Slashdot Mirror


Practical File System Design with the Be File System

erikharrison writes "Dominic Giampaolo's Practical File System Design with the Be Filesystem has been around since 1999 - not exactly a new book. The book has been out of print for a time now, however, so Dominic made the book available in PDF form on his website. With this public release of the book, and the BeOS rising to join the ranks of OSs that won't die (hi Amiga!) it makes sense to take a look at what the book has to offer us today." Read on for the rest of Harrison's review below to see just what that is -- it covers a surprisingly broad range. Practical File System Design with the Be File System author Dominic Giampaolo pages 227 pp publisher MORGAN KAUFMANN PUBLISHERS, INC. rating 8.5 of 10 reviewer erikharrison ISBN 1558604979 summary Discusses implemeting a file system, using the Be file system as example

Table of Contents

  • Chapter 1 Introduction to the BeOS and BFS
  • Chapter 2 What Is a File System?
  • Chapter 3 Other File Systems
  • Chapter 4 The Data Structures of BFS
  • Chapter 5 Attributes, Indexing, and Queries
  • Chapter 6 Allocation Policies
  • Chapter 7 Journaling
  • Chapter 8 The Disk Block Cache
  • Chapter 9 File System Performance
  • Chapter 10 The Vnode Layer
  • Chapter 11 User-Level API
  • Chapter 12 Testing
  • Appendix A File System Construction Kit

First thing to note is that Giampaolo is not a great writer, nor is he a bad one. He does not have the gift that some tech writers have of making both an interesting technical document and a fun read. His style is very straightforward - introduce idea, explicate idea, summarize idea. On the other hand, he knows his topic inside and out, and has an obvious enthusiasm for the material, and a real talent for saying things simply without dumbing it down, and his occasional dry wit makes the book a surprisingly easy read.

Giampaolo is doing two things - discussing designing filesystems in general and documenting the Be filesystem. He does both well. BeFS has some advanced features - arbitrary metadata, attribute queries, and indexing. The desire to support these features influences the overall design of the system, but Giampaolo shows how changes to that design change implementation details. The result is a good overview of how a file system works, the trade-offs in optimizing for a particular usage pattern, and how to design one yourself.

The book can be roughly divided into three sections: the first is an overview of how filesystems work and some of the concepts that you encounter - extents, inodes, B-trees, superblocks, and the other standard pieces of a filesystem. Included in this early section is a good high-level overview of the design of five other file systems: BSD FFS, Linux's ext2, Macintosh HFS, Irix XFS, and Windows NT's NTFS. The coverage here strikes a proper balance between too much and too little information. Giampaolo prefers to show rather than to tell, and these filesystem overviews make the connection between design, performance, and features perfectly clear, and provide a solid background to talk about a specific implementation in detail - namely BeFS.

The second section is the bulk of the book - how to implement a filesystem from the ground up, leaning heavily on the BeFS implementation for examples. This is the most straightforward part of the book. Giampaolo covers a single issue in design and implementation in a "Here's the problem, here's and overview of possible solutions and their drawbacks, here's how I did it, now lets summarize" manner. Again, Giampaolo's style makes this an easy if somewhat dry read. As a filesystem and kernel ignoramus, I would have appreciated a slightly more detailed coverage of how all of the various data structures get to disk - how are they serialized, whether endianess is an issue, etc. The BeOS was pretty portable, running at one time or another on the AT&T Hobbit processor, PowerPC, and x86 - I would have liked to have seen portability issues discussed, however, BeFS wasn't written until after the move from the Hobbit to PowerPC, and the book was written prior to the move to x86, so the lack of coverage is reasonable.

Even considering the plain Jane style of this middle section, there are a few gems. The coverage of journaling is excellent, and while I've long understood journaling from a 10,000 foot perspective, this really made me understand the underlying concepts, combined with simple code snippets that helped understand implementation. The Allocation Policies chapter showed in clear terms that disk access is a major bottleneck, and filesystems have become very sophisticated in their optimizations.

The third section of the book deals with some of the more indirect concerns in implementing a file system; specifically, interacting with the kernel, designing a user level API and the major role of testing in filesystem development. This is the one place Giampaolo's writing shines. He really is a good teacher, and this section affords him the chance to talk about the broader perspective of OS design, and even recount a few war stories. For example, in terms of parentage, the BeOS has BSD and classic MacOS as its father and mother. In a few places, such as the Storage Kit API covered in chapter 11, this heritage shows some signs of less-than-seamless integration, and this offers Giampalo a chance to wax philosophical on the nature of OS design, company politics, and the pressure of shipping dates.

In short, the book lives up to it's title. The author is a pragmatist, and offers a clear roadmap for those who have a need to work with low level filesystem implementation. His emphasis on testing, careful optimization, and data structure protection not only helps to show the pitfalls of filesystem work, but also offers a Swiss army knife of techniques to dodge them. The book concludes with a short appendix which covers a file system construction kit, allowing a would-be implementor to begin work on his own filesystem safely without worrying about killing his hard disk. All in all, a solid read.

Here's a link to Practical File System Design with the Be File System as a PDF; you can also look for a used copy at Barnes & Noble. Slashdot welcomes readers' book reviews -- to see your own review here, carefully read the book review guidelines, then visit the submission page.

258 comments

  1. Mirror with PDF by BiggestPOS · · Score: 5, Informative

    http://biggestpos.com/pfs/ It took me so long to download the file I thought I'd put it on a faster server for you guys.

    --
    What, me worry?
    1. Re:Mirror with PDF by nacturation · · Score: 4, Informative

      This reminds me of the guy who put up an Adobe Golive PDF book for download. Wired article here. Since nobody reads links, the short story is that it was downloaded 10,000 times in 36 hours and faced a possible $15,000 bandwidth bill (which was later rescinded by Level 3, his hosting company).

      Earthlink wasn't so charitable to a Halo fan who put up a movie previewing Halo and saw it downloaded 100,000 times. Earthlink charged him $30,000 for the 4500 gigabytes downloaded.

      The moral here is that if you're going to put up anything which might be downloaded quite a lot (or if you're expecting a slashdotting), make sure your host doesn't charge through the nose for extra bandwidth. Or, if the file is over 5MB in size (and under 1GB), make a freecache.org link and let others mirror it for you automatically.

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
    2. Re:Mirror with PDF by BiggestPOS · · Score: 1

      The file is just over a meg, and I've got tons of extra bandwidth to burn, I think :)

      --
      What, me worry?
    3. Re:Mirror with PDF by Neil+Blender · · Score: 1

      make sure your host doesn't charge through the nose for extra bandwidth.

      Or just make sure you have 95th percentile billing. That way, you could survive a short burst (up to 36 hours/month) of high bandwidth since you don't pay for the top 5% of your usage. And if you can't burst beyond a certain number, you'd best rate-shape at the firewall.

    4. Re:Mirror with PDF by nacturation · · Score: 1

      Coincidentally, the banner ad showing at the top (yes, I don't disable ads... some are interesting) has the $59 ValueWeb server special with 1 terabyte of bandwidth.

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
    5. Re:Mirror with PDF by Profane+MuthaFucka · · Score: 1

      The solution to this is to host your site yourself on your own unmetered DSL line. It's slow, but if there's a big demand for something, there's always someone else who will cache it for you.

      --
      Fascism trolls keeping me up every night. When I starts a preachin', he HITS ME WITH HIS REICH!
    6. Re:Mirror with PDF by kryptkpr · · Score: 4, Informative

      Have you been living under a rock? The solution is BitTorrent!

      --
      DJ kRYPT's Free MP3s!
    7. Re:Mirror with PDF by N1KO · · Score: 1

      Set up a torrent, Post the thing on newsgroups.

