FreeBSD 14 Alpha 2 Available For Testing - The Last Series For 32-bit Platforms

[phoronix]

Phoronix: FreeBSD 14 Alpha 2 Available For Testing - The Last Series For 32-bit Platforms

Last week the FreeBSD 14 alpha phase kicked off and available today is the second weekly alpha release for this upcoming major BSD operating system update...

https://www.phoronix.com/news/FreeBSD-14-Alpha-2

[JEBjames]

Michael

Typos

"for detching firmware packages." should be "for fetching firmware packages."

and

"Thw fwget utility initially can get firmware for Intel and AMD GPUs." should be "The fwget utility initially can get firmware for Intel and AMD GPUs."

[hiryu]

I'm okay with Linux and FreeBSD dropping support for running on 32-bit platforms, but I don't know that I'm okay with dropping support for 32-bit applications. Which isn't the case here... yet.

Anyway, I figure OpenBSD and NetBSD will continue supporting 32-bit platforms for a long time to come anyway?

[kpedersen]

Anyway, I figure OpenBSD and NetBSD will continue supporting 32-bit platforms for a long time to come anyway?

OpenBSD is doing a more natural migration away from i386. Some packages are not ported for example. The developers also suggest people just use an amd64 machine because i386 isn't particularly interesting to maintain.

However, yes, for the basics, it will remain available; mostly because building for this target potentially exposes a range of code errors.

[stormcrow]

OpenBSD is doing a more natural migration away from i386. Some packages are not ported for example. The developers also suggest people just use an amd64 machine because i386 isn't particularly interesting to maintain.

However, yes, for the basics, it will remain available; mostly because building for this target potentially exposes a range of code errors.

Usually for as long as hardware is available for testing on and people willing to pay the bills for their electricity. That may become increasingly problematic though as i386-ish systems run hot and most of the non-embedded systems don't have power states to use less electricity when idle... While I understand and have used alternative platforms to help diagnose coding flaws in the past, I'm no longer sure that this argument holds a moral weight in the real world we live in where period x86 systems like the Pentium 4 uses 90-100W of power constantly and doesn't have an idle power state but new ARM and AMD64 can drop to a few Watts while idle, and even my AMD Ryzen 5 3600 CPU uses less electricity when going full blast at around 65-70 Watts. We just hit 105F here today, which is around 5-10F hotter than it should be this time of year. And it's impossible to legitimately deny part of how we got here is due to the computing revolution and dirty electricity generation - and will be for some time to come still, no matter how many clean power initiatives there are.

[curfew]

I'm okay with Linux and FreeBSD dropping support for running on 32-bit platforms, but I don't know that I'm okay with dropping support for 32-bit applications. Which isn't the case here... yet.

And I'm not okay with Linux and FreeBSD devs drinking blood of newborns... Which they don't. What was the point in my message? I have no idea.

[skeevy420]

Usually for as long as hardware is available for testing on and people willing to pay the bills for their electricity. That may become increasingly problematic though as i386-ish systems run hot and most of the non-embedded systems don't have power states to use less electricity when idle... While I understand and have used alternative platforms to help diagnose coding flaws in the past, I'm no longer sure that this argument holds a moral weight in the real world we live in where period x86 systems like the Pentium 4 uses 90-100W of power constantly and doesn't have an idle power state but new ARM and AMD64 can drop to a few Watts while idle, and even my AMD Ryzen 5 3600 CPU uses less electricity when going full blast at around 65-70 Watts. We just hit 105F here today, which is around 5-10F hotter than it should be this time of year. And it's impossible to legitimately deny part of how we got here is due to the computing revolution and dirty electricity generation - and will be for some time to come still, no matter how many clean power initiatives there are.

I've been getting between 95F and 115F (actual) for the past month. That's a range of about 35-46C for y'all metric users. This week this highs are between 103 and 104 (~39C). It's basically par for the course here and it sucks so much. I'm in central Arkansas. That means the humidity is in the range of 40-80% which brings the index upwards to 100-120F (48C). A couple weeks ago it was literally 110 in the shade. That week I was doing a job for a guy from Arizona who was telling me how he was missing the cooler feeling dry 115F in Arizona compared to the wet 100F in Arkansas. Humidity is fn oppressive. At least it isn't Arabian Peninsula oppressive humidity.

Y'all Canadians can turn your fans on now. I'd really appreciate an arctic blast.

But your comment does highlight a very valid point about inefficient hardware, how many resources they use for how little they get done, and the overall effect that has on the climate and power grid. To me, it also shows that we need some sort of hardware recycling and replacement program to encourage people to upgrade for efficiency and not necessarily performance.

[Sonadow]

Microsoft drops 32-bit computers from Windows 11: "Microsoft is forcing obsolescence!"

FreeBSD drops 32-bit systems after FreeBSD 14: "This is a natural progression, old x86 systems run hot and don't have idle states."

[skeevy420]

Microsoft drops 32-bit computers from Windows 11: "Microsoft is forcing obsolescence!"

FreeBSD drops 32-bit systems after FreeBSD 14: "This is a natural progression, old x86 systems run hot and don't have idle states."

FWIW, I thought MS should have done that back when they released Windows 8. As best as I recall, it was Intel pushing those piece of crap 32-bit Atoms on us that was the reason everyone kept saying how 32bit needed to still be supported with the "Well, they're still making the hardware so we still need to provide the software." argument like Windows 7 and RHEL 6 weren't viable options. Like, sorry not sorry, not all hardware needs bleeding edge software support.

That said, if Intel or AMD actually started making modern 32-bit processors I'd feel differently. When it's basically Intel Atoms, nah.

I was one of the weirdos that was hoping that X32 would take over for desktop computing.

[Sonadow]

FWIW, I thought MS should have done that back when they released Windows 8. As best as I recall, it was Intel pushing those piece of crap 32-bit Atoms on us that was the reason everyone kept saying how 32bit needed to still be supported with the "Well, they're still making the hardware so we still need to provide the software." argument like Windows 7 and RHEL 6 weren't viable options. Like, sorry not sorry, not all hardware needs bleeding edge software support.

That said, if Intel or AMD actually started making modern 32-bit processors I'd feel differently. When it's basically Intel Atoms, nah.

I was one of the weirdos that was hoping that X32 would take over for desktop computing.

Because Intel fucked up with Bay Trail and Cherry Trail that had a 64bit architecture but stupidly required 32bit UEFI, thus Microsoft had to make Windows 8 compatible with these hardware at launch. Fortunately Intel had the sense to have their successor Braswell use a proper 64bit UEFI.

A friend of mine just recently gave me an old Bay Trail netbook and I honestly want nothing to do with it. I honestly have no love for any 32-bit OS, be it Windows or Linux. And I have a rotten experience compiling a vanilla kernel from kernel.org on a 32bit x86 processor back in 2014 where an entire portion of code ( i forgot which one) had to be commented out just to get the kernel to build on 32bit.