Carl George :fedora: :centos:

#RPM packaging nerd. Works on #EPEL, #Fedora, and #CentOS. Texas forever.

Public Key

npub1awkqz8sdqd5kqx5q8l6rc07gxgdv64hyc0v9frgwvssdntgxtjuq6jshpr

NIP-05 Address

Profile Code

nprofile1qqswhtqprcxsx6tqr2qrlapu8lyryxkd2mjv8kz5358xggxe45r9ewqpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0ds070syg

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1awkqz8sdqd5kqx5q8l6rc07gxgdv64hyc0v9frgwvssdntgxtjuq6jshpr

Show more details

Last Notes

2024-07-06 19:06:38

- reply

@npub1upk…tcwy I saw other replies about ANSI escape codes to do something similar, but Tilix's approach doesn't require individual apps to implement support for it. It just works with any text inside the terminal.

2024-06-25 03:08:47

Hot take, apparently: "patches welcome" or "pull requests welcome" is just fine as a response to complaints. It's a gentle reminder of how open source works. If you are offended by it, that says more about you than the person saying it.

2024-05-23 20:37:57

- reply

Just to be clear, I've been on the committee since 2020. Historically the committee seats have been appointed, but we decided to implement proper elections going forward. About half the committee including myself volunteered to step down and effectively run for re-election.
https://docs.fedoraproject.org/en-US/epel/epel-policy-steering-committee/

2024-05-23 20:35:30

I'm running for the EPEL Steering Committee. If I've ever helped you with EPEL, or you've enjoyed my conference presentations about EPEL, please consider voting for me.
https://communityblog.fedoraproject.org/epel-steering-committee-election-interview-with-carl-george-carlwgeorge/
#epel #fedora #centos #rhel

2024-04-23 23:15:08

- reply

@npub1lny…65xd Hey buddy, how you doing

2024-04-01 05:12:45

In case you haven't heard, @npub1um6…atg0 is back this year, April 12-13 in Austin! We took a long hiatus during COVID, but everyone is excited to bring the show back. Get ready for two days of nerding out on Linux and open source topics.
https://2024.texaslinuxfest.org
#TXLF

2024-03-21 17:04:55

- reply

@npub1tny…z6px The asymmetrical aspect of the SSPL is what bothers me the most. That license is extremely viral, but companies choosing it exclude themselves from those viral requirements. At its core that license is designed to provide an asymmetric advantage to a single company.

2024-03-19 14:39:43

Cursed random morning thought: when sorted alphabetically, the word "after" is before the word "before".

2024-03-13 19:59:42

- reply

@npub1pxy…htpa I didn't claim that's what it meant, I just said it was in conflict with being five months behind. I understand the volunteer aspect of open source, many of the packages I maintain are purely on a volunteer basis and have nothing to do with work. I was asking to learn more about the differences in how Python is handled between the distros, not to shame anyone. Because I asked, I got a reminder about zombie-imp (which I had forgotten about) and also learned about the Arch todo page.

2024-03-13 04:20:17

- reply

Debian Sid and openSUSE Tumbleweed both still use 3.11 by default, but at least have 3.12 available as an alternative in the default repos. Fedora switched to 3.12 as the default in Fedora 39. Is there something I'm missing about Python 3.12 that makes it harder for distros to adopt?

2024-03-13 04:11:53

Does anyone know what's going on with Python in Arch Linux? 3.12 was released over five months ago, but Arch is still using 3.11. This doesn't affect me directly, I just noticed and I'm curious. I found an issue about it, but it was closed with no answer.
https://gitlab.archlinux.org/archlinux/packaging/packages/python/-/issues/8
#archlinux #python

2024-03-12 22:36:52

- reply

@npub1tuv…5jr2 @npub1hlt…ck4e Awesome, I believe you just volunteered for booth duty. Hey @npub1hr7…2z4y, let's get Jim on the roster.