    8. Re:Mirror with PDF by G-funk · · Score: 1

      Much appreciated.

      --
      Send lawyers, guns, and money!
    9. Re:Mirror with PDF by Anonymous Coward · · Score: 0


      Ah yes, but try getting that terabyte quickly. You may very well find that your TB is rate limited over a month.

    10. Re:Mirror with PDF by BiggestPOS · · Score: 1

      I disabled SOME ads because it said you had to in order to see stories early, but I'm certainly seeing many ads anyway, is that a good thing or bad thing for /. considering I've never in my life so much as glanced/clicked or even as such considered one?

      --
      What, me worry?
    11. Re:Mirror with PDF by nacturation · · Score: 1

      Normally, I don't even notice those banner ads. And I don't subscribe, so I don't know if there are any benefits to showing them. As to whether it's good or bad for you, judge for yourself. For me, an ad is relevant if I happen to be in the market for what they're advertising. Otherwise, like you, I don't even consider it.

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
  2. Crud more file Systems? by LoveTheIRS · · Score: 0, Flamebait

    Aren't there enough filesystems? All you need is one!

    1. Re:Crud more file Systems? by Anonymous Coward · · Score: 1, Insightful
      All you need is one!

      Surely all you need is love?

    2. Re:Crud more file Systems? by FictionPimp · · Score: 5, Funny

      isn't love a journaling filesystem? My girlfriend can remember every little issue I would like to delete.

    3. Re:Crud more file Systems? by spellraiser · · Score: 3, Insightful

      Aren't there enough filesystems? All you need is one!

      Yes, and 640K is enough for anyone, and everything worth inventing has already been invented.

      But seriously, what's wrong with creating new things? A new filesystem just might be better than the current ones. Stranger things have happened. Much stranger things, in fact. Especially in Computer Science. Remember hierarchical databases? No? Well, they used to be all the rage back in the old days, before relational databases took over. These days, the idea of using any other database model seems ludicrous. However, even those might be superseded at some point in the future.

      It never hurts to keep an open mind you know...

      --
      I hear there's rumors on the Slashdots
    4. Re:Crud more file Systems? by .com+b4+.storm · · Score: 3, Funny

      isn't love a journaling filesystem? My girlfriend can remember every little issue I would like to delete.

      You think that's bad? Wait until you upgrade to the wife filesystem. It won't even let you delete entries for mowing the lawn, etc. from your crontab! Talk about rights management...

      --
      "Wow, you're like some kind of superhero able to ward off happiness and success at every turn."
      -- Ryan Stiles
    5. Re:Crud more file Systems? by Anonymous Coward · · Score: 1, Interesting

      These days, the idea of using any other database model seems ludicrous.

      I would kill to get a PROPER object-database instead of spending 1/3 of the development time on manually making O/R mappings, Factory-objects, etc. In most situations I don't care how data is stored in the memory so why should I care how it's stored on the hardrive? Too slow, you say? Buy a bigger box! Cheaper than wasting $70.000 salleries on redundant stupid work... (by the way, done correctly, in many scenarios I can't see a big performance punishment)

      Totally off-topic, sorry ;)

      Cheerios
      Peter

    6. Re:Crud more file Systems? by gavinjolly · · Score: 1
      You think that's bad? Wait until you upgrade to the wife filesystem. It won't even let you delete entries for mowing the lawn, etc. from your crontab! Talk about rights management...

      Dont worry, DHL is a method for filtering out some of the entries from your inputs. DHL (Domestic Hearing Loss) may be installed once a domestic situation develops.

      • DHL enables effective filtering of Noise
      • Generally, it will not filter out Hardware issues such as body gestures, slamming doors or tears.

      --

      The weathers here - Wish you were beautiful

    7. Re:Crud more file Systems? by BiggerIsBetter · · Score: 4, Funny

      You think love is bad? Wait till you see the vengence in action. Sure it remembers everything little issue, but using an advanced AI system it creates arbitrary yet realistic scenarios in a multi-threading manner. These can be variously intepreted based on context by others systems reading the disk, which may result in locks and race conditions or possibly even data loss. Love is nice, but wouldn't recommend the post-love vengence patch - best bet is to keep plugging away with what you've got. Don't forget to keep up with maintenance and the system's support needs.

      --
      Forget thrust, drag, lift and weight. Airplanes fly because of money.
  3. Future is relational databases by leandrod · · Score: 2, Insightful

    One more backwards-looking text... it simply ignores that the future is relational databases as a filesystem.

    --
    Leandro Guimarães Faria Corcete DUTRA
    DA, DBA, SysAdmin, Data Modeller
    GNU Project, Debian GNU/Lin
    1. Re:Future is relational databases by happyfrogcow · · Score: 4, Insightful

      and just where are you going to store that database? on another database acting as a filesystem? or will you have a db built into the OS, right next to your flight simulator, and supercolliding monkey accelerator?

    2. Re:Future is relational databases by master_p · · Score: 1

      You are right, but not only relational...the future is object-oriented databases.

    3. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Agreed. Link

    4. Re:Future is relational databases by Anonymous Coward · · Score: 1, Funny

      t next to your flight simulator, and supercolliding monkey accelerator?

      No, it's next to the frog-exaggerator.

    5. Re:Future is relational databases by LWATCDR · · Score: 1

      Maybe but maybe not. There are some performance issues to look at not to mention that even if the relational database uses a raw partition it is still use some sort of filesystem to store its data on disk and to manage free space. Even if it is one of it's own design. An how do you know ti will be a realtional database? Maybe it will be an OO database.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    6. Re:Future is relational databases by Neil+Blender · · Score: 1

      the future is object-oriented databases.

      And flying cars. Seriously, is there any decent object oriented db out there? Something that you could trust in production? I have been hearing about them for years but where are they?

    7. Re:Future is relational databases by ceswiedler · · Score: 4, Interesting

      It's interesting that one of the features of BeFS is its metadata indexing capabilities (which are the beginnings of a relational model).

      Clearly the BeOS designers agree with you.

      It's also interesting that the author spends quite a while discussing how difficult it is to do well (particularly performance-wise) and how they almost left it out (IIRC) and/or had to limit its scope.

      Clearly the BeOS developers think you're wrong.

      And personally, I'll believe people who have actually tried to implement the technology in question over people who say others should do so.

    8. Re:Future is relational databases by AKAImBatman · · Score: 1

      Maybe you're trying to be funny, but BeFS was renowned for its database-like properties that allowed for easier and quicker searches to find data.

    9. Re:Future is relational databases by 91degrees · · Score: 1

      True. And back in 1999, BeOS was heading in that direction. Not exactly a database, but it had some nice features that are associated with databases.

    10. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Well that seems like such a bad idea that you would expect Microsoft to adopt it. Wait, maybe they have.
      Why put a bloated product like a database between a program and the filesystem, unless you want to slow things down and obfuscate. Oh thats right, that is some peoples goal, too.

    11. Re:Future is relational databases by leandrod · · Score: 1
      > maybe not

      Only if inertia and ignorance continue to hinder us.

      > here are some performance issues

      No, this is a physical issue. The relational says nothing about the physical level, thus leaving the implementor total freedom to achieve the best performance possible. This is typically better than what is possible navigationally.

      > even if the relational database uses a raw partition it is still use some sort of filesystem

      Not at all. What makes you think so?

      > how do you know ti will be a realtional database? Maybe it will be an OO database.

      OO has already been laughed off the contest.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    12. Re:Future is relational databases by ckaminski · · Score: 2, Informative

      www.objectstore.net

      Amazon.com uses it.

