Announcement

Collapse
No announcement yet.

Reiser4 Brought To The Linux 5.0 Kernel

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

  • #61
    Wow - Im surprised of how ignorant this community really is.

    "The once promising file system [...]"

    Once? Its one of the more modern approaches to innovative features and design written entirely from scratch.

    And since many of you obviously dont know a whole lot of the key features it provides, let me summarize some of them for you and even prove the author as well as some of the people posting here wrong.

    To date, Reiser4 is one of the few which support for inline crc32 checksumming through its (also quite unique) plugin design, with their 'node41' plugin which will only require 3 cpu cycles. This is one of the later plugins to be created for the FS and not many other FS besides ZFS have in common. There is also subvolumes support with mirroring and failover support recently introduced not to mention the added support for Zstd compression in addition to lzo and gzip. Compression is the only plugin that is enabled by default which the author never seems to understand when publishing bemchmarks vs other filesystems which do not have this enabled by default.

    Also a very important features is the wandering logs-design which is way more efficient in *every* aspect vs either metadata or regular journaling. It has also taken the very much praised feature from XFS with it's allocate-on-flush support. Besides that it *is* in fact the only FS with a atomic transaction design in which the IOP entirely occur or entirely don't (btrfs comes to mind). Instead of using the B-tree design it uses the 'dancing tree' feature to balance the B*Tree which is undoubtedly a much smarter design than a regular B-tree design like btrfs uses.
    In addition it also features Mount options (transaction models) to both suit HDD and SSD disks optimally. These include:

    Write-Anywhere (SSD), Journaling (HDD) and a heuristic model called hybrid which alternates between the two for optimal performance on SSHD and HDD users. The bemchmarks posted here is using the defaults for all of the filesystems and WILL result in sub-optimal performance in synthetic bemchmarks because they all differ in their defaults.

    Perhaps the most direct alternative is btrfs which does NOT enable compression by default while Reiser4 does.

    What is also written which is just wrong is that multiple requests for inclusion has been done. What Edward Shishkin said about the topic was that without a corporation backing up financially for the people involved in the development, there is only so much a handful of people can do to get the system included in the mainline Linux-tree. Many others (like those of you that are suggesting a new name for a filesystem because of the person who initially started on its design is convicted of murder is in fact a solid piece of evidence for) are suggesting that the reasons for not being mainlined is in fact political reasons and not flaws or lack of innovation or other technical challenges which the author of the article implies when saying that the other filesystems is a step ahead of reiser4 when the obvious fact is that that is not the case *at all*. I have used reiser4 since 2012 on my laptop and not even once experienced dataloss or corruption. Google btrfs and read up on the fact that the percentage of users having issues with it although being mainlined is disturbingly high.

    I hope the next bemchmarks will include results from the most equivalent of both plugins and Mount options to show how incredibly well the performance and features of Reiser4 in 2019 actually is.


    Comment


    • #62
      Originally posted by petterk View Post
      Wow - Im surprised of how ignorant this community really is.

      "The once promising file system [...]"

      Once? Its one of the more modern approaches to innovative features and design written entirely from scratch.

      And since many of you obviously dont know a whole lot of the key features it provides, let me summarize some of them for you and even prove the author as well as some of the people posting here wrong.
      You've told us nothing new. The reason it is considered "once promising" is its inability to get accepted into mainline. Very few users are going to go with out-of-tree file systems which don't seem to be getting consistent developer support. As great as reiser's features are, without mainline kernel support and consistent and up-to-date maintenance, very few users are going to bother with it.

      Comment


      • #63
        I've told many of you plenty new. The article even mentions ZFS which is incompatible with the license of Linux and will always stay out of tree - does that make it not worth writing articles about because of it? Where is your proof that there is lack of ability to maintain it? Thats not the case and the developers have become more in numbers with the once-leading developer still actually sticking to the design and principles that were the foundation of its design. And if you fail to acknowledge that reiser4 is just as easy or easier to implement than say ZFSoL, then you are missing out on things that in the future could have saved your data WHEN that happens for those of you with btrfs. And if its the actual name some of you want to change because it's so offensive - well then I feel sorry for how overly sensitive you are and lack of ability to see that Hans Reiser is serving his time, its predecessor is already mainlined and was default in SuSe Linux for a long time - why dont you complain to Novel for still being one of the few companies to provide support for Reiserv3 then... person vs technology - you cant mix those because of a single individual.