2024-03-05 19:38:51

- reply

@npub1nve…qvau @npub1um6…atg0 Our site theme doesn't seem to handle narrow resolutions like mobile phones well. We're working on some theme adjustments to use a grid layout which may help. In the mean time, try viewing the schedule from a desktop.
https://cdn.fosstodon.org/media_attachments/files/112/044/531/416/232/499/original/0aaf2688c095c175.png

2024-03-04 19:02:16

- reply

@npub13dg…ejem The minimum required RAM for EL7 is 1 GB, and for EL8/9 it's 1.5 GB, so yeah I'd expect only using 0.5 GB will eventually bite you in the rear. Getting OOMs during a leapp upgrade definitely sounds like the worst possible time for that biting to happen.
https://access.redhat.com/articles/rhel-limits#minimum-required-memory-3

2024-02-19 17:37:25

- reply

@npub1hex…x0cp @npub1sfq…dq36
> Red Hat "lifers" should label their posts accordingly.
Why is it members of the Rocky community routinely jump straight to personal attacks rather than addressing or even considering the substance of a criticism?

2024-02-19 17:36:20

- reply

@npub1hex…x0cp @npub1sfq…dq36
> As you know but take pains to hide
Brendan isn't the author, he's just sharing an article he enjoyed.
> There is little difference between the approaches applied to this and Red Hat's approach to Cent0S stream
Which is the whole point of the article. Did you read the whole thing?
> except that the Recky versions are more stable and accessible.
False, Rocky can't accept pull requests that change the OS, therefore it's less accessible and less collaborative.

2024-02-19 07:18:30

- reply

@npub1vp0…kkyd I'm 100% certain that losing their marbles (hehe) is happening in the near future.

2024-02-19 01:55:06

- reply

@npub18tj…9qfz My kids wanted it, and I like putting smiles on there faces 😁
I figured "it can't be THAT difficult to get out". It sure put up a fight, but in the end my Dremel won and no one got cut by glass.

2024-02-19 01:45:34

Tonight, I get to be a hero to my kids because I figured out how to reach the forbidden marble.
https://cdn.fosstodon.org/media_attachments/files/111/955/367/975/755/035/original/7dc7702e494c1bb0.jpg

2024-02-12 08:41:49

I got this awesome hoodie at the FOSDEM Fringe CentOS Connect event. Haven't you heard? CentOS is the source.
#CentOS #RHEL #RedHat #FOSDEM
https://cdn.fosstodon.org/media_attachments/files/111/917/319/827/059/506/original/627f21445352a9f0.jpg

2024-01-04 07:27:25

- reply

@npub1rg5…tdr8 @npub1mjv…k4w8 @npub12h3…9qyg Bundling isn't a free pass on doing work though, as there is other work involved to bundle correctly. Whether bundling or not, someone has to do the work, which is why it's better to get involved directly to make things happen.

2024-01-04 07:26:54

- reply

@npub1rg5…tdr8 @npub1mjv…k4w8 @npub12h3…9qyg So your complaint isn't actually about language servers, it's that software in general has too many dependencies? That's on the upstream software projects, not Fedora. Fedora policy doesn't force you to personally package every dep. For python, rust, and go, most common deps are already packaged. If not all deps are already packaged, you're allowed to bundle. For node packages, it's actually required to bundle.

2024-01-04 06:58:19

- reply

@npub1rg5…tdr8 @npub1mjv…k4w8 @npub12h3…9qyg Have you considered packaging the language servers you care about using and submitting them for inclusion in the #Fedora repos? Packages don't exist by happenstance, someone has to add them and maintain them. Fedora's philosophy is that it is *our* operating system that we build together.
https://docs.fedoraproject.org/en-US/package-maintainers/New_Package_Process_for_New_Contributors/

2023-11-24 16:33:08

- reply

@npub1rgx…6y88 Have you seen the sh library? It has basically the API you described, and let's you run any command that way. It's still shelling out, but it's useful until the underlying program implements a proper API.
https://sh.readthedocs.io

2023-11-24 16:10:14

- reply

@npub1zgy…u85a The most likely reason is that Thinkpad's CPU doesn't support the x86-64-v2 baseline required by RHEL 9. This was proposed and rejected in Fedora.
https://fedoraproject.org/wiki/Changes/x86-64_micro-architecture_update
https://pagure.io/fesco/issue/2198
https://developers.redhat.com/blog/2021/01/05/building-red-hat-enterprise-linux-9-for-the-x86-64-v2-microarchitecture-level

2023-11-16 17:00:24

- reply

@npub1mjf…p676 @[email protected] @npub1eqg…fncn Even better, he replied with some "so you admit..." strawman bullshit and then blocked me. It's always a gamble trying to talk to people like that. Sometimes they listen to reason and facts, and sometimes they just reveal they were arguing in bad faith all along.