    13. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Maybe that's why his post was modded +5 Funny. Clearly everybody else understood the joke but you.

    14. Re:Future is relational databases by leandrod · · Score: 1
      > will you have a db built into the OS

      Why not?

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    15. Re:Future is relational databases by Elwood+P+Dowd · · Score: 1

      OO databases are myriad and variegated. Yes, there are many that you could trust in production. Of course, if your current problem is convincing your boss that any non-MS-SQL solution is viable, then no, no there is not one that you can trust in production.

      If your current problem is that you have important data analysis that cannot be completed with your current Oracle/whatever relational DB due to an algorithm that could lower it's order a few exponents with an OO design, then yeah. There are all kinds of production solutions for you.

      --

      There are no trails. There are no trees out here.
    16. Re:Future is relational databases by leandrod · · Score: 2, Insightful
      > its metadata indexing capabilities (which are the beginnings of a relational model)

      No, they could be the beginnings of an implementation of the relational model.

      > the BeOS designers agree with you

      If they did they would have gone full ahead instead of taking half measures.

      > It's also interesting that the author spends quite a while discussing how difficult it is to do well (particularly performance-wise) and how they almost left it out (IIRC) and/or had to limit its scope.

      No one ever said data was easy. But implementors usually do this particular blunder because they never understood the relational model, confounding it with SQL.

      > I'll believe people who have actually tried to implement the technology in question over people who say others should do so

      No, I am not talking about their technology. I am talking about the relational model.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    17. Re:Future is relational databases by rapett0 · · Score: 2, Funny

      Of course, you know the quote that old lady gave to I believe Carl Sagan, "its relational databases all the way down"

    18. Re:Future is relational databases by leandrod · · Score: 1
      > BeOS was heading in that direction

      Not quite... they were more doing something SQL-like.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    19. Re:Future is relational databases by no+reason+to+be+here · · Score: 4, Informative

      Actually, Be started with a full database instead of a file-system. They found it to be incredibly slow and crash-prone, and so they developped the marvel that is the Be file system.

    20. Re:Future is relational databases by gwappo · · Score: 4, Informative

      Amusing sarcasm but there's no reason for a database not to write directly to any random access device.

    21. Re:Future is relational databases by Tom7 · · Score: 1

      You are right, but not only relational...the future is object-oriented databases. ... implemented in XML.

    22. Re:Future is relational databases by leandrod · · Score: 1
      > Maybe you're trying to be funny

      Am not.

      > BeFS was renowned for its database-like properties

      That's not what we need. What we need is a full relational database as the core of the OS, with a relational-enabled language as the primary systems programming language.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    23. Re:Future is relational databases by dasmegabyte · · Score: 2, Funny

      Because that's what Microsoft is doing. So it must be wrong (up until Longhorn sells its first 10 million copies, then we'll scramble to implement it).

      --
      Hey freaks: now you're ju
    24. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Yes, that's why we're all using desktop AS/400s.

      I do agree with you, actually but it will probably take quite a while yet.

    25. Re:Future is relational databases by leandrod · · Score: 0, Troll
      > Be started with a full database instead of a file-system. They found it to be incredibly slow

      Ergo, it was not relational.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    26. Re:Future is relational databases by Tim+C · · Score: 2, Informative

      I'm no DBA (I'm a programmer), but I *believe* that Oracle can do just that if you configure it to - write directly to a raw disk partition. I can't imagine that it's the only db to do so, either.

    27. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Why?

      I once knew a CompSci prof that claimed that every computer science problem was a database problem.

      While certainly tongue-in-cheek, he did mean to imply that many or most problems in computer science that deal with structured data (i.e.: almost all of them) can benefit from a proper data storage/retrieval mechanism.

      This includes "file stores"/filesystems. I can only imagine the day being able to perform proper set-theoretic operations on a collection of files. If you can't imagine why it's better, you should probably read a database fundamentals book.

    28. Re:Future is relational databases by wintahmoot · · Score: 1

      Shameless plug: You might want to check out kSpaces. kSpaces lets you manage metadata that is associated with your files. It's not as neat and clean as direct integration into the filesystem, but it works today, and sure as hell beats waiting for Longhorn. :)

    29. Re:Future is relational databases by Suidae · · Score: 1

      Oh, right, because any relational DB is always fast and stable. right.

      It was slow and crashed because it was a POS implimentation.

      Seriously though, what would the user experiance on a system built on a database file system be like? Mostly I see this as useful for data that requires/benefits from lots of metadata, like MP3 files where I want rich tagging. But would the actual file contents be stored off in another table (the 'MP3FRAME' table for example, which contains all the frames of all MP3 files stored on the drive).

    30. Re:Future is relational databases by zurab · · Score: 1
      There are some performance issues to look at not to mention that even if the relational database uses a raw partition it is still use some sort of filesystem to store its data on disk and to manage free space.

      OK, IANAFE (filesystem expert).

      It wouldn't need a full-blown filesystem. That "filesystem" would be extremely fast and simple for storing and retrieving data. The database itself would be the real filesystem in terms of abstraction.

      From the book:

      The two fundamental concepts of any file system are files and directories.

      Why can't those be files and categories? That way each file can belong to one of more categories, completely indexed with the help of a database, with proper security, and search capabilities.
    31. Re:Future is relational databases by Carewolf · · Score: 1

      Because large parts of a database is more like a shell or a compiler and doesnt belong in a kernel. Ofcourse if you argue bash and gcc is part of the _GNU_/Linux OS, then a database can be too. I just wouldnt want to be the one to upgrade it ;)

    32. Re:Future is relational databases by DeSigna · · Score: 1

      In which case, won't all these optimizations and tips come in handy, or we'll have another inefficient, slow FS ?

      Btw, there are a fair few database systems that do read and write raw partitions.

    33. Re:Future is relational databases by Vancorps · · Score: 1
      You take a very large performance hit and you lose some redundency since the rollback logs aren't kept for any longer than one transaction. Generally it is better to keep every thing on a filesystem that can handle large files.

      Perhaps in the near future will remove the filesystem as a layer of abstraction, but not yet.

    34. Re:Future is relational databases by leandrod · · Score: 2, Informative
      > any relational DB is always fast

      Know you are joking, but for the benefit of others:

      A DB isn't fast, it is just stored at a DBMS. Now a RDBMS can be faster than anything else, because of data independence, that is, the separation of the logical and physical levels of any given database.

      > a POS implimentation

      What does that mean? OneLook gave me so many meanings, I gave up.

      > what would the user experiance on a system built on a database file system be like

      Anything you'd like. Hierarchies can be quite nicely stored in a RDBMS, so one does not need to change much in the user level.

      > would the actual file contents be stored off in another table

      It would be stored in a relation. RDBs don't have tables.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    35. Re:Future is relational databases by LWATCDR · · Score: 1

      The main issue here is not if this is possible but is it going to happen.
      The amount of just I am sure I am right and all others are wrong is mind numbing. I can see how a file system based on a relationl database could work. I am just not sure that it is the ultimate solution.

      As to it still requireing a file system. Yes even if they database uses a raw partiton it will need to keep track of free and used blocks. To me this is a filesystem.

      As to OO databases already being considered useless. So far you are correct from what I have seen. What I was trying to say is that there may be an even better model for a file system then a relational database and we should keep our minds open.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    36. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Wow. Not just monkeys. Not just colliding monkeys, no!

      But, supercolliding monkeys. What a glorious mess.

