Clear Linux Further Enhances Its Desktop Installer, Launches Help Forums

[ChadD]

Considering the long history of Intel bucking Intel software wise. I think its interesting that as MS seems to be embracing more Linux and open source stuffs... Intel seems to be getting a little less worried about ruffling MS feathers. 20 Years ago MS would have went well out of their way to keep keep Intel from making even a media player that didn't bow to all MS APIs. Now Intel is not only producing a performance monster OS... they are making it more and more user friendly.

More proof we live in bizzaro future.

[edwaleni]

I must have downloaded a release just prior to the one Michael just showed us. I was missing many of the GUI screens he displayed as every component installed as a text based terminal window inside the live GUI.

I have been trying to run some PTS (8.6) tests (linux-system) on the Nouveau based install and it punts. This time I ran PTS and about 45m into the test script the system mysteriously rebooted itself and started up in a kernel panic and hung. I hadn't even reached the graphics part yet. So it seems I will download this new release (again!) and see if it will survive another PTS run.

For those who might ask the donor PC is a HP corporate desktop with a Sandy Bridge i7 and the Nvidia GF7 with a SATA SSD. So yes, it is old.

I have loaded several Linux spins through this PC over the last year and PTS runs fine until now with Clear, so I will keep plugging away and see what gives. It's very possible Clear gets its performance by extracting more from its host hardware and the generation I am working with is just too old.

[Guest]

I have loaded several Linux spins through this PC over the last year and PTS runs fine until now with Clear, so I will keep plugging away and see what gives. It's very possible Clear gets its performance by extracting more from its host hardware and the generation I am working with is just too old.

From what I've read, Nehalem might be the oldest compatible generation, but the documentation states that AVX is required, so Sandy Bridge would be the oldest supported generation of CPUs.

https://clearlinux.org/documentation/clear-linux/reference/system-requirements#system-requirements

[skeetre]

I must have downloaded a release just prior to the one Michael just showed us. I was missing many of the GUI screens he displayed as every component installed as a text based terminal window inside the live GUI.

Same for me...

I have been trying to run some PTS (8.6) tests (linux-system) on the Nouveau based install and it punts. This time I ran PTS and about 45m into the test script the system mysteriously rebooted itself and started up in a kernel panic and hung. I hadn't even reached the graphics part yet. So it seems I will download this new release (again!) and see if it will survive another PTS run.

Mine crashed twice in a row trying to do a generic overall type benchmark too. I didn't look into where/why though, no time right now.

[edwaleni]

It seems I had downloaded the desktop beta release 29250 and Michael is working with release 29340. So I missed the new release by a few days.

Another thing Intel has finally done, put it in a ISO format. Can't tell you how many blogs I have read on people complaining about lack of ISO based images for Clear.

So we are off once again.

[edwaleni]

I tried to install it on my MSI MEG Creation X399 / Threadripper 2950X and it just can´t find the NVMe´s Any help would be appreciated

See this bug report:

https://github.com/clearlinux/distribution/issues/392

Seems different NVMe's work and some don't. Someone tried to pin it to AMD based boards, but that is not the case.

In other cases its the firmware in the NVMe module itself. Did you check if yours is current? Seems different brands are responding differently to Clear's enumerator, especially if the firmware is dated.

[mumar1]

See this bug report:

https://github.com/clearlinux/distribution/issues/392

Seems different NVMe's work and some don't. Someone tried to pin it to AMD based boards, but that is not the case.

In other cases its the firmware in the NVMe module itself. Did you check if yours is current? Seems different brands are responding differently to Clear's enumerator, especially if the firmware is dated.

Thanks, I already found this thread and I get pretty much the same output for dmesg | grep nvme and the pcie device list as given in the first post there. But there is no solution offered in this thread.

[edwaleni]

Thanks, I already found this thread and I get pretty much the same output for dmesg | grep nvme and the pcie device list as given in the first post there. But there is no solution offered in this thread.

Did you check the firmware of your NVMe module? Is there an update? It was one of the recommendations.

[mumar1]

Did you check the firmware of your NVMe module? Is there an update? It was one of the recommendations.

Not yet, I´ll do it after lunch

[ms178]

"Gesture to the community". Honestly I'm going to be blunt. The "community" has never had it so good. I grew up when software was *very expensive*, and that included the whole stack. Now you demand concessions from an excellent and already-free, yet expensively developed, piece of software, to go even further and literally shoot at itself by supporting other architectures. Then you point to all these hypotheticals about how giving away even more free stuff to the (already very spoiled) community might make them love you a little bit more. You've forgotten the wonderful world that you live in. It could be much much worse than this. Actually, we should all be paying a lot more money towards Open Source than we currently are. But No. We prefer to pay zero, and still ask for more to somehow build more goodwill from an entitled "community" that doesn't appreciate what it has.

That mindset of yours implies that these open source strategies are a way of selfeless generosity of these companies which we should be thankful for. Let me tell you, thankfulness is not a category business think of. They wouldn't have come to the conclusion that this strategy is the better one than the closed-down expansive silos of the past if it wouldn't make business sense for them. And there is a broader picture to this: Open Source furthers collaboration and portability. Other parties have an incentive to invest into an open ecosystem and build on top of all this because the barrier to entry is very low (relative to the past) thus building an ecosystem around your technology. That is a much better way to cement your market power than how they did it in the past which stiffled all of that innovation.

By the way, it would cost them very little (build time, disk space, automated testing) and flipping of a compiler switch. At least the OpenMandriva guys told me that the added costs for their Zen-builds are very minor compared to the support of a totally different architecture (like Aarch64). I don't expect Intel to do the same level of in-depth optimizations they do for their own architectures though. I guess they wouldn't reject pull requests from AMD either (most of these low-level work is done collaboratively upstream anyway).