2023-11-16 07:19:05

- reply

@npub1rxj…q8q6 @npub1eqg…fncn You fell for the FUD from other distros trying to disparage the project.
CentOS is not dead, it's alive and well. Stream isn't a different distro, it's an improved variant that can accept contributions (the classic variant couldn't). Within each major version it's 90-95% the same software versions as RHEL at any given time. That's because it's the next minor version of RHEL, which has to stay highly compatible.

2023-11-16 06:57:09

- reply

@npub1rxj…q8q6 @npub1eqg…fncn Because of the changes to CentOS, RHEL is _more_ open than it has ever been before.

2023-11-16 06:48:08

- reply

@npub1a2e…57lf RHEL 9.3 was released on 2023-11-07, six days before Alma 9.3. Not sure why Distrowatch has the wrong date.
https://access.redhat.com/articles/3078#RHEL9

2023-11-09 00:03:39

- reply

@npub1zlr…pr7v I asked around in the CentOS Promo SIG, and no one thinks it's intentionally being held back. It's just a matter of someone updating the logos repo with the new images, which is being tracked here:
https://github.com/CentOS/centos-logos/issues/4

2023-11-01 13:28:34

- reply

@npub1rnj…vtmy I think Lenovo refers to them as "preloads", so maybe #FedoraPreload or #FedoraPreloaded?

2023-10-31 16:00:07

- reply

@npub1djf…klr2 I noticed the same thing. It matches what is in bodhi, so it seems to be waiting on the maintainer to submit an additional build.
https://bodhi.fedoraproject.org/updates/?packages=firefox

2023-10-31 06:27:51

- reply

@npub1tdp…w6tq This you?
https://cdn.fosstodon.org/media_attachments/files/111/327/562/557/622/200/original/ee5fe2b5d5495c3e.png
https://cdn.fosstodon.org/media_attachments/files/111/327/563/192/580/951/original/40f0339cc79e24e0.png
https://cdn.fosstodon.org/media_attachments/files/111/327/563/530/602/285/original/1d679d7eee4cd871.jpg

2023-10-31 01:33:07

- reply

@npub1q0y…wk42 Yes that should work fine. Regardless of spin or edition, packagers are supposed to ensure the upgrade path of their packages over at least two releases.

2023-10-30 19:02:33

Here's a tip for those looking to upgrade to Fedora 39 before the official release: enable the updates-testing repo first. Due an extended release freeze, there are quite a few updates stacking up in the updates-testing repo that normally would have been pushed to stable by now. One of these updates was even necessary for me to complete my upgrade transaction.
Also you may want to set a reminder to turn updates-testing back off after the final release.
#fedora

2023-10-25 19:55:18

- reply

@npub1ryg…uwgd Yeah those particular docs are targeted specifically at golang packages in the official Fedora repos, certainly not ideal as a starting point. If you're new, I recommend https://rpm-packaging-guide.github.io/ instead (which is linked to from https://rpm.org/documentation.html). I also created https://red.ht/rpm as an interactive lab if that is more your learning style.

2023-10-25 18:50:28

- reply

The recording for this talk is now available.
#LFNW #EPEL #CentOS #RHEL
https://youtu.be/-7VwIV-Qnww?si=k7a6LP1veyVLhvid

2023-10-21 23:32:21

- reply

@npub1ryg…uwgd Can you tell me more about the problems you ran into with spec file macros? Did you have problems finding the documentation? Or did you find the docs but they were confusing?