    37. Re:Future is relational databases by leandrod · · Score: 1
      > large parts of a database is more like a shell or a compiler and doesnt belong in a kernel

      A DBMS can have an API like anything else, no need for shells.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    38. Re:Future is relational databases by Anonymous Coward · · Score: 0

      I agree with you and am considering building a simple one as an experiment. In my model there would only be associations (a group of files/records which are associated, with no file marked 'special') and dependancies (a group of files/records which are associated, with a specific file being marked 'special', so that dependancy trees can be constructed) along with an arbitrary key/value pair list for each association group (of either type), and another arbitrary key/value pair list for each file. Groups can only hold records/files, not other groups. A file is considered unique if it differs in any key/value pair, OR in the combonation of groups to which it belongs. Files which find themselves without any dependants are unlinked, their space returned to the database. Of course one file can belong to an arbitrary number of groups, in any mix of dependancy groups and associative groups, and any group can hold an arbitrary number of files. I believe this is sufficiently simple to be powerful, yet sufficiently complex to be useful. There would be no structure for heiarchy, except the implied structure of heiarchy in the "dependancy" type group.

      There would be a layer over the filesystem to give meaning to specific keys and to implement ownership and security and so on. Standard keynames such as "string name" for filename, for example, "string owner" for the owner, "boolean readable" for it's readability. OS would only call this layer, the filesystem layer itself would never be directly called. Each OS could implement its own set of "meaningful" keys. It would be theoretically possible to implement a standard heiarchal filesystem using this one (with the dependancy trees and "blocking" behavior at the layer above this one), but I have no interest in that.

      100% Unicode for all string values.

      My project would be open-source, would write directly to a random-access device, would export a complete API, and would have an example implementation. A strong searching mechanism would be applied and fast searching would be paramount, in both directions ("Which files does this group contain? Which groups does this file belong to? Which files have the key/value pair, "string owner", "Erik"?). It would be intended to act as a replacement for a filesystem, and would not depend on a filesystem to function. There's no technical reason however it wouldn't run alongside a filesystem on the same system. (I would imagine booting one's system on an ordinary heiarchical filesystem, but working with all of one's data (any produced work, non system files) on this system. Eventually we could have an OS designed to boot entirely on it and sort everything using it.

      In order to do it I need someone with filesystem experience and database experience. I have a small amount of database experience, I'm considering doing this to improve it.

      If anyone has any interest, reply to this and give me a way of responding to you so I can give you my e-mail and so on.

      Not sure how good my real-time collaboration for this would be (for example IRC and so on) because I live in Hawaii so my timezone is "a bit off" from everyone else's in the world.

      Some notes:

      The germs of this idea came to me before I'd heard of relational databases, when I had an Amiga. I was frustrated by the fact that if I installed 10 programs, and they all used let's say "mylib.library" which was stored in LIBS:, if I were to delete the 10 directories which held those 10 programs, the mylib.library would still be sitting there, useless, in the LIBS: I came up with this idea to solve it. Over the years I built experimental models, not of code, but what a filesystem would look like if it were organized in this way, instead of heiarchally. I came to the conclusion that it would be far easier to search and have far fewer objects (groups instead of directories) in total. It would be simpler, yet more useful.

      Erik

    39. Re:Future is relational databases by leandrod · · Score: 1
      > is it going to happen

      Not soon. There are currently several problems:

      First, OS research is stagnated. It tends to focus on low-level stuff like distribution or microkernels, or GUI stuff like Sun's. Anything else just doesn't happen, because it is considered that the POSIX model can't be overthrown, be it in its implementations like the GNU system or in its corruptions like MS Windows.

      That said, there is the GNU Hurd. I have a suspicion it is meant by Stallman to be a stepping stone from POSIX to a rosy Lisp future. I hope it, or something the like, can be used as a stepping stone to a bright relational future. Perhaps Gnome Storage or something the like could be a proof of concept to something even better.

      Second, there is currently too much data mistification. OO, XML, SQL... and too few people who really understand data. Worse, these few people tend to work with IBM mainframes or MS Windows. I doubt there are more than half a dozen hackers who do grok the relational model, and thus we have Opus and Duro, but they are not ambitious enough -- yet, I hope.

      Third, while current relational free software projects are simple data access libraries, there are at least two full RDBMSs in development -- Dataphor and Alfredo Novoa's. But none is done yet, both are MS.Net, and none is free.

      But it is not hopeless yet.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    40. Re:Future is relational databases by erikharrison · · Score: 1

      Well. I'll bite.

      The text isn't "backwards looking". It is a discussion of how one man implented a file system, and how filesystems are generally implemented. Dominic himself comments that his book is a textbook of sorts (or the textbook he wishes he's had in Graduate School, where he did FS work) and that BeFS is an attempt to implement already existing functionality of the BeOS (where they were using a true database as a file system) using tried and true methods. In fact he bemoans that he couldn't do any real research.

      Additionally, the relational model is a great way to organize your data. You want to know how x relates to y in your data set. A file system is a way to store and retrieve arbitrary, heterogenous data, with loose or no relationship. File system metadata should be arbitrary, extensible, and changable on the fly. I don't want a "MAIL:From" entry to persist in any sort of DB where said entry exists for the kernel image, or my C compiler.

      A file system is not a data set where a mathematical model can be used to express tight relationships between said data. The so called "database" like qualities of the BeOS were seperate from a true relational model. The true essence of the relational model is just that - the ability to talk about complex interactions and tight linkages between data, the fact that I can then use that to do things I'd like my filesystem to do, is not a reason to build all that ridiculous overhead, in terms of performance, disk usage, and potentially buggy implementation complexity into my filesystem.

    41. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Let me clarify a few things:

      Finite and specific types for the datatypes that a key/value pair can be (number, datetime, boolean, (unicode) string, positiveinteger).

      All key/value pairs and group information is stored in RAM. It is sent to disk only in the sense of data security (to protect from power loss) and during shutdown. The one exception is the key/value pairs which hold strings. Only the first 16 characters of a string value are stored in RAM. Keys are of type string, and are limited to 16 characters in length. When the filesystem is mounted, this information is copied from disk to RAM -- as needed. It's blocked together on the disk (and backed up elsewhere on the disk) such that a single read for some metadata can pull in other metadata likely to be associated with it, with no further time cost.

      There is a single anchor file in each filesystem. It cannot be unlinked, and it cannot depend on any other file (but other files can depend on it). This is to keep files from dissapearing automatically because they don't have dependancies. It is possible to create smaller anchor files which depend on the big anchor file and therefore will not dissapear. In this way for example you could have per-user anchor files. Downloading a file from the internet with your browser would automatically link it to the anchor file for your user, but installing software would not (except possibly the main executable). These are all implementation details though.

      Erik

    42. Re:Future is relational databases by VitaminB52 · · Score: 2, Informative

      Yes - Oracle can.
      Oracle also has an iFS, short for Internet File System. It's actually a filesystem inside the Oracle RDBMS, and can be connected to through HTTP, FTP, IMAP4, SMTP, SMB/NTFS, NFS, AFP, etc..
      So Oracle has an relational database that can be used as a file system.

    43. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Glovin!!! I love Frink :)

    44. Re:Future is relational databases by big+tex · · Score: 1

      POS = Piece of Shit.

      i.e., load of crap.

      --
      I think I need a new sig here.
    45. Re:Future is relational databases by leandrod · · Score: 1
      > It is a discussion of how one man implented a file system

      Without taking in account 30 years of research and development of the relational model.

      I can understand it may sound strange, but it has been that long since hierarchies became obsolete as a way of organising -- not necessarily of presenting -- data.

      > the relational model is a great way to organize your data. You want to know how x relates to y in your data set

      No, that's a relationship, not a relation. The relational model isn't about relationships (which aren't even stored), but relations.

      > A file system is a way to store and retrieve arbitrary, heterogenous data, with loose or no relationship.

