Ubuntu 17.10 Will Drop The 32-bit Desktop ISO

[Delgarde]

His point was about the time of announcements, they just drop stuff in the middle of things, with no warning.

It's not exactly without warning... if you search Phoronix history, you'll see that the change has been proposed several times, going back as early as 2014. The plan to drop 32-bit ISOs in 17.10 was proposed in mid-2016, so this shouldn't surprise anyone.

[R41N3R]

Removing the 32bit installer is a good decision.

[JeansenVaars]

Won't hear much about Arch and Tumbleweed because they just work, and just stay up to date. Hear a lot about distros with timed releases, because so many people have trouble at the same time as they struggle to upgrade. Also, those distros fall months and months behind on important packages, and then try to quickly catch up all at once.

Thing about timed distro is that they usually provide features of their own, see for instance Fedora, which include new developments and features in general, or Mint where the system grows along the improvements made in Cinnamon. As opposed to Arch or Debian testing where they are more pure distros and things change only when packages change.

IMHO a good balance makes it right, hence the success of Fedora, Manjaro and Mint. it also ensures a snapshots where things "just work", and most package-makers can take a release as a reference, where making such package work in rolling releases will usually require some work-arounds or soft linking to different versions of a dependency (most common in external proprietary software).

[Anvil]

Fedora should do the same, 32bit is a dying breed , RHEL dont produce 32bit, so i cant see why Fedora needs to keep it

[AndyChow]

Finally!

32Bit machines cannot handle Ubuntu as it should run IMHO. Besides, the world should be moving into 64bit a long time ago... (anyone knows where's that 128Bit computer i've been dreaming on??? ahahah)

Lubuntu will keep running on 32bit for sure (at least i hope so)! It makes total sense.

I've heard talk that the first 128-bit silicon might be produced by 2020... for riscv. You can already emulate it with riscvemu. From the little I understand, it's to have separation of memory management and memory protection, and have truly persistent pointers, potentially over networks or between exascale clusters. It won't be 128 bit flat indexing, but rather 64-bit indexing and 64-bit Object-ID. The 128-bit will be domain wide, not just local. So it can be defined (potential) on a global scale network of exascale clusters.

And those exascale clusters are going to get built real soon. At least three are planned in Europe, one in the USA (officially, I'm sure DARPA wants a few), Japan, and China.

To give an idea of the scale, Japan's goal is to have their exacomputer consume less than 30 megawatts. That's roughly the amount of energy needed to power 20 000 american residential homes.

[Gymnasiast]

Will lubuntu keep making 32 bit desktop images?

Yes. It's up to the flavours themselves to decide whether to keep making them, and Lubuntu has not announced they're dropping it. Not surprising, as Lubuntu will have more 32-bit users, and for the Lubuntu team, this move will have no impact: they built their own ISOs before and can just continue doing so. The package archives are not going away just yet.

[rockworldmi]

I think NetBSD does better 32 bit and for old system OS..

[devius]

In before someone starts moaning that they're running one of those 32-bit Atom EEE-style netbooks from 10 years ago....

I'm running one of those 32-bit Atom EEE-style netbooks from 10 years ago.... on Ubuntu 14.04. No intention of upgrading to a newer version though.

[beniwtv]

I was running a 32 bit machine until two months ago, and I can vouch for the "no effective QA testing" on 32 bit hardware.
I experienced crashes often (not as often to bother me, but they were there). Seems indeed like there is not much testing on 32 bit these days.

I ended upgrading the motherboard / CPU to 64 bit components, and all is rock solid again.

[chithanh]

Glad to see they're keeping the i686 Server images. The smaller memory footprint of i686 vs x86-64 allows me to run about 30% more VM instances.

Maybe dropping i386 will free up resources to work on x32.