HN user

mattdm

147 karma

I am not, in the developer sense, a Hacker, but I like to build things. I am currently the Fedora Project Leader.

[ my public key: https://keybase.io/mattdm; my proof: https://keybase.io/mattdm/sigs/QI11RVVFG5wScQLINySUn5P5qnYdFVCVNG9PeSxhuUc ]

Posts4
Comments93
View on HN

In Fedora, we don't "support" third-party packages or installation of software because we can't do much about it if something is wrong. You should go to the provider of the software for help.

But we certainly support your _ability_ to install and run whatever you want. It's your computer, and it's your OS.

It's not necessarily about being "one program". It's this part:

"The “Corresponding Source” for a work in object code form means all the source code needed to generate, install, and (for an executable work) run the object code and to modify the work, including scripts to control those activities."

I get that it's really hard to make money as an open source company. (That's why I am one of your paying customers.)

The exclusion you are putting on your SDK seems very similar to that of the "bitkeeper" version control software used for the Linux kernel for a short time. Look how that turned out.

ChatGPT, for all its amazingness, _never_ follows instructions. It just appears to, as it generates likely text. This is incredibly important to understand — it isn't a generalized AI.

Larger and more sophisticated models will do the party trick more convincingly... but actually following instructions will require a different approach.

There is no any guarantees Stream doesn't break ABI compatibility.

This is incorrect. Or, sure, there are no _guarantees_, but any such break would also break a future RHEL release, and is therefore a bug.

2. Five years.

The distribution needs to do almost nothing to support either EFI or the legacy BIOS or any other booting method.

I think you're seriously underestimating the amount of effort the bootloader and hardware enablement teams who work on Fedora put in to making _systems boot Linux at all_.

In complete seriousness, I think there are three broad categories:

* Use cases which actually despite the fears will be covered just fine by Stream * Use cases which will be covered by the upcoming expanded no-cost/low-cost RHEL programs * Use cases which, yeah, aren't covered

It's my estimation that the first two are actually the vast majority. I don't mean to be flippant (it was a long day). understand that people who are mostly in the third case are angry and disappointed, but from what I've seen in talking to people both when Stream was launched and after this announcement, at least some large number who are worried that that's their situation are going to be actually getting something better when the dust settles.

Okay, this is a serious question. For me, not an official RH position. In my time in HPC, nodes were baked with a specific image and then that basically never ever got updates. As I came to that as a sysadmin from other areas, I found that somewhat horrifying, but it seemed pretty universal. Have things changed such that applying patches regularly (like, more often than once a month or so except in emergencies) is a thing?

Fedora releases every six months. RHEL has a major release every three years, with minor releases every six months.

The major release of RHEL is usually forked from a Fedora version about a year before the major release, so with some exceptions software versions tend to be about a year behind those in a Fedora release initially, and then as RH backports patches rather than revving versions, drifts further.

It's the patches and smaller updates — and new hardware enablement in the full-support phase — that will be landing in CentOS Stream. Previously, this stuff was developed internally and only released to the world at the minor-release drop every six months. Now, it'll be developed in a shared space. So what you'll have in CentOS Stream is whatever is intended to ship in a RHEL minor update within six months.

There may occasionally be a situation where an update introduces bugs into Stream that would have been caught by QA before a public minor RHEL release and subsequent CentOS Linux rebuild. But I don't actually expect that to happen enough to worry about in any case where you can justify not paying for actual supported RHEL in the first place.

Fedora doesn't like to be put in that particular niche anyway. Yes, Fedora is fast moving and it is what Red Hat uses as the base for major releases, but we're more than "Redhat-unstable" in so many ways.

CentOS Stream will be the upstream for RHEL minor branch development. This is actually a huge thing that I think people are missing: previously, once branched from Fedora, RHEL development did not happen in the public eye. With Stream, it will. This is huge and awesome good news. However, that development will still be entirely Red Hat curated. This is different from Fedora, where we make community decisions with Red Hat's engineering input as a stakeholder but not the decider. (See for example btrfs as the default filesystem.)

I think there might be some confusion about "rolling" in the context of CentOS Stream. The updates are continuous and there's not released minor versions, but all changes are changes that are intended to _very shortly_ land in the next every-six-months minor release of RHEL. So you aren't going to suddenly see more "environment changing out from under me" than you would on RHEL or CentOS Linux.

I'm a little confused by this. Well, first, I'd encourage you to try Fedora Workstation for your family -- we've worked on making upgrades painless, so that they're basically an automated thing that happens while you go for coffee once or twice a year (at your option). But second, if a "Fedora LTS" would fit your needs, why not give CentOS Stream a look? It's not actually going to be that different from CentOS Linux, and almost certainly will have less constant change than a theoretical Fedora LTS would.

Also, I am not one of the highest upptity-ups in the company or anything, but from the inside: I see no evidence whatsoever that this is the result of IBM anything.

And this will be expanding to cover more use cases. See https://www.redhat.com/en/blog/faq-centos-stream-updates#Q10:

_In the first half of 2021, we will be introducing low- or no-cost programs for a variety of use cases, including options for open source projects and communities, partner ecosystems and an expansion of the use cases of the Red Hat Enterprise Linux Developer subscription to better serve the needs of systems administrators and partner developers._

Yes, this is exactly it. The conspiracy theories about buying-CentOS-to-kill it are understandable but totally off-base. Red Hat brought CentOS in-house at a time when the company was trying to grow from being a single-product company to a portfolio one, and it became clear that Fedora wasn't working for what at the time RH called "layered products". The hope was that CentOS would provide a more-RHEL-like community place for work like RDO to happen. That was partially successful, but the plan wasn't really realized — and speaking from a Fedora point of view, that "Fedora is failing at a thing we need so we'll turn to CentOS" wasn't a healthy dynamic for either project. CentOS Stream serves the initial purpose better, and now RH's distro ecosystem story is actually linear rather than a crazy MC Escher contortion.