      You are pressuposing file systems are hierarchical. OK, the relational model allows one to store a hierarchy at a more efficient and powerful way.

      > The true essence of the relational model is just that - the ability to talk about complex interactions and tight linkages between data

      Not at all, the relational model simplifies all this complexity. It all becomes simple relations, without relationships.

      > is not a reason to build all that ridiculous overhead/i>

      You are thinking SQL, not relational. Relational is simple and implies little overhead.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    46. Re:Future is relational databases by Anonymous Coward · · Score: 0

      I have a feeling that if there is going to be a jump from the traditional file system to another approach, it will end up being native xml storage. XML structure maps well to a traditional filesystems folder and file structure. Any file attributes would be stored as XML attributes. XPath can solve most basic forms of searches you would want to execute. There is also a lot of work being done on encryption of xml data using all sorts of encryption algorithms.

      I have worked with relational databases for simple and more complex projects and know first hand how valuable a well thought out schema is. If you try to make changes later on, it can easily turn into a clunky mess. I think XML will be an incredible way to represent the data stored on your hard drive due to its flexibility.

      Probably not fast enough yet for servers, but for desktops I think the current implementations avaible from xmlsoft.org are fast enough. XML is the one way all data can be stored, represented, validated, converted and accessed from most compiled and scripted programming languages.

    47. Re:Future is relational databases by drfreak · · Score: 1

      Generally you would store you rollback logs on a separate physichal disk anyways.

    48. Re:Future is relational databases by the_greywolf · · Score: 1

      i honestly don't see why the future is so written in stone as being "relational databases."

      BeFS has all of the capability that a relational database has in a hierarchical arrangement with POSIX properties.

      i don't see any need for anything more powerful than that.

      what is it about relational databases that makes it so critically important? why isn't BeFS good enough for you? (as a side note, have you read the book?)

      --
      grey wolf
      LET FORTRAN DIE!
    49. Re:Future is relational databases by leandrod · · Score: 1
      > why the future is so written in stone as being "relational databases."

      Because the relational model is the only real data model in existence, giving us the power and performance we need by being simple, expressive, and by separating the logical from the physical level.

      > BeFS has all of the capability that a relational database has in a hierarchical arrangement with POSIX properties.

      That is part the problem. More powerful views of files should be possible.

      > have you read the book?

      No.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    50. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Please, in the future, preface your posts with:
      Warning: I am a pedantic relational asshole and will quibble over everything db related, including insisting that SQL and relational have nothing to do with each other, but I won't spell it out until you are all begging for mercy for me to stop my quippy, dismissive, curt, not-backed-up posts and replies.

      The relational model is the only salvation for computing - Join me in waiting for the mothership, won't you?

      --

      Sybase Certified DBA
      OCP (though I don't admit it)
      DB2 DBA
      Formerly Informix DBA
      MySQL DBA and developer
      Postgres DBA
      backplane user

    51. Re:Future is relational databases by LWATCDR · · Score: 1

      I have to agree that many parts of CS is not just stagnate but pretty dull.
      OSs, ISA, and GUIs are pretty much stagnate.
      I was pondering this relational database as a file system. But I was wondering how do you map files and directories to tables and tuples?

      Yea Hurd is... Well not moving fast. I know Unix is good but is it the best that we can do? I wonder if something like Plan9 or openmosix is an interesting direction to look in. Lets face it networks of computers is the dominate system and is likeily to get more so in the future. What I worry about is that OSs will eventualy be so burdened with Security functions that they will spend all there time be secure and very little time running apps

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    52. Re:Future is relational databases by Anonymous Coward · · Score: 0

      I doubt there are more than half a dozen hackers who do grok the relational model

      For some reason, I don't think this is the kind of accusation you can just throw around without some kind of backup. Why don't you explain some of these concepts that the rest of us are "just not getting"?

    53. Re:Future is relational databases by leandrod · · Score: 1
      > I was wondering how do you map files and directories to tables and tuples?

      First, it is relations, not tables. SQL has tables, but it is not relational.

      Then, remember a directory is also a file. Just a special type of file.

      In a very simple logical model, one would have a relation Files (name, type etc), and another Hierarchy (name, name_father). The Hierarchy relation would have just the pairs of files and their parents. The parentless file is the root directory.

      This is done everyday all over the world, even if SQL lacks proper support.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    54. Re:Future is relational databases by Anonymous+Cow4rd · · Score: 1

      Heh, I've noticed a large number of low UID users are incredibly unintelligent so don't let that fool you.

    55. Re:Future is relational databases by Anonymous Coward · · Score: 0

      You are thinking SQL, not relational. Relational is simple and implies little overhead.

      No, actually I was thinking asshole.
      Please enlighten us Mr. Pompous Asshole, what is the magical, low overhead, fast, light-weight non-SQL, relational language and implementation that you are so coyly alluding to?
      Current RDBMS's and their SQL-based implementation of relational calculus are the best people could do so far, unless Codd and Date are like the Oil and Auto companies - buying up the best relational systems and hiding them, least their lucrative relational model (as well as the extended relation model) monopoly is rendered worthless!
      Please fucking tell us since you have been dancing around this since your first post on this story.

      I am hereby officially biting.
      Show your hand, troll boy.

    56. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Apparently, its all hot air - but air that was efficiently heated by his secret relational (non-SQL) system.

      And of course, it is the data store chosen for Duke Nukem Forever, but they got stuck during development because he wouldn't allow them to store data in tables, they had to use relations , but he wouldn't tell them how to do that...

    57. Re:Future is relational databases by dmccunney · · Score: 1

      It's not. Doing that used to be a feature of the Unify RDBMS on Unix platforms. You could choose to place it on a raw device and bypass the Unix filesystem for performance. You needed a fairly big DB to justify it, but it wasn't hard to do.

    58. Re:Future is relational databases by dmccunney · · Score: 1

      Why relational databases? Why not a post relational design? Why not an object oriented database? Everything doesn't map cleanly to the relational model.

      Some folks may recall the file systems used on IBM mainframes, where awareness of what various file system objects were was part fo the OS. (VSAM, anyone?) For that matter, see PickOS, a relational DBMS built into the operating system, with a BASIC like language for development as well. A version of Pick was available as a layered product on top of Unix the last I knew, and a PC version called Revelation was also offered.

      PickOS grew out of a US government project, but it isn't the only such thing. The IBM System 38 was a lot like that, and the current AS400 is, too.

      The predominant Unix model seperated the file system from the application. Unix neither knows nor cares what a file contains. MS-DOS folloewd that model, and Windows did too.

      Future Windows versions apparently feature a file system based on SQL Server (which itself derives from Sybase). It will be interesting to see how that works out.
      ______
      Dennis

    59. Re:Future is relational databases by dmccunney · · Score: 1

      A file system is a logical construct layered atop a physical medium. The disk controller sees things in blocks, clusters, and cylinders, but the logical view can be what you like.

      There's no reason you *couldn't* use an RDBMS as a file system, and some RDBMS's will let you place the DB on a "raw" device, and provide thier own low-level calls to access it, and bypass the overhead of an OS file system.

      And there are systems now where the DBMS is built into the OS instead of being a layered product. See IBM's AS400 for a current example.

      Whether it makes sense depends upon what you need to do. The Unix model has the virtue of simplicity, but isn't necessarily the only way to do things.
      ______
      Dennis

    60. Re:Future is relational databases by leandrod · · Score: 1
      > Why relational databases?

      Because there is nothing else good enough.

      > Why not a post relational design?

      Because there is not quite such a thing... all are actually worse, lacking the fundamentals.

      > Why not an object oriented database?

      It is a 30-years throwback to pre-relational systems.