2023-10-15 04:13:33

A little while back I posted about doing both a talk and workshop at LinuxFest Northwest this year. As you may have heard, the conference had their original venue cancel on them and had to downsize. The new venue didn't have room for any workshops, but my regular talk is still happening. Check it out to hear about our journey to EPEL 9, with a sneak peek of EPEL 10.
#LFNW #EPEL #CentOS #RHEL
https://lfnw2023.sessionize.com/session/494071
https://cdn.fosstodon.org/media_attachments/files/111/236/575/755/544/040/original/33cd50b7ffccc66e.png

2023-10-14 17:46:50

- reply

@npub1e2p…aqmc Nothing wrong with playing slightly older games, they tend to work better on Linux anyways. I just bought and started playing Jedi Fallen Order this week.

2023-09-24 22:54:13

I dedicate this meme to @npub1usy…5y4m's herculean efforts maintaining texlive in #fedora. :fedora:
https://cdn.fosstodon.org/media_attachments/files/111/122/073/517/481/507/original/b1dd2ede32ec2f47.jpg

2023-08-31 03:08:19

- reply

@npub1tnk…j0hy @npub1zlr…pr7v Thanks y'all for adding comments to that bug with extra details. Until it's resolved I suggest configuring your system to default to a previous kernel such as 5.14.0-354.el9.
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/9/html/managing_monitoring_and_updating_the_kernel/assembly_the-linux-kernel_managing-monitoring-and-updating-the-kernel#setting-a-kernel-as-default_assembly_the-linux-kernel

2023-08-30 18:34:43

- reply

@npub1zlr…pr7v Which kernel version-release is that happening on? @npub1tnk…j0hy mentioned seeing something similar in IRC earlier today with kernel-5.14.0-361.el9. Narrowing down exactly which kernels are affected would be really useful for a bug report.

2023-08-30 14:31:54

- reply

@npub17kf…hsl3 Inside the container you can run `dnf config-manager --set-enabled <repoid>` or just edit `/etc/yum.repos.d/redhat.repo` directly.

2023-08-11 04:20:53

I'll be honest, seeing CIQ and SUSE *voluntarily* associate themselves with Oracle puts a big ol' grin on my face.

2023-07-20 04:55:45

- reply

@npub15hl…3knw @npub1jrp…xxnn @npub1hkw…jkge RHEL 8 has PHP 7.2 as default, but also has optional 7.3, 7.4, and 8.0. It doesn't make sense to compare the default PHP version of a distro released in 2019 with the default PHP version of a distro released in 2022. RHEL 9, which was released in 2022, has PHP 8.0 as the default, with optional 8.1.
Alma could ship additional optional PHP versions, but changing the default PHP version would break the ABI compatibility they're aiming for.

2023-07-20 04:55:19

- reply

@npub15hl…3knw @npub1jrp…xxnn @npub1hkw…jkge RHEL customers asking for this because Alma has it would be a great outcome. The maintainer specifically mentioned customer feedback/demand. But really it's probably going to depend more on the severity rating it ends up with (it doesn't have one yet).

2023-07-20 04:46:44

- reply

@npub1jrp…xxnn On the third point, it's also no. Things are only merged into CentOS Stream if they are planned for the next minor version of the corresponding RHEL major version. Merging things that aren't intended for RHEL would be against what we've described CS as. It's not a sandbox to throw out new features/fixes to see if they work. It's the staging area for QA-tested, RHEL-compatible updates that are approved for the next minor version.

2023-07-20 04:41:58

- reply

@npub1jrp…xxnn @npub1hkw…jkge
Merge decisions for CentOS Stream contributions are made by Red Hat. CentOS Stream is still more of a community-oriented distro than CentOS Linux. Both of these things are true.
I can't help but feel the "shouted down by the centos-stream advocates" is directed at me. If I've ever made you feel "shouted down", I truly apologize. I'm a believer in the strategic direction of the CentOS project. The execution is a work in progress.

