft UEFI CA 2011. If there is no timely
> response, those affected, millions of users, could find themselves
> with black screens from the moment they turn on their PCs, without
> getting past POST, and in some cases, with computers rendered
> virtually unusable due to graphics cards with an unsigned UEFI GOP by
> 2026.
>
> Virtually no one had realised the significance of what we are about to
> see, but after collating the information and based on what Microsoft
> has revealed, it does appear to be true, as it all stems from an
> analysis on Reddit by user Gaseousgalaxy. This raises the key and most
> pertinent question: why is this issue so important and what does the
> so-called UEFI GOP have to do with all this?
>
> (... continues on the link)

I think this is largely scaremongering.  See
https://mjg59.dreamwidth.org/72892.html for a better-informed source.

Briefly:

Certificate expiry is not likely to be an issue here. UEFI firmware does
not enforce expiry dates when it verifies certificate. This is pretty
common practice for code signing.

The real issue is that at some (currently unknown) point in the future
new code will stop being signed with the ‘expired’ key, and only signed
with its replacement. When that happens, platforms that only trust the
old key won’t be able to install video cards or boot operating systems
that were only signed with the new key.

AFAIK disabling secure boot will work as a workaround.

-- 
https://www.greenend.org.uk/rjk/

[toc] | [prev] | [next] | [standalone]

#73388

On 9/4/25 09:47, Richard Kettlewell wrote:
> "Carlos E.R." <robin_listas@es.invalid> writes:
>> June 2026, graphic cards die.
>>
>> Spanish link - I don't have an English one:
>>
>> <https://elchapuzasinformatico.com/2025/09/tarjetas-graficas-no-arrancaran-pc-junio-2026-secure-boot-microsoft-uefi-gop/>
>>
>> DeepL translation:
>>
>> In June 2026, your PC may not start up because of your graphics card's
>> UEFI GOP
>>
>> By Protector Indefinido  01/09/2025
>>
>>
>> In a few months, the world of PC hardware could face a silent but
>> serious, very serious problem. The reason is not a physical failure or
>> planned obsolescence, but something much more bureaucratic: the expiry
>> of a certificate that until now signed the UEFI firmware of many GPUs,
>> specifically the Microsoft UEFI CA 2011. If there is no timely
>> response, those affected, millions of users, could find themselves
>> with black screens from the moment they turn on their PCs, without
>> getting past POST, and in some cases, with computers rendered
>> virtually unusable due to graphics cards with an unsigned UEFI GOP by
>> 2026.
>>
>> Virtually no one had realised the significance of what we are about to
>> see, but after collating the information and based on what Microsoft
>> has revealed, it does appear to be true, as it all stems from an
>> analysis on Reddit by user Gaseousgalaxy. This raises the key and most
>> pertinent question: why is this issue so important and what does the
>> so-called UEFI GOP have to do with all this?
>>
>> (... continues on the link)
> 
> I think this is largely scaremongering.  See
> https://mjg59.dreamwidth.org/72892.html for a better-informed source.
> 
> Briefly:
> 
> Certificate expiry is not likely to be an issue here. UEFI firmware does
> not enforce expiry dates when it verifies certificate. This is pretty
> common practice for code signing.
> 
> The real issue is that at some (currently unknown) point in the future
> new code will stop being signed with the ‘expired’ key, and only signed
> with its replacement. When that happens, platforms that only trust the
> old key won’t be able to install video cards or boot operating systems
> that were only signed with the new key.
> 
> AFAIK disabling secure boot will work as a workaround.
	
	And using GNU/Linux which require shutting off the so-called Secure-Boot
  as a preparatory step in the installation of the majority of the 
distributions.

	By the way dual booting of Windows with Linux is to be discouraged as
  everytime Windows updates its kernel it messes up the UEFI settings.  And
  it never tells the users when it is updating the kernel so typically 
the user
  ends up with a non-bootable Linux install which require repairs i.e. 
reinstall
  of the boot tools usually GRUB2.
	Better if you must use Windows is to use a tool like Virtual Box to isolate
  Windows for your work and not worry about messing up the boot or the basic
  insecurity of Windows as it will be protected inside of Linux.

  bliss- Dell Precision 7730- PCLOS 2025.08- Linux 6.12.44-pclos1- KDE 
Plasma 6.4.4

[toc] | [prev] | [next] | [standalone]