      > Everything doesn't map cleanly to the relational model.

      You are thinking SQL. The relational model really can map anything.

      > see PickOS, a relational DBMS built into the operating system

      Pick is MV, not relational. Check news:comp.databases.theory for more on that, or DBDebunk.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    61. Re:Future is relational databases by Anonymous Coward · · Score: 0
      WTF?
      WTF is right, don't be confused he is being pedantic to try to muddy the waters...
      but you are slightly mixed up about the term relation as used in the relational model

      From his seminal paper:
      A Relational Model of Data for Large Shared Data Banks, by E. F. Codd
      in Communications of the ACM, Vol. 13, No. 6, June 1970, pp. 377-387.
      (Copyright © 1970, Association for Computing Machinery, Inc.)
      <longquote>
      The term relation is used here in its accepted mathematical sense. Given sets S1, S1, , Sn, (not necessarily distinct), R is a relation on these n sets if it is a set of n-tuples each of which has its first element from S1, its second element from S1, and so on. We shall refer to Sj as the jth domain of R. As defined above, R is said to have degree n. Relations of degree 1 are often called unary, degree 2 binary, degree 3 ternary, and degree n n-ary.

      For expository reasons, we shall frequently make use of an array representation of relations, but it must be remembered that this particular representation is not an essential part of the relational view being expounded. An array which represents an n-ary relation R has the following properties:

      1. Each row represents an n-tuple of R.
      2. The ordering of rows is immaterial.
      3. All rows are distinct.
      4. The ordering of columns is significant - it corresponds to the ordering S1, S1, , Sn of the domains on which R is defined (see, however, remarks below on domain-ordered and domain-unordered relations ).
      5. The significance of each column is partially conveyed by labeling it with the name of the corresponding domain.
      </longquote>
      He goes on to describe a relational system for parts inventory, 34 years later we would consider some of the ideas fairly obvious.
      People who live in the real world where data has to be stored and used put their tuples of data in a fucking table that represents the relational mathematical relation that the table name intends.
      Its like insisting that people use tuple and on top of it intimating that there is something special and magical about tuples vs rows or records.
      Using your example, it is more like him saying cars don't have motors they have ICE's (Internal Cumbustion Engines) and then in his other posts he says ICE's are much more beautiful, low overhead and efficient than motors...
    62. Re:Future is relational databases by Anonymous Coward · · Score: 0

      You're wasting your time. leandrod has not responded with a single informative or useful comment to anyone in the entire thread, yet he has responded to almost everyone.

      All he seems capable of is regurgitating the words of Fabian Pascal, harping on semantics, and disappearing whenever people want some actual information.

    63. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Warning: buzzword central!

    64. Re:Future is relational databases by Anonymous Coward · · Score: 0

      OO has already been laughed off the contest.

      I realize you are too self-important to "do anyone else's homework" or offer anything more than short quippy responses but instead of putting up a bunch of links to the front page of your friends (crappy, not much unpaid content) website, can you please link to the actual page that "laughs OO off the contest"?

      TIA

    65. Re:Future is relational databases by Anonymous Coward · · Score: 0

      Mod parent up. +5 true.

      I don't understand why all leandrod are getting modded up, he hasn't posted anything insghtful, interesting, funny or informative... just vague, more relational than thou crap...

    66. Re:Future is relational databases by ynohoo · · Score: 1

      The Unix model has the virtue of simplicity

      roflm

    67. Re:Future is relational databases by Anonymous Coward · · Score: 0
      I thought your daddy Pascal doesn't like NULLs so how does your "tuple" in the Heirarchy "relation" not have a name_father value - OMFG! its not "Pascal approved" relational! Call the pomposity police!

      Or the leandrod response to this post would be:

      The parentless file is the root directory.

      That has a missing value, ergo its not relational, ergo its SQL, ergo bad, ergo my daddy didn't teach well enough how to be a pompous, pedantic, asshole I still in training.

      Actually that is too long of a sentence for a smug leandrod response...
    68. Re:Future is relational databases by Anonymous Coward · · Score: 0

      When you can show us a relational database as a filesystem then we can talk again. Until then, keep your pretentious claptrap to yourself.

    69. Re:Future is relational databases by Suidae · · Score: 1

      would the actual file contents be stored off in another table

      It would be stored in a relation. RDBs don't have tables.


      I'm having difficulty picturing this.

      'Relational Database' typically refers to something like DB2 or MSSQL that support features like foreign keys and referential integrity, allowing one to create relations between tables.

      It sounds like you are taking about some kind of abstraction of the relational concept. Thats fine, there are ways to represent data other than the table format most systems use, but I don't understand how you are suggesting it be organized.

      I know that many database systems support raw partion access for disks, and that a traditional filesystem could be implimented in such a database, but I don't see why this would be as much and improvement as some people suggest.

    70. Re:Future is relational databases by LWATCDR · · Score: 1

      Okay. For me and I think a lot of the world for right or wrong SQL=A Relational Database. I stand corrected.
      While this logical model works it still leaves out the nuts and bolts of the database/filesystem. What index system do you use how do you handle files as they grow in size? Even forgeting these issues let me see if I can translate this into an SQL.
      You would have a table of files that would contain name, mime type, dates..
      You would then have a table that defines the directory stucture. You could then have say a table of mime types and a table for each mimetype that contains mime type specific fields.
      For security you could use views for each user.

      It would also seem to open up the possibility to make a file of the type table. It would make sense to use the database in all sorts of programs. So basicly you would have a unified database for the entire OS.
      Yea I can see now that would work. I can think of some pratical issues with things like random access to the data in a file. I would assume that the data would be stored in a blob. Would that entire blob have to be read into memory when you did a seek. While the average memory has grown there are still some HUGE files that often get read in a line at a time and then the data writen out a line at a time.
      It make an interesting project to create a file system for Linux that uses a postgres or maybe SapDB server as a back end just to get a feel for it.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    71. Re:Future is relational databases by leandrod · · Score: 1
      > For me and I think a lot of the world for right or wrong SQL=A Relational Database

      Yeah, IBM brainwashed the world on that some twenty years ago, together with Oracle. That was one of the reasons for Dr Codd, the creator of the relational model, to leave IBM.

      > While this logical model works it still leaves out the nuts and bolts of the database/filesystemDatabase

      That's the beauty of it. By fixing an interface language at the logical level, the implementors have total freedom at the physical level. They can use tables, hiearchies, pointers, hashes, indexes, whatever, as long as only relations and constraints appear to the userland.

      > Would that entire blob have to be read into memory when you did a seek.

      Not necessarily... perhaps, but then perhaps the file contents could be parsed for the search engine's benefit when the file was created or every early morning or whatever. This wouldn't be much different than today's systems.

      > It make an interesting project to create a file system for Linux that uses a postgres or maybe SapDB server as a back end just to get a feel for it.

      That's Gnome Storage for you. But SQL has way too much limitations, the real thing would be something like Opus, Duro, Dataphor or the like at the backend.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    72. Re:Future is relational databases by leandrod · · Score: 2, Insightful
      > 'Relational Database' typically refers to something like DB2 or MSSQL that support features like foreign keys and referential integrity, allowing one to create relations between tables.

      There are some misunderstandings here.

      First, both IBM DB2 and MS SQL Server are based on the SQL standard -- actually IBM DB2 defined the standard --, and SQL is in frontal violation of serveral fundamental characteristics of the relational model. It does, however, incorporates some relational ideas and is generally saner than alternatives like XML, MV, OO DBs.

      Second, foreign keys are constraints to enforce referential integrity. This does not 'create relations between tables', it documents and enforces some relationships, and does that non-exclusively. The user is free to query the data and find any other relationships.

