Ubuntu 22.04 LTS To Shift Its PPC64EL Baseline To POWER9 CPUs, Dropping POWER8

[tildearrow]

ppc64el just means little-endian byte ordering. ppc64 is big-endian. There are, to my knowledge, no PPC64 devices with no floating point.

armel originally just was to distinguish from the long-dead big-endian ARM port, "armeb", at a time when neither had floating-point. "el" has nothing to do with FP on its own - "armhf" was just newer and got its own name as a result.

That's odd because it would then be read as "endian little". Are you sure it isn't supposed to be ppc64le?

But if "el" in armel indeed means "little endian" then I mean I thought "el" meant "EmuLated (floating point ops)"...

[mroche]

tildearrow It's just a chosen naming convention:

https://lists.debian.org/debian-powe.../msg00040.html

https://wiki.debian.org/ppc64el

https://lists.ozlabs.org/pipermail/l...er/120650.html

ppc64le and ppc64el refer to the same platform as far as I'm aware. PowerPC (or ppc64) is the big-endian platform.

Cheers,
Mike

[chithanh]

only those contributing to Canonical's support by doing the (hard) work of full development and QA, or paying for that support, should expect Canonical to give a darn.

not everyone likes them but the become necessary as time moves on depending on install base trends, performance and optimization, and support burdens. These trends come from existing customers and potential business opportunities, not the home/after market which is relatively trivial in comparison and not either company's business focus.

That is a very short-sighted approach. Sure those users do not directly add to the bottom line and are therefore often overlooked by corporate beancounters. But they are important gatekeepers and multiplicators.

If you want an enterprise deployment without pushback from your admin staff then you choose technology which they are already familiar with. If you want software to work on your hardware, then you need the developers to dogfood it. AMD is currently learning that lesson, and I guess IBM will learn this too now thanks to Canonical's move.

Owners of Power8 hardware can still use a supported version of an Enterprise Linux family distribution or Ubuntu until those points in time, just not the newest and shiniest.

While home lab operators are willing to compromise on getting newest and shiniest hardware due to cost, they will not compromise on the software side. There is no point for them in sticking with old software. Once the support from most recent release is gone they will switch away.

[King InuYasha]

RHEL 8* and derivative projects choosing to match will be supported until end of May 2029 (2031 for ELS subscribers). Customers of Canonical's ESM program will have until (end of?) April 2030 with Ubuntu 20.04 and non-customers until 2025, should this change not be reverted. Owners of Power8 hardware can still use a supported version of an Enterprise Linux family distribution or Ubuntu until those points in time, just not the newest and shiniest. It's not as if when RHEL 9/Ubuntu 22.04 comes out Power8 support is going to vanish. Hardware cycles are a known occurrence in the enterprise space; not everyone likes them but the become necessary as time moves on depending on install base trends, performance and optimization, and support burdens. These trends come from existing customers and potential business opportunities, not the home/after market which is relatively trivial in comparison and not either company's business focus.

Enterprise groups do _not_ move quickly from the foundational software perspective, newer hardware is often purchased well before operating system major upgrades occur. And they definitely don't jump when a version is first released, often 2-3+ years into its cycle. That hardware won't go bust when it enters the aftermarket, currently existing distribution versions will still handle it fine.

* Though oddly the ppc64le build is not available for download in my personal account using the Developer Subscription for Individuals. I might post this internally next week and see if anyone bites. The RHEL 9 beta version is, but that doesn't help this conversation.

Cheers,
Mike

It also might be worth noting that Raptor Systems' Talos II computers are POWER 9 systems. For most people, POWER 9 is easier to acquire than POWER 8, and with POWER 10 launching for enterprises this year, a lot of POWER 9 enterprise systems will be making its way to the secondary markets over the next couple of years, which will make it a lot easier for people to pick it up.

Would I have preferred it if POWER 8 was supported a bit longer? Maybe. It's still supported by Fedora, which is commonly used by hobbyists too.

It would be nice to have RHEL for POWER available in the dev sub for individuals because of the availability of Talos systems for individuals, too.

[Dawn]

That's odd because it would then be read as "endian little". Are you sure it isn't supposed to be ppc64le?

But if "el" in armel indeed means "little endian" then I mean I thought "el" meant "EmuLated (floating point ops)"...

"el" is "le" written in little-endian order. It has never stood for "emulated," regardless of what you think.

https://unix.stackexchange.com/quest...and-for/254892 has some discussion of the matter.

[spanky]

This is a fine example of why Linux users cannot have nice things. Linux is like 1% of the desktop market and let's say 90% of the server market. Of that desktop market, PowerPC is probably 0.000001% of that, and probably not much better in the server market. Of those markets, how much market share does Ubuntu have, and how much resources do they devote to supporting it? Yet, complete and total outrage that aging, obscure hardware will be expected to stick with an aging LTS release. Demanding real company and community resources be diverted from supporting features used by most users, to support something that close to zero people use.

Just like when a desktop ISV (proprietary or open source) offers their WIndows/Mac software on Linux via a deb and rpm package, they are immediately flooded with support requests of "power users" trying (and failing) to use it on some obscure hobbyist-grade distro that doesn't even work properly on a good day. As if the (probably small) ISV should devote massive resources to support all 50,000 Linux distros (when it is almost certainly the distro's fault it doesn't work) to fully address that potential 1% of the market, when about 49,980 of those distros don't even work correctly in the first place. Even the well supported, well funded distros are hard enough to support, given that they too are still often broken, especially when most people are going to use forks of these distros that add their own incompatibilities. Then there is the bleeding edge distros, where the assumption that foundational packages like the kernel, glibc, gcc, clang don't have bugs when they are new, and that it is the downstream developer's fault if their software doesn't work on it.