2023-07-20 04:41:01

- reply

@npub1jrp…xxnn @npub1hkw…jkge Your "not interested in this fix" comment is leaving out the "at this time" and "keep it open for evaluation" parts that were actually said. At least you didn't crop those parts out like the person who posted this on Reddit did. The CVE isn't even rated yet, so let's see how that turns out before rushing to judgement. Not every CVE gets fixed in RHEL. The severity rating and customer demand will influence this being merged.

2023-07-14 09:10:53

- reply

@npub19xr…gkcq So that's another issue you could file. Compare and contrast how in Fedora you can look at dist-git and bodhi and see the contribution path, but looking at the CentOS dist-git makes you feel lost. That's a valid observation and one that can hopefully be addressed via the contribution guide. Start the conversation in an issue and see where it goes.

2023-07-14 09:05:52

- reply

@npub19xr…gkcq You can certainly suggest improvements. You just helped identify that the docs mention GitLab but don't explain how to find the right repo within the linked namespace. I'm just asking you to file that as an issue.
These docs are a contribution guide. The goal is to guide people to contribute. The audience is people who want to contribute. If you're not interested in contributing, then sure they're "not for you". But that's not the same thing as "unwelcoming to contributors".

2023-07-14 08:43:52

- reply

@npub19xr…gkcq If you have suggestions for improving this page, the source is here and pull requests are welcome.
https://github.com/CentOS/docs-contributors-guide/blob/master/modules/ROOT/pages/quickstart.adoc
Alternatively, just enumerating your concerns as issues in that repo would be helpful.

2023-07-14 08:43:29

- reply

@npub19xr…gkcq I don't know what's going on with the "waiting for content" pages, but this page you linked is the meat of the process.
It doesn't assume you know where bug reports go, it links you directly to bugzilla in the "file a bugzilla" section, with pre-filled fields to categorize the report correctly.
For patches, the "open a merge request" section links to the GitLab namespace. This should probably be expanded to explain how to find the correct repos within that namespace.

2023-07-14 08:05:36

- reply

@npub19xr…gkcq Do you have specific feedback about what in that doc you see as unwelcoming? Do you have a problem with the steps of the process or with the documentation itself? We've worked hard to build out a contribution path that meets our requirements, but are happy to discuss ways to reduce friction points further. On the scale of RHEL lifecycles this is all still very new.

2023-07-14 06:54:23

- reply

@npub1nz7…5q62 RH decided to turn that tooling off because we're not using it anymore, while at the same time encouraging derivatives to get their sources from CentOS Stream, the same way we do to build actual RHEL.
I'm of course biased here, but I think the only thing that's really justified for folks to be mad about is the suddenness of the legacy tooling being turned off. Or the fact that it was allowed to start for RHEL 9 at all.

2023-07-14 06:51:53

- reply

@npub1nz7…5q62 With CentOS Stream 9 we actually built RHEL fully in the public for the first time ever, in collaboration with the community. Sources are now public in a way that facilities contributions, real git repos that you can send pull requests to. This made the extract/export tooling obsolete. However, this legacy tooling was left in place for longer than it should have been, and the new rebuild distros set up their pipelines to depend on it.

2023-07-14 06:50:45

- reply

@npub1nz7…5q62 Originally it was a dump of SRPMs to ftp.redhat.com, the dubious "private development, then throw it over the wall" model. For RHEL 7 this was adjusted in collaboration with CentOS to extract the SRPM contents and push them in big obscure commits to git.centos.org (with some limited automatic debranding). This was still fully private development with no contribution path.

2023-07-14 06:49:29

- reply

@npub1nz7…5q62 Not quite right, no new license or subscription terms were introduced. The no redistribution clause has always been there for RHEL binaries and sources obtained via subscription. The minimum requirement of some of the licenses involved is to provide sources to those you provide binaries to. RH went beyond that by posting the sources for the latest minor versions of RHEL (not older EUS or ELS versions) in a few additional ways, in public and not bound by the no redistribution clause.

2023-07-14 06:14:02

- reply

@npub199l…ljf5 When that comes up inside Red Hat I like to tell folks that we have no standing to correct people on it until we rename the redhat-release package to red-hat-release.