      Third, relation has a very definite mathematical meaning. It is the defining feature of the relational model, and corresponds to what SQL calls tables -- only that relations are sets, while tables are bags.

      > I don't understand how you are suggesting it be organized

      One can still have a hierarchy, but it would be just one way of accessing files, not the way. The position(s) of a file in the hierarchy would be just yet another attribute of it.

      --
      Leandro Guimarães Faria Corcete DUTRA
      DA, DBA, SysAdmin, Data Modeller
      GNU Project, Debian GNU/Lin
    73. Re:Future is relational databases by dmccunney · · Score: 1

      Compared to things like IBM OS/MVS, or VAX VMS, why yes, Unix *does* have the virtue of simplicity.

      Rather less than it once did, but still present...
      ______
      Dennis

  4. Need more than one filesystem by Anonymous Coward · · Score: 5, Informative
    All you need is one!

    No. You need four:

    1. Ext2 or Ext3 on primary hard drive
    2. IS09660 on CDs
    3. minix for initrd
    4. FAT for floppies/USB devices
    1. Re:Need more than one filesystem by Short+Circuit · · Score: 4, Insightful

      You know, I hope nobody ever tries a "one size fits all" approach with filesystems.

      Some are good for a specific purpose, like ISO9660. It's got error correction in it, so it's good for write-once removable media.

      Others are extremely widespread, like FAT. That's good for interoperability.

      Others support advanced permissions like ACLs. That's good for system security.

    2. Re:Need more than one filesystem by Anonymous Coward · · Score: 1, Insightful

      How about Ext2 for /boot and ReiserFS for everything else on the primary hard drive? :)

    3. Re:Need more than one filesystem by dasmegabyte · · Score: 1

      I'd trade Ext3 for Reiser, which is B-tree based, extremely scalable, has support for metadata journaling and outperforms Ext* for quick access to small files (such as web pages, perl scripts and config files).

      And let's not forget UDF for our DVDs.

      --
      Hey freaks: now you're ju
    4. Re:Need more than one filesystem by Etyenne · · Score: 1
      3. minix for initrd

      Huh ?!? My initrd is an ext2 fs ...

      --
      :wq
    5. Re:Need more than one filesystem by SmittyTheBold · · Score: 1

      Unless I'm mistaken, it's the CD media writing process that involves error correction, not the ISO9660 format itself. And mistaken I could be.

      --
      ± 29 dB
    6. Re:Need more than one filesystem by Short+Circuit · · Score: 1

      There are ECC blocks in the filesystem, which help correct corrupt data. Including corruption that happened after the data was written.

    7. Re:Need more than one filesystem by SmittyTheBold · · Score: 1

      Is it actually in the FS though? I don't think so. I can put four moderately-sized files (~125 MB) into an ISO file, with a net gain of 50 KB or so over the raw files. You can't tell me besides containing the actual FS info that image includes ECC.

      The reed-solomon encoding/decoding is all taken care of at the hardware level by the CD drive, unless you choose to record in DAO/96 mode or something similar, where the recording software takes charge of subcodes.

      It was always my understanding that ISO9660 was a sort of lowest-common-denominator standard designed to be useful on most computer systems at the time of its design. Most systems have moved on from where they were years ago, so companies have added extras (Windows Joliet support, Rock Ridge, High Sierra, etc.) as the years have gone on. The original idea, though, was to be a sort of minimum standard so as to avoid problems with different computers.

      In a Mode 1 CD the drive hardware takes care of the ECC. You feed the drive 650MB (or, recently, 700MB) of data, and it generates error correction as it burns. Mode 2 removes the ECC for slightly more space. This is useful for things like VCDs, where a missed bit is easily concealed, but an extra 300 bytes per sector adds up to significantly more space.

      This is, interestingly enough, the secret used in many copy protection schemes. The original CD was mastered with invalid ECC entries on specific sectors. Your computer would correct the data when reading/ripping the image. Then, when you burned the CD with "correct" ECCs, the software simply had to look for the sectors it knew should be bad. If they looked pristine, then the software knew you were running a copy.

      So, my summary: ISO9660 is a crappy FS that's been extended far too much, but much like FAT it's here to stay.

      --
      ± 29 dB
    8. Re:Need more than one filesystem by Short+Circuit · · Score: 1

      Whoops. You're right. ECC isn't part of the ISO image, but part of the CD file format.

      But that doesn't explain why I couldn't read the data off (though I could mount it) CD I accidentally burned with the Apple filesystem.

    9. Re:Need more than one filesystem by SmittyTheBold · · Score: 1

      You're right, that doesn't. BUT, I've used a BFS CD before, (with BeOS of course) so using filesystems other than ISO9660 is not unheard of. I could have sworn most Apple CDs were HFS as well. In fact, I'm almost positive - that's why some nifty tricks, like building pictures out of strategically-placed cutom icons for empty folders - worked. You wouldn't be able to do that with any of the ISO formats, I'm almost positive.

      --
      ± 29 dB
  5. It's a nice thought. by DiscordOfFive · · Score: 5, Interesting

    I generally think it's a good thing when books get put up (legally) on the web like this. My personal view on BeOS is that it's more of a hobby OS than a production one, but a book that details the workings of a complex system is useful. After all, maybe it'll help form the basis of a new, advanced FS.

    --


    Only the purest of souls seek enlightenment. Everyone else just wants power.
  6. Without being too cynical by Timesprout · · Score: 4, Insightful

    He does not have the gift that some tech writers have of making both an interesting technical document and a fun read.

    But for most people this is seriously dull subject matter. Oscar Wilde would stuggle to get a chuckle out of this stuff.

    --
    Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
    What truth?
    There is no dupe
    1. Re:Without being too cynical by Anonymous Coward · · Score: 0

      Oscar Wilde would stuggle to get a chuckle out of this stuff.

      I think Oscar Wilde would be more likely to struggle to get out of his coffin... :)

    2. Re:Without being too cynical by daeley · · Score: 1

      "There is no such thing as a moral or an immoral book. Books are well written or badly written."

      --
      I watched C-beams glitter in the dark near the Tannhauser gate.
  7. Mirrors: by Anonymous Coward · · Score: 3, Insightful

    Mirrors of the PDF:

    http:/beos.spb.ru/program/105/practical-file-sys te m-design.pdf.zip

    http://users.aber.ac.uk/mmb9/data/practical-file -s ystem-design.pdf

    http://www.funtech.org/downloads/Temp/practical- fi le-system-design.pdf

    1. Re:Mirrors: by Anonymous Coward · · Score: 0

      Take out some of the spaces the slashdot engine put in the URLs in order to get to the file.

    2. Re:Mirrors: by uberdave · · Score: 3, Informative

      Links to the mirrors:

      Link 1

      Link 2

      Link 3

      Really folks, is it so hard to throw a <a></a> tag around a url?

    3. Re:Mirrors: by WhoDang · · Score: 1

      Its always nice when someone makes it a easier for us lazy folk.

    4. Re:Mirrors: by EvanED · · Score: 1
      In this case, no. But often, it's also a matter of tossing in

      or
      tags around the paragraphs so the breaks are preserved and remembering to chose HTML from the format menu.

    5. Re:Mirrors: by UserGoogol · · Score: 1

      You oughtta try out plain-old-text mode. There must've been an update or something, it's quite good now. It handles paragraphs just fine AND automatically parses HTML quite nicely. Although it does some weird stuff with blockquote.