Yet, these universal package formats (like Snap) are a solution to this problem, even if none of them are ideal, yet Canonical is widely slandered for using Snap. The biggest beneficiary of these package formats is your obscure toy distro that might actually be able run large, complex desktop software. But then it would be "too mainstream", and you would have to switch distros. Supporitng Linux offers a virtually unlimited support burden in exchange for potentially 1% of the market who will fight any attempts to be monetized in any way, who won't even let you send anonymous telemetry to fix bugs, who will complain about literally anything you do.

[andre30correia]

they continue support until 2030? is more than time small enterprise can have the canonical extension for free

[ElectricPrism]

spanky -- That was some methed-up Hitler level rage-posting you got there. The Strawman of 50,000 linux distros, slippery sloping into theoretical infinite support, blame & income, complaining about what hobbiests and Information Scientists.

And then ADHD dick sucking Canonical -- and here's the Hyperactive part

Yet, these universal package formats (like Snap) are a solution to this problem, even if none of them are ideal, yet Canonical is widely slandered for using Snap.

Slandered? Really?

slander
slăn′dər
noun
- Oral communication of false and malicious statements that damage the reputation of another.
- A false and malicious statement or report about someone.

Let's start over.

1. "50,000 distros". ( If you are targeting more than 1 to 3 distro families you are doing it wrong. You can immediately delete 99.9% of that list. )

• Game-devs should target SteamOS (Arch),
• Servers should target RHEL (CentOS) and Debian (Ubuntu),
• Desktop should target Ubuntu (Mint), Fedora (RHEL) and maybe OpenSUSE. (Or they can just target [1] Flatpak [which is universal] and optionally [2] Snap [which is ubuntu only])
• Depending on the software that's on average 1 or 2, sometimes 3 if you want to cast your dragnet wide.

2. As per your theoretical support paradox, game-devs already work around this by simply targeting SteamOS, the scientific community generally targets a single distro, and the video editing industry does the same -- they used to target CentOS for example.

3. Blame & Income -- Linux professionals spend lots of money on equipment, software, support, get over your fantasy that somehow the most used software on the planet is some kind of "hobbiest" thing "wit no muns" just because it also appeals to hobbiests, and people interested in learning.

As for blaming those hobbiests for "not wanting the windows way" of "snaps" or whatever, you can have all the snaps you want for all I care. Have you even read and critically thought and reasoned out the criticisms of Canonical with their NIH, and Snaps? or the Criticisms of the centralized nature, technical flaws, and pushing back against the greater community? I doubt you have.

4. Complaining that a hobbiest working on a Rasberry Pi is somehow hurting development of Ubuntu is very naive, it'd be like complaining that Muh Ford Pickup isn't getting to be popular because some engineer is building Drone technology.

This is what you do when you try to complain that X Linux is different than G Linux, or "not necessary" -- maybe for you personally, or "in your opinion".

In the same vein, complaining about the diversity of Linux would be like complaining that Honda, Mercedes, Tesla, Ford, Acura, Lexus, Chevy all exist and then arguing that why can't we have 1 homogenized crappy car/truck that compromises and sucks at everything instead of diversity with strengths. I'm not as adept at cars, but if I recall correctly places in Asia have their homogenized state-made average commie cars and they are often underwhelming and poor quality but keep the state factories employing people.

We get it though, to a noob with ADHD, it's difficult to navigate and fully understand the differences. So put on your blinders and settle down -- ignore what hobbiests are doing, also -- like -- legit, consider chilling out and doing something to improve self-wellness, these are hard crazy times and most of us have been the rager before, but maybe go exercise, improve diet, sleep or even unplug a day or two if you are overstimulated by the BS of the net.

[ElectricPrism]

Back to my intent to comment on the article:

Remember when Canonical dropped 32-bit support and blindsided Valve who recommended Ubuntu to their customers? Dropping Power8 reminds me of that. In the history of Canonical, they have this kind of "We'll do what we want, when we want, and not communicate with others" attitude, (I partially blame their geographical location being the UK and not California and thus the timezone, people who coordinate with workmates in other timezones know what I am talking about)

I have watched this the last 10 years on Phoronix:

• making things like Upstart, then nuking it,
• Ubuntu Phone got millions (18?) in funding and then canned,
• then it was Mir display protocol which had the goal of bridging Android GPU binary blob Drivers to Linux only to get axed,
• Ubuntu Touch was a huge deal then thrown out,
• Unity Desktop,
• etc.. etc..

I deeply believe that Canonical underestimates the value of the gatekeeper admin, as per other comments in this thread -- those users will re-use old hardware and then deploy technology in their enterprises they are already skilled at. Canonical is making a grave mistake to give up their "All Roads Lead to Rome" seat -- it's no wonder they are no longer the center of gravity and not viewed as favorably.

Also, to those whining and complaining in this thread that Canonical is somehow in danger of going out of businesses -- is that really valid or just a projection of our own hard economic situations on to them? Because last time I checked the enterprises and wealthy were doing amazing and it was the plebs that weren't.

billionaire-graph.png

[jo-erlend]

So those who have three or fewer POWER8 systems running, will have 8 years to figure out what to do. Those who have more than three will have to choose between either finding an alternative in three years or paying a little extra. How popular will these systems be after 2030 anyway?

It seems to me that people are looking for things to be angry about.