2023-07-14 06:07:39

- reply

@npub1wmh…rv0p @npub1hkw…jkge Nah you've got it all wrong. Us Red Hatters are cheering on this move. It's exactly what we've been asking derivatives to do.

2023-07-14 05:46:18

- reply

@npub1hzg…zc23 Perhaps this one from @npub1tjg…gye3?
https://quantum-integration.org/the-curse-of-bug-to-bug-compatibility

2023-07-14 00:51:35

- reply

@npub129r…0vg0 @npub167l…2g0n Agreed. Bug-for-bug compatibility isn't a valuable thing to pursue. The Alma folks get it now. Now CentOS Stream and Alma will be the only Enterprise Linux distros that can accept contributions that change the OS.

2023-07-14 00:36:08

> We will also start asking anyone who reports bugs in AlmaLinux OS to attempt to test and replicate the problem in CentOS Stream as well, so we can focus our energy on correcting it in the right place.
👏
#CentOS #RHEL #AlmaLinux
https://almalinux.org/blog/future-of-almalinux/

2023-07-12 08:53:49

- reply

@npub15zz…q2s8 Correct, subscription-manager doesn't have a dnf5 plugin yet, so if you're using that already then you can avoid it. At some point that will change as they have to integrate it into dnf5 eventually (my guess, by Fedora 40).

2023-07-12 08:15:44

- reply

@npub1wls…fdfu The dev sub for individuals wouldn't be a good fit for that anyways. Projects that do self-hosted CI should look into the ROSI (RHEL for Open Source Infrastructure) program. For shared CI providers like GitHub, it's an ongoing discussion figuring out how to make RHEL available. In the meantime, CentOS Stream works well for that use case.

2023-07-12 08:08:23

- reply

@npub1wls…fdfu It's not surprising that it would gather that data for support reasons, considering its origins of subscribing supported RHEL systems directly. I agree that data is unnecessary on unsupported Fedora systems. Perhaps if you filed a bug the dev team would consider an off switch?
I'm just trying to show folks how easy it is these days to consume real RHEL in mock and podman. If you don't want to, then don't.

2023-07-12 06:32:48

- reply

@npub1u7f…fc7j I've never heard of IBM auditing Red Hat customers. Sounds a bit far fetched.
Would it help convince them if you told them that the devsub for teams gave them up to 25,000 RHEL instances for free?

2023-07-12 04:51:20

- reply

This was really easy y'all. If you were previously put off by frustrations with subscription-manager, give it another shot.

2023-07-12 04:50:09

- reply

dnf commands on my workstation were slightly slower afterwards due to the dnf subscription-manager plugin. Since I don't need any subscription repos directly on my workstation, I disabled the plugin and the subscription-manager setting that would re-enable it.
sudo subscription-manager config --rhsm.auto_enable_yum_plugins 0
sudo sed -e '/^enabled=/ s/1/0/' -i /etc/dnf/plugins/subscription-manager.conf
Disabling this plugin doesn't affect the subscription passthrough to mock or podman.

2023-07-12 04:49:18

- reply

My main use cases for RHEL are building packages with mock and running repoquery commands inside local podman containers. How easy is it to use actual RHEL for these use cases? It's literally one command.
sudo subscription-manager register
I ran that on my Fedora workstation, it prompted me for my username and password, and I was done. RHEL mock chroots just worked. Podman UBI containers just worked, with access to the full RHEL repos, not just the limited UBI repos.

2023-07-12 04:48:18

I regularly see folks state that they use a RHEL rebuild distro because managing RHEL subscriptions is too painful. 5 years or so ago, I might have been inclined to agree with them. That said, things can improve over time, so let's check out the current state. The most notable improvement in the last few years is a thing called Simple Content Access (SCA).
#RHEL #RedHat
https://access.redhat.com/articles/simple-content-access

2023-07-12 03:27:56

- reply

@npub129r…0vg0 @npub167l…2g0n I agree with Mike. I'll also point to @npub1pzn…y2as's most recent article as a good reflection of my own feelings on those announcements.
https://dissociatedpress.net/2023/07/11/red-hat-and-the-clone-wars-iv-knives-out/

2023-07-12 00:24:39

- reply

@npub1fw3…q5nd @npub1352…c7qs
> stream is the stuff RHEL might want & is testing
More than just "might want", it's literally the stuff that is planned for the next minor version of RHEL. Things are tested before they're merged and released, not after.