#73401

On 2025-09-04 18:47, Richard Kettlewell wrote:
> "Carlos E.R." <robin_listas@es.invalid> writes:
>> June 2026, graphic cards die.
>>
>> Spanish link - I don't have an English one:
>>
>> <https://elchapuzasinformatico.com/2025/09/tarjetas-graficas-no-arrancaran-pc-junio-2026-secure-boot-microsoft-uefi-gop/>
>>
>> DeepL translation:
>>
>> In June 2026, your PC may not start up because of your graphics card's
>> UEFI GOP
>>
>> By Protector Indefinido  01/09/2025
>>
>>
>> In a few months, the world of PC hardware could face a silent but
>> serious, very serious problem. The reason is not a physical failure or
>> planned obsolescence, but something much more bureaucratic: the expiry
>> of a certificate that until now signed the UEFI firmware of many GPUs,
>> specifically the Microsoft UEFI CA 2011. If there is no timely
>> response, those affected, millions of users, could find themselves
>> with black screens from the moment they turn on their PCs, without
>> getting past POST, and in some cases, with computers rendered
>> virtually unusable due to graphics cards with an unsigned UEFI GOP by
>> 2026.
>>
>> Virtually no one had realised the significance of what we are about to
>> see, but after collating the information and based on what Microsoft
>> has revealed, it does appear to be true, as it all stems from an
>> analysis on Reddit by user Gaseousgalaxy. This raises the key and most
>> pertinent question: why is this issue so important and what does the
>> so-called UEFI GOP have to do with all this?
>>
>> (... continues on the link)
> 
> I think this is largely scaremongering.  See
> https://mjg59.dreamwidth.org/72892.html for a better-informed source.
> 
> Briefly:
> 
> Certificate expiry is not likely to be an issue here. UEFI firmware does
> not enforce expiry dates when it verifies certificate. This is pretty
> common practice for code signing.
> 
> The real issue is that at some (currently unknown) point in the future
> new code will stop being signed with the ‘expired’ key, and only signed
> with its replacement. When that happens, platforms that only trust the
> old key won’t be able to install video cards or boot operating systems
> that were only signed with the new key.
> 
> AFAIK disabling secure boot will work as a workaround.

Not on W11. I have a virtual machine with W11. I double booted to a 
Linux CD to force delete some file, and for that I had to disable secure 
boot. Subsequently I booted Windows, to find that network was gone. So I 
had to reboot and enable back secure boot.

So people double booting need secure booting enabled.


-- 
Cheers, Carlos.

[toc] | [prev] | [next] | [standalone]

#73336

Fritz Wuehler <fritz@spamexpire-202509.rodent.frell.theremailer.net>
wrote:
>Besides personal computing, 10-20 year old computers are still
>more than adequate even for (almost hidden) *enterprise*
>applications: log servers, employee sign-in/sign-out systems,
>IVR systems, smart building monitoring, access control.

Things like that get replaced in sane companies because the company
can't afford them to break.

-- 
----------------------------------------------------------------------------
Marc Haber         |   " Questions are the         | Mailadresse im Header
Rhein-Neckar, DE   |     Beginning of Wisdom "     | 
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fon: *49 6224 1600402

[toc] | [prev] | [next] | [standalone]

#73353

On 04/09/2025 08:39, Marc Haber wrote:
> Fritz Wuehler <fritz@spamexpire-202509.rodent.frell.theremailer.net>
> wrote:
>> Besides personal computing, 10-20 year old computers are still
>> more than adequate even for (almost hidden) *enterprise*
>> applications: log servers, employee sign-in/sign-out systems,
>> IVR systems, smart building monitoring, access control.
> 
> Things like that get replaced in sane companies because the company
> can't afford them to break.
> 
We used to end up putting linux on all the old kit and running BIND on 
them. Perfectly fast enough.

There was nothing unreliable about them.

And with two BINDc servers one can fail ...and not be a huge issue

-- 
  “A leader is best When people barely know he exists. Of a good leader, 
who talks little,When his work is done, his aim fulfilled,They will say, 
“We did this ourselves.”

― Lao Tzu, Tao Te Ching

[toc] | [prev] | [next] | [standalone]

#73376

The Natural Philosopher <tnp@invalid.invalid> wrote:
>On 04/09/2025 08:39, Marc Haber wrote:
>> Fritz Wuehler <fritz@spamexpire-202509.rodent.frell.theremailer.net>
>> wrote:
>>> Besides personal computing, 10-20 year old computers are still
>>> more than adequate even for (almost hidden) *enterprise*
>>> applications: log servers, employee sign-in/sign-out systems,
>>> IVR systems, smart building monitoring, access control.
>> 
>> Things like that get replaced in sane companies because the company
>> can't afford them to break.
>> 
>We used to end up putting linux on all the old kit and running BIND on 
>them. Perfectly fast enough.
>
>There was nothing unreliable about them.
>
>And with two BINDc servers one can fail ...and not be a huge issue

Yes. And to fix the door control system you need to enter the server
room, which is protected by said door control system. I know of
companies that had half a day of outage because they didn't find the
actual key.

That's why such systems get replaced.

-- 
----------------------------------------------------------------------------
Marc Haber         |   " Questions are the         | Mailadresse im Header
Rhein-Neckar, DE   |     Beginning of Wisdom "     | 
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fon: *49 6224 1600402

[toc] | [prev] | [next] | [standalone]

#73410

On Thu, 04 Sep 2025 17:23:40 +0200, Marc Haber wrote:

> And to fix the door control system you need to enter the server
> room, which is protected by said door control system.

You can just hear the voice saying “I’m sorry Dave, I’m afraid I can’t do 
that” ...

[toc] | [prev] | [next] | [standalone]

#73423

On 9/4/25 8:18 PM, Lawrence D’Oliveiro wrote:
> On Thu, 04 Sep 2025 17:23:40 +0200, Marc Haber wrote:
> 
>> And to fix the door control system you need to enter the server
>> room, which is protected by said door control system.
> 
> You can just hear the voice saying “I’m sorry Dave, I’m afraid I can’t do
> that” ...

   HaHaHaHaHa !!!

   Try a fire-axe  :-)

   Let's see the software encryption cope with brute force ...

[toc] | [prev] | [standalone]

Back to top | Article view | comp.os.linux.misc

csiph-web