        Comment


        • #64
          Personally I think its a much bigger concern that filesystems like btrfs which ARE mainlined can end in it's users having huge data loss and corruption - yeah it's still not marked as stable and THAT is what a failure to maintain something looks like. I'd chose reiser4 over something that is broken by design despite that btrfs is mainlined. It's 6 pages of useless comments on a positive news article which unfortunately is somewhat wrong in it's statements, and 90% of them talk about renaming the project and calling it 'ninafs' - thats just sick...

          Comment


          • #65
            In most parts of the world, a diagnosis within the autism spectrum disorders called 'Asperger Syndrome' - The guy who first wrote about the medical condition, Hans Asperger which also named the condition 'autistic psychopaths', has been directly proved to be part of the child euthanasia in nazi-germany at tye Am Spiegelgrund, when Hitler was their leader. He selectively chose the faith of up to multiple hundreds of child ( (murders) for 'research'. In 2019, the name is still not changed because of it in the ICD 10/11 which among others is us used by WHO, even though this in this case, it was in fact the murders and his actions which made it possible to describe the diagnosis which in almost all parts of the world still using the name Asperger syndrome, except for the ECD which put the diagnosis under the autism spectrum in general because the allegedly lack of medical descriptions about the disorder. TLDR; Even today the name is used despite being a child murderer and part of the nazi party. This is as close as you get to 'Hitlers syndrome' yet it still holds it's original name.

            For those of you who are so offensed by the file system whose name was also used because he was at the very least the visionary behind it's design and model, compare the two identical cases (in principle) and see how little the name of something actually means.

            Comment


            • #66
              Originally posted by andyprough View Post
              Yup.
              Any one of them could fork it tomorrow and call it "NinaFS" or whatever, but none of them will do squat.
              I don't want to work with people either who can only see values in code but little else. It's fun to write great software with a bunch of young people. Nobody cares how you do it as long as the result is great. That's freedom. But when someone turns out to be a murderer then it kicks the doors to society wide open, everyone gets exposed and judged because of one person doing something unthinkable. That's where the fun ends.

              Many ran from the project back then and they'll know why. I can very much understand and appreciate their reaction. Those who remained and even kept the name are then simply mislead by false ideals. Writing software isn't everything in life, but you won't find out if you never do anything else.

              Comment


              • #67
                Originally posted by labyrinth153 View Post
                Yeah I’m a “social justice warrior” because I can’t fathom why they would continue to develop a half assed filesystem and continue to name it after it’s murdering creator. Yeah, you are a sociopath.
                Well, posting on forums demanding that people should rename a FS due to it's original creator is a convicted murderer is exactly what the anti-SJW crowd claims is SJW behaviour so in fact this thread is the first time that I've seen those allusive SJW:s that apparently have infected every aspect of Linux and open source.

                Comment


                • #68
                  Originally posted by petterk View Post
                  Personally I think its a much bigger concern that filesystems like btrfs which ARE mainlined can end in it's users having huge data loss and corruption - yeah it's still not marked as stable and THAT is what a failure to maintain something looks like. I'd chose reiser4 over something that is broken by design despite that btrfs is mainlined. It's 6 pages of useless comments on a positive news article which unfortunately is somewhat wrong in it's statements, and 90% of them talk about renaming the project and calling it 'ninafs' - thats just sick...
                  What you have missed here is that it's certain configurations of btrfs that is marked as unstable (especially raid 5/6), for other uses such as raid1,0 and 1+0 it has been stable for years now.

                  Comment


                  • #69
                    Originally posted by hreindl View Post
                    where can I get the mushrooms you are on?
                    the technical stuff aka code is completly independent of who wrote it no matter if the person is dead or in jail - period
                    No its NOT!
                    In this case,
                    The event of developing that software is the base of a monstrosity, which ended in the killing of Nina Sharanova..

                    This is not a *tipical case*, like when you work has for a Company, that then works for a Client..
                    Here everything is the same thing, since they were married, and was a "family business"

                    This his different,
                    And the fact that some very well known key Open-Source developers left the Community was exactly because it started to become too much toxic..
                    This Toxicity, come when people that arrive from a Sauvage Capitalism, enters into the Open-Source sphere..

                    This people are toxic, and have no values or moral, they don't feet into the spirit of Open Source World,
                    But they don't stop to arrive...unfortunately.

                    Its pollution all over the place nowadays..





                    Comment


                    • #70
                      I’m not demanding anything. I’m saying it’s stupid and bad taste.

                      Comment

                      Working...
                      X