2023-07-12 00:20:08

- reply

@npub1352…c7qs I have a pinned thread that has some diagrams that might help clear things up.
https://fosstodon.org/@carlwgeorge/109985597904896856

2023-07-12 00:18:33

- reply

@npub1352…c7qs
> Can I just install "Stream 9" and know it will get updates for X years?
Yes, at least 5 years, corresponding to the RHEL "full support" phase. CS9 came out about 6 months before RHEL, so it will have about a 5.5 year lifecycle. That's what I'd expect for future versions as well.
https://access.redhat.com/support/policy/updates/errata#Full_Support_Phase

2023-07-12 00:18:04

- reply

@npub1352…c7qs
> I also can't figure out if updates from RHEL come _from_ Stream or are pushed _to_ Stream.
They're maintained as separate minor version branches. Right now Stream is effectively 9.3, while RHEL is on 9.2. Most updates happen in Stream first, won't get put in RHEL 9.2, but will be part of RHEL 9.3 when it is released. Some updates happen in both 9.2 and 9.3, usually near the same time but sometimes based on how something is implemented it might land in 9.2 slightly before 9.3.

2023-07-12 00:16:46

- reply

@npub1352…c7qs
> Stream is literally documented as a rolling release
It was originally announced as a rolling release to convey the idea that it "rolls between minor versions", but it still had major versions and EOL dates. Of course this was confusing so we stopped using that term. It's easier to think of it as just not having minor versions. If you find the term rolling on a centos.org site please let me know and I'll work on getting it removed.

2023-07-12 00:15:29

> Stream is literally documented as a rolling release
It was originally announced as a rolling release to convey the idea that it "rolls between minor versions", but it still had major versions and EOL dates. Of course this was confusing so we stopped using that term. It's easier to think of it as just not having minor versions. If you find the term rolling on a centos.org site please let me know and I'll work on getting it removed.

2023-07-11 20:46:20

- reply

@npub1x09…jlxu @npub1tjg…gye3 It's almost like RH operates independently from its parent company...

2023-07-11 20:07:10

- reply

@npub1x09…jlxu @npub1tjg…gye3 Correct. I'm guessing that product (like many others) also has delays on supporting new minor versions of RHEL, holding customers back from upgrading their OS. I'm not sure why anyone would suggest behavior like this is anything but "doing it wrong".

2023-07-11 19:41:27

This article from @npub1tjg…gye3 is awesome. Some key points, paraphrased:
- CentOS Stream 9 is RHEL 9 mainline.
- RHEL minor versions are branched from CentOS Stream major versions.
- CentOS Stream is the open reference implementation of the RHEL ABI compatibility standard.
#CentOS #RHEL #RedHat
https://quantum-integration.org/the-curse-of-bug-to-bug-compatibility

2023-07-09 20:27:00

- reply

@npub18yv…xxyn @npub1pwn…xme7 What are you talking about? I don't know any RH employees that care about what buildsystem Rocky uses. I feel like you're drastically misrepresenting something I said.
https://fosstodon.org/@carlwgeorge/110621103762874827

2023-07-09 20:17:32

- reply

@npub18yv…xxyn @npub1pwn…xme7 I don't know much about SUSE but an openSUSE board member told me that significant portions of SLE are not available here, some product stuff but also buildroot-only packages (which would be critical for creating a rebuild).

2023-07-09 19:53:34

- reply

@npub18yv…xxyn @npub1pwn…xme7 Seriously dude, stop it with the gotcha attitude. source.redhat.com is an employee knowledge base and has nothing to do with source code. The equivalent for your comparison would be here:
https://gitlab.com/redhat/centos-stream

2023-07-07 18:27:43

- reply

@npub1e2p…aqmc @npub10e3…07m7 RH has multiple answers to that.
- CentOS Stream is excellent for this purpose, major version compatible and ensures compatibility with the next minor versions. No subscription needed, unlimited use.
- Developer Subscription for Individuals, 16 instances.
- Developer Subscription for Teams, pay for RHEL in prod, get free RHEL in non-prod (dev, test, staging, etc).
- Official partner program (I don't know all the details on this one).

