Announcement

Collapse
No announcement yet.

openSUSE ALP Prototype "Les Droites" Releasing This Week

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

  • #11
    Originally posted by nadir View Post

    It's named after the mountain in the ALPs.
    It's still not a good name if you have to explain that it doesn't mean what everybody thinks it means at first.

    Comment


    • #12
      Originally posted by nadir View Post

      It's named after the mountain in the ALPs.
      Who had that idea?

      Comment


      • #13
        Originally posted by StefanBruens View Post

        No, you don't. The TPM just stores the key, and you can keep a copy in another safe place.
        ​
        Originally posted by nadir View Post

        And you can add as many additional keys and passphrases as you like.
        That depends on how the FDE is implemented. If it is LUKS, you can hope the master key is encrypted by one or several access keys and the result stored in the LUKS header keyslots: in this case the TPM vault needs 'only' to store at least one of the access keys in the TPM vault, and could store more.
        A naïve implementation would store the master key in the TPM vault, and not have it on disk, which leads to problems with motherboard/TPM failure rendering the disk decryption key inaccessible, unless you have a copy elsewhere (hopefully secure).

        Comment


        • #14
          Originally posted by StefanBruens View Post

          No, you don't. The TPM just stores the key, and you can keep a copy in another safe place.
          I don't know why I always thought TPM was a per device/dongle/doohickey type thing...

          Comment


          • #15
            I criticise that x86_64-v3 isn't used. So now modern computers cannot use there full power and efficency because the system is compiled with old parameters. This is a waste of energy especially in this times.

            Comment


            • #16
              Originally posted by skeevy420 View Post
              Does anyone from SUSE know if the end-game is to replace LEAP with this?
              I am not from SUSE, but it was confirmed multiple times, that Leap 15.5 is the last traditional version.

              Comment


              • #17
                Originally posted by skeevy420 View Post

                Damn shame, ain't it? You'd think the market leader wouldn't be the one releasing the products that hold everyone back by 10 years.
                Yep it is, didn't know we were allowed to use Damn on the forums. Intel has always held back the atom lineup and other value oriented SKUs. My first laptop was one my grandpa got me with an atom inside that was 32bit running windows xp in 2009! 2009 and a 32bit CPU with sata drive running in IDE mode, eMachines should have been ashamed that was my laptop for college. Next I get to grad school and obviously need a replacement for the atom and get an Acer Chromebook 15 that is miles better but still lacks AES-NI instruction set so is a little slow (because Chromebooks use FDE for the /home directory). It goes end of life and my mom helps me get a new Chromebook with a Jasper Lake and I'm like here I go again! So Yeah I've been bitten several times by Intel's crap SKUs of processors

                Comment


                • #18
                  Originally posted by Malsabku View Post
                  I criticise that x86_64-v3 isn't used. So now modern computers cannot use there full power and efficency because the system is compiled with old parameters. This is a waste of energy especially in this times.
                  1. glibc already uses multiversioning, so it uses newer instructions where there is a benefit
                  2. lots of other packages do the same, either using GCCs (now also in Clang) native function multiversioning, or using some home-grown dispatcher.
                  3. lots of software does not benefit from newer instruction sets

                  In the few cases where the package is not using dispatching of any kind **and** would benefit from AVX-512, you can still rebuild this single package with better optimization (and then even incorporate changes which are not covered by instruction set alone, but also considers cache sizes, instruction latency, ...).

                  Comment


                  • #19
                    Originally posted by kylew77 View Post

                    Yep it is, didn't know we were allowed to use Damn on the forums. Intel has always held back the atom lineup and other value oriented SKUs. My first laptop was one my grandpa got me with an atom inside that was 32bit running windows xp in 2009! 2009 and a 32bit CPU with sata drive running in IDE mode, eMachines should have been ashamed that was my laptop for college. Next I get to grad school and obviously need a replacement for the atom and get an Acer Chromebook 15 that is miles better but still lacks AES-NI instruction set so is a little slow (because Chromebooks use FDE for the /home directory). It goes end of life and my mom helps me get a new Chromebook with a Jasper Lake and I'm like here I go again! So Yeah I've been bitten several times by Intel's crap SKUs of processors
                    Shit, man, there are a lot of fucking words you can use on these forums. Pretty much all of Carlin's 7 dirty words and words considered hate speech are allowed.

                    What's funny is a lot of us get flagged when using p#ac#ful words like l@ve and fr!#ndsh!p. The forum's censor bot has its own special logic about things.

                    Comment


                    • #20
                      Originally posted by Leinad View Post

                      I am not from SUSE, but it was confirmed multiple times, that Leap 15.5 is the last traditional version.
                      Thanks, I wasn't sure if I read that as a possibility or as a sure thing. Do you happen to know if the replacement will be ALP or something else?

                      Comment

                      Working...
                      X