Announcement

Collapse
No announcement yet.

Clear Linux Gets Questions Over Steam Integration, Other Plans For This High-Perf Distro

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Clear Linux Gets Questions Over Steam Integration, Other Plans For This High-Perf Distro

    Phoronix: Clear Linux Gets Questions Over Steam Integration, Other Plans For This High-Perf Distro

    Auke Kok of Intel / Clear Linux carried out the distribution's first ask-me-anything session today where he fielded questions ranging from Steam to under-served software projects...

    http://www.phoronix.com/scan.php?pag...ux-Initial-AMA

  • mgtelve
    replied
    As someone who has worked within the Clear Linux eco-system for the past year or so, I can probably clarify some of the community concerns about the distro's future direction, at least from a lot of developers/users.

    The core of the problem, as many see it, is that the bundle system is inadequate as a software management system, and as Clear has grown in popularity, the broadening use cases have highlighted this weakness. It is incredibly inflexible as a desktop package manager. People wonder and complain "how could this ever be designed!". That's because it was never designed to be a Desktop package manager. It was designed as an Infrastructure OS image manager. There is a system limit to the number of bundles, and packages are not easily accessible at a lower level of granularity. Many developers encouraged by the performance of Clear were asking for an "AUR-like" user repository, but such a thing would destroy the stateless nature of Clear by its definition.

    A lesser issue, though related to the above, is that bundle/packaging priorities become a critical factor when they are limited in number. My general sense from being on the github and forums is that the criteria are: 1) if a Flatpak exists, use that. Clear devs won't make an optimized build. 2). Absent a Flatpak, they seem to use an "Ease of build" vs "criticality" judgment when deciding, leaning heavily toward ease. I've seen small, causal apps be included on first reply, and large numerical queries left open or not a priority. The ultimate problem is that Clear presents a brick wall to the developer interested in a self-made optimized desktop workflow. There is no way to easily build and manage optimization at the package level, so you are stuck either using a vanilla Flatpack or compiling outside of a package system. In which case, Clear is hindering your work and in no way benefitting it, outside glibc/kernel patches/flags. And further, there is no roadmap for even allowing users to develop something like this on their own.

    The design of Clear Linux makes sense when you consider its original niche - infrastructure. You can replicate and update Clear installs without worrying about granular package problems. This is HEAVEN for a devops person populating clusters or virtual networks. It is HELL for the individual attempting to create an install/config specific to their machine and usage.

    As far as Clear on the desktop, and what the devs are saying. If your distro were gaining popularity (and beta testers), would you turn people away by saying "Clear isn't for the desktop", or "we don't care if Steam runs"? Probably not. The devs are being paid by Intel to build an Infrastructure OS, not a Desktop OS. And everyone hoping to find the next Gentoo in Clear should remember that.

    Leave a comment:


  • ms178
    replied
    Originally posted by Vistaus View Post

    ... because Ikey has more talents than just developing LSI?
    If Valve chose to partner with Clear Linux (after abandoning Ubuntu due to their 64bit announcement), this would be an area where they could use his talents very well. But of course it is up to his managers at Intel where to make best use of his talents.

    Leave a comment:


  • Mario Junior
    replied
    Originally posted by Weasel View Post
    Does Clear Linux support proper multi-lib to run at least Wine and Steam 32-bit flawlessly? I don't care about the "Clear Linux performance" with this question, the libs don't have to be hand-tuned and optimized.
    Steam just works on CL.

    Leave a comment:


  • edwaleni
    replied
    Originally posted by Michael View Post

    Right so some UEFI obstacles. Basically what I was getting at was for the older CPU support there are UEFI issues like this as opposed to just not wanting to support older CPUs. If your laptop does properly support UEFI with other Linux distros, I guess there's a buggy problem either with your system or somewhere in the kernel.
    Yes I agree.

    My issues with Clear Linux and Class 1 UEFI are posted in the forum.

    Leave a comment:


  • torsionbar28
    replied
    Originally posted by sturmen View Post
    Having never tried Clear Linux in any capacity, what's holding it back from being a desktop distro? Just the lack of package management that Kjell mentioned?
    It's not intended as either a desktop or a server distro. Its purpose is to act as a reference for performance optimization. I don't believe intel ever intended people to actually use Clear Linux in a production capacity. The fact that you cannot buy a software support contract for Clear Linux proves this. Basically, it's for software developers who are interested in doing things according to intel's preferences and best practices.
    Last edited by torsionbar28; 06-20-2019, 12:51 PM.

    Leave a comment:


  • spstarr
    replied
    Could my posts stop being marked as spam/unapproved :P

    Leave a comment:


  • Weasel
    replied
    Does Clear Linux support proper multi-lib to run at least Wine and Steam 32-bit flawlessly? I don't care about the "Clear Linux performance" with this question, the libs don't have to be hand-tuned and optimized.

    Leave a comment:


  • Vistaus
    replied
    Originally posted by ms178 View Post

    I have the same problem with my Sandy Bridge laptop, so sadly no Clear Linux for us. I also wonder why they brought Ikey Doherty back to Intel but "don't have the ressources" to work on the Linux Steam Initiative. If Gaming were a newfound focus of Clear Linux, this project (or a successor) should get a higher priority.
    ... because Ikey has more talents than just developing LSI?

    Leave a comment:


  • spstarr
    replied
    I've love to see a Fedora Clear Linux SIG, RPMs.. this would require resources though for building, rpmmacros to use newer x86 CPU models...
    Last edited by spstarr; 06-20-2019, 12:32 PM.

    Leave a comment:

Working...
X