Changes coming to Clear Linux' direction

The Clear Linux OS project was created with the aspiration of being able to revise and adapt quickly. Most of the time we use this to incorporate security fixes and other improvements and roll them out to users quickly. For large content features, we do regular long term strategy reviews and align our resources with them - just as any good project would.

We’ve made significant changes to experiment and improve on the developer experience over the last 2 years or so. We are proud of the improvements we made to the developer experience, which was possible thanks to the input and support from our active community.

We made this effort based on the information we had a while back when we were looking forward. Today we’re looking forward again and see that things have changed for the Clear Linux OS team - we still want to attract developers but we are not as invested as we were in supporting a diverse and complex desktop environment, or even multiple desktop environments.

It is with that in mind that we’ve decided to streamline our content offerings with a bias to cloud and server use cases. We do find that it remains critical to developers that we offer actual optimized components unrelated to desktops - that is, cloud and server workloads. This is absolutely where things matter most - being able to develop, deploy and operate those workloads.

As for the specifics on what’s changing - we will announce these to our usual channels: the community forum and mailing list. This won’t happen overnight, we’ll likely take a few months. If you have any specific questions, please don’t hesitate to reach out to us.

2 Likes

Does this imply “the end of Clear Linux Desktop Support” (ie. the capability to be used as a regular desktop for a developer’s environment)?

3 Likes

Most unfortunate. I always seem to have the bad luck of getting into an alternative OS that the creators abandon. BeOS. OpenSolaris. And others. Clear Linux has the potential to be a really great alternative desktop. It will be missed.

3 Likes

I’m newbie and here my two cents:

It’s good news for market, but not for casual users, who continue work (develop) at home, in similar environment but on Desktop edition
I vote YES to keep Desktop edition

PS I uses Clear about month, or more even (not hours, minutes, etc., like other distros for “just try”) as main OS (replaced Windows).

5 Likes

This makes me sad…

Just keep the desktop, only Gnome, only open software as is…everything else via Flatpak and/or 3rd party repositories. Common, why throw such a great desktop OS into trash?!

You don’t have to support every user scenario…give us a secure, stable and clean basic desktop OS and leave every thing else to the community.

I love Clear Linux as my main OS for work and private…it would be a shame to have to switch yet again. I don’t get why so many distributors try to fullfill every single user need.

You say Clear is for developers…only on servers or cloud stuff…or what?

…i’m really not happy with your decision.

6 Likes

No! We will continue to have GNOME, for instance, but it will be a more vanilla approach - we’re not going to sink in as much time as we did last year.

Having a desktop allows us to work on performance related to graphics, media/sound and other related things - so we have to have a desktop environment no matter what.

6 Likes

That’s what the plan is - largely. I think we’ll keep Plasma (KDE) and Xfce4 around as well. We’re definitely happy to see some people are making 3rd party repos with applications, as that is a good way for us to focus on the things we need to, and for users to get the end-user applications they want, without us fighting over resources.

3 Likes

re: Changes

This change in direction is unfortunate but understandable - intel made an assessment of effort vs resources and providing end user support to a diverse set of users lost out.

Ideally we won’t have dozens of different repos coming and going.

I would much rather see or start the equivalence of EPEL (as to Red Hat) – a larger and more formal collection of volunteers working together curating 3rd part Clear Linux packages. less duplication of effort and more possibility of a repo lasting.

3 Likes

Given the new Clear Linux main focus is (or will be) on cloud & servers, will you invest more resources on getting better domain join compatibility w/o 3rd party commercial applications?

Namely, I was interested on extended support for sssd (in order to join Clear Linux server to a Windows domain and apply some basic group policies like RHEL already allows you to).

sssd is part of the enterprise-login bundle. If there’s anything we can do to improve that bundle, I encourage you to open a github enhancement request and document it.

I would personally love to see it, too.

1 Like

Thank you @ahkok for clearifing the situation. After reading your initial post i was a bit shocked, maybe because english is not my native language. I tought you would stop the support for a desktop altogether.

This sounds very good to me and i can fully understand it. It is a mass of work to maintain a distribution in a good way, you and all the other in this distribution involved people do a fantastic job with the developement and all the support and documentation.

I don’t know if this would be even a option or possible business wise, but i would be happy to donate a small amount of money. Maybe give the option to do so on the download page, like other ditributors do or maybe enable the Sponsor option at GitHub.

I would love to support the team with developement, but unfortunately i don’t think i would be much of use in this case. :slight_smile:

At the bottom line, good news. Thank you all for the hard work! :love_letter:

2 Likes

Hello! I’m a big fan of the project overall and let me see if I understand it right:
Clear Linux will deliver optimisation to server and cloud packages and let desktops (such as gnome) extremely similar to their source codes (therefore without many optimisations)
Is this right?

I think as things evolve, it would be helpful to know pretty clearly what is and what is not in scope. I’ve posted comments saying that before. Most of us enthusiasts try to “read the tea leaves” of developer comments but honestly, even in this comment I’m not sure what is going to be in or out of scope (but I gather you haven’t fully figured that out yet). Prior to this the general recommendation was “file a github request” for software but that opens you up to a request for basically anything and I don’t think you have the resources to do that.

As an independent developer who is typing this on a ClearLinux desktop, has a ClearLinux laptop and who’s home development box runs Clear, I would say that I probably use more ClearLinux software on the desktop than on the server. The server runs a base ClearLinux server install which … runs docker. Some of the docker containers (mariadb, nginx, php-fpm) are ClearLinux based but many are not. (Plug to implement the HSTORE extension for Posgress and I could flip that one over). Cups doesn’t run very well as a printer server in Clear so there’s a docker for that. home-assistant uses the docker; not the ClearLinux package.

On the desktop, I think the big challenge is you’re crippled by the inability to include codec support your average person wants. I know that’s not your fault, but “not-ffmpeg” is the first thing I replace on any install. This is coupled with the buggy, crash prone nature of the flatpaks. Midori will crash and take the whole system down with it in a flatpak. VSCodium works ok as a flatpak and I use the the flatpak Gimp as it supports codecs which again, your compiled gimp doesn’t. At some point I just gave up and decided “I’ll just use my compiler” and /usr/local has most of what I use these days.

1 Like

Best of luck. I will stay on CL until I can run it on my laptop. Thank you.

We will still compile desktop software with all our optimizations. We may decide to remove specific high-cost tweaks, but it should still be better than an average distribution without our compiler and build tools.

1 Like

If you haven’t already, that is definitely a feature we’d want to support, so please open a github ticket for that?

Thanks. I’ve seen lots of users express their opinion in a similar way. My personal take is: if you do feel like you are willing to contribute more than just your interest in Clear Linux, please figure out if you and others in the community outside of Intel can do something together to work on things that benefit the community (and not Intel directly).

1 Like

Pity, I just installed Ubuntu 20.04 on my laptop, and as expected, about 10% decline in performance.

That said, you won’t be credible as a production server OS without support for ZFS.