2023-07-06 23:42:20

- reply

@npub1e2p…aqmc @npub1a9x…rdjj Agreed, CentOS Stream is very RHEL-like. It literally becomes the next minor version of RHEL. I've measured it repeatedly and at any given time it's 90-95% the same software versions. That works perfectly well as a server OS for most people. Not so much if you have 3rd party software vendors that have a hard requirement on RHEL minor versions (and usually lag far behind on certifying new minor versions).

2023-06-23 20:08:35

Are people ready to hear the hard truth? Nah, I don't think they are.

2023-06-22 18:07:23

Free idea for any fediverse client that wants it: an option to hide posts with more than 20 hashtags (or a configurable number). I really like following hashtags here, but some people abuse it and tag their posts with dozens of unrelated (or only tangentially related) hashtags.
#fediverse #mastodon #calckey
https://cdn.fosstodon.org/media_attachments/files/110/588/696/020/448/111/original/715c5716f0c7065b.png

2023-06-10 21:40:57

- reply

@npub1gzl…kk4u Was it worth it? I.e. should I do the same thing?

2023-06-04 19:18:20

- reply

@npub1evu…uf5t Thanks for the kind words. I contributed to Fedora before RH hired me, and I likely will after I'm gone. So I guess I'm a Fedoran first and a Hatter second.
It's not like RH had an army of people working on LO. As I understand it, a leading contributor changed employers from RH to Collabora, so no net loss or gain. Maybe it's more risk to LO if Collabora goes out of business. But if that's the concern then the focus should be on bringing in more corporate sponsorship than 1 or 2.

2023-06-04 04:23:20

- reply

@npub1e2p…aqmc Sweet, I'll be there too, we'll have to grab a meal and/or drink.

2023-06-03 23:29:00

- reply

@npub1evu…uf5t Red Hat is going to make product decisions for RHEL that guide their investment of paid engineer resources in Fedora. Whether each individual decision is good or bad or meh fully depends on how you feel about the thing in question. It's subjective. Those engineering resources are not infinite.

2023-06-03 21:29:10

- reply

Red Hat certainly contributes greatly to Fedora in other ways, but direct maintenance of a specific package or set of packages is not a reliable indicator of Red Hat's overall investment. Fedora is going to be just fine.
#Fedora #RedHat #RHEL #LibreOffice

2023-06-03 21:27:34

I've been seeing lots of "the sky is falling" takes around Red Hat pulling out of LibreOffice RPM maintenance in Fedora. Here's a reality check: the vast majority of packages in Fedora are not maintained in an official capacity (i.e. part of their job) by Hatters. That's because only a small subset (~10%) of Fedora packages make it into RHEL. Currently RHEL 9 has 6,501 RPMs (from 2,393 SRPMs), while Fedora 38 has 60,807 RPMs (from 23,281 SRPMs).
#Fedora #RedHat #RHEL #LibreOffice

2023-05-04 05:47:08

Calckey looks like it has some pretty awesome features. However, using it as a primary account would mean migrating and changing my handle. This makes me think that Bluesky may be onto something regarding "account portability" in the AT protocol. Being able to keep the same handle when migrating between providers sounds really appealing.
#calckey #mastodon #fediverse #bluesky

2023-03-26 16:29:35

Surprising no one, ProPublica reveals that Cigna prioritizes profits over people by mass rejecting legitimate healthcare claims without review, counting on most people to just pay instead of appeal. I guarantee they aren't the only insurance provider doing shit like this. The U.S. healthcare system is fundamentally broken, and will remain that way as long as it is profit driven.
#[0] #[1] #[2] #[3]
https://www.propublica.org/article/cigna-pxdx-medical-health-insurance-rejection-claims