      --
      "Never attribute to malice that which can be adequately explained by stupidity." -- Hanlon's Razor
  8. Mostly a good book, but one comment..... by SollyCholly · · Score: 1, Redundant

    Perhaps the most annoying thing about this book though is that he doesn't finish his thoughts. I felt that often, just as he was getting to the interesting part after cutting through the fluffy descriptions of his design choices, he would leave the topic and not come back. The must frustrating part of this was that after skipping over many pertinent details of how he actually built the BeFS, he spends an excruciating amount of time describing the vnode layer and the exact API that the file system driver must write too -- something I feel would have been better left to a Be-specific API programming manual.

    1. Re:Mostly a good book, but one comment..... by Anonymous Coward · · Score: 0

      This is just cut 'n paste from the amazon.com customer review here by Paul R. Nash. (second paragraph of paul's review)

    2. Re:Mostly a good book, but one comment..... by NetNinja · · Score: 1

      Excellent point! That's why it's not a mainstream OS.

  9. Silly submitter... by grub · · Score: 5, Funny


    the BeOS rising to join the ranks of OSs that won't die

    An OS isn't dead or dying until Netcraft confirms it.

    --
    Trolling is a art,
    1. Re:Silly submitter... by nocomment · · Score: 2, Informative

      and since Be lives on, it will probably be awhile before netcraft confirms it. At least Yellowtab, is releasing something whereas amiga hasn't released anything tangible (although they say they have) since os v4.

      --
      /* oops I accidentally made a comment, sorry */
      /* http://allyourbasearebelongto.us */
    2. Re:Silly submitter... by Power+Everywhere · · Score: 1

      This is a nice effort, but I wish the various splintered post-Be BeOS update communities would unite. Precious skills and time are being pushed in different directions. Short of Palm throwing one of the groups an official bone, I think it would make more sense for the (what is it, three? four?) groups working on BeOS hacks to start working together.

    3. Re:Silly submitter... by nocomment · · Score: 1

      actually be threw yellowtab and "official bone" right before palm bought them. A little zeta history.

      <snippet>
      Before Be, Inc. sold its assets to Palm, Inc., we managed to close a deal allowing us to distribute the PE version and had started negotiations over the future of the Pro version. Koch Media was ready to reissue the copies of the Pro Version that they hadn't managed to resell, to make it a part of a new distribution first called BeOS NG (New Generation), now renamed to "Zeta".
      </snippet>

      --
      /* oops I accidentally made a comment, sorry */
      /* http://allyourbasearebelongto.us */
    4. Re:Silly submitter... by nocomment · · Score: 1

      oh and ps, I think it's 4 ;-)

      --
      /* oops I accidentally made a comment, sorry */
      /* http://allyourbasearebelongto.us */
    5. Re:Silly submitter... by Power+Everywhere · · Score: 1

      This was a nice gesture, but where is Be now? Nowhere. Its assets are owned by Palm and Palm isn't extending this favor to Yellowtab, or anyone else, sadly. :_(

    6. Re:Silly submitter... by Power+Everywhere · · Score: 1

      Forking makes Jesus cry. :_(

    7. Re:Silly submitter... by nocomment · · Score: 1

      Jesus has bigger things to worry about than that. :-)

      --
      /* oops I accidentally made a comment, sorry */
      /* http://allyourbasearebelongto.us */
    8. Re:Silly submitter... by commodoresloat · · Score: 1

      Actually, Netcraft only confirms what is already present in the most reliable data measure we have about the life of an OS: The number of posts about it to usenet. So always check usenet postings before shouting "Red ink! River of blood!"

      Also, it would help to be aware of how long it takes to copy a 17 M file from one disk to another under the OS.

    9. Re:Silly submitter... by Anonymous Coward · · Score: 0

      "At least Yellowtab, is releasing something whereas amiga hasn't released anything tangible (although they say they have) since os v4."

      You're confused. v4 has just been released to beta developers, before that it was 3.9. Neither, incidentally, were released by Amiga. As far as I can tell BeOS is as dead as Amiga; which is to say, not at all because people still use it.

      I'll never understand why trolls like yourself spend so much time moaning about other people's choice of operating system.

  10. Just what I always wanted by Anonymous Coward · · Score: 5, Funny

    A practical file system with an impractical OS!

  11. Holy Crap. I'm a Be-diot. by torpor · · Score: 4, Funny

    All those years I've slagged off Amiga blow^H^H^H^Hdie-hards, and here am I with a rev-a. BeBox I refuse to part with. I love the damn thing. I never noticed it, but I've got Amigazoids' Disease.

    Why, oh why can't someone invent a "Parallel Universe Chunnel" so I can get myself a laptop BeBox. tiBooks come close, but Apple crack is still crack!

    Ah well. At least I have something else (the .PDF) to put on my Amig^H^H^H^HBeBox, beautiful and blue though it sits in the corner, gathering dust, as if there's nothing else to do with it ... ;)

    --
    ; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
  12. ironic fact by Power+Everywhere · · Score: 5, Informative

    Apple considered buying Be a few years ago for BeOS and opted for NeXT instead. Now, years, later, they have hired several Be engineers to work on the Mac OS X filesystem. It looks like Apple is getting Be without having to buy the whole company. Be fans, look at Tiger as an upgrade for your favorite OS.

    1. Re:ironic fact by nacturation · · Score: 2, Interesting

      ...especially when you consider that both NeXT and Be were started by former Apple VPs (Steve Jobs and Jean-Louis Gassee).

      --
      Want to improve your Karma? Instead of "Post Anonymously", try the "Post Humously" option.
    2. Re:ironic fact by ckelly5 · · Score: 2, Interesting

      Yes, I loved BeOS back in its heyday. When I learned that Giampaolo had gone to Apple, I figured he'd be beefing up the FS. His first work was probably the Journaling FS that is prevalent in Panther, and the rumored metadata in Tiger has his name all over it...

    3. Re:ironic fact by Power+Everywhere · · Score: 1

      Good point, interesting coincidence, but Jobs was more than just a VP at Apple. He was founder and Chairperson of the Board.

    4. Re:ironic fact by Power+Everywhere · · Score: 1

      Take this with a large grain of salt, but supposedly the work on Mac OS Extended for Tiger is going to be pretty far-reaching and will be fleshed out for years to come. I've heard the name Mac OS Extreme thrown around for the update to HFS+.

    5. Re:ironic fact by shawnce · · Score: 1

      Mac OS Extended?

      You mean HFS+ (HFS Plus)? That is the name of the file system.

    6. Re:ironic fact by Power+Everywhere · · Score: 2, Interesting

      The original Hierarchal Filesystem (HFS), introduced well before System 6 was called "Mac OS Standard" after Mac OS 8.1. It was this same release that Apple released Hierarchal Filesystem Plus (HFS+), which they opted to call "Mac OS Extended." In Tiger, using the metadata and relational aspects tacked onto HFS+ will be called Mac OS Extreme when you go to format your drive and turn on different features of your filesystem.

    7. Re:ironic fact by iMacGuy · · Score: 1

      No, it's an HFSX filesystem. This can be already created in 10.3; it's the Case-Sensitive Mac OS Extended (newfs_hfs -s). There's an Apple technote about some of the new features that aren't implemented yet, but I don't know the number.

      --
      Why won't slashdot let me change my terrible username :(
    8. Re:ironic fact by scorpioX · · Score: 1
      No, he means something beyond HFS Extended, namely HFS Extreme. In fact, this new extension of HFS already exists in some form for Mac OS X Server. It is used to implement case sensitivity in HFS. See Apple's HFS Technote for the details. Here's a snippet:

      HFSX is an extension to HFS Plus to allow additional features that are incompatible with HFS Plus. The only such feature currently defined is case-sensitive filenames.

    9. Re:ironic fact by shawnce · · Score: 1

      Yes, I know that (it is HFS+ not HFS Extended).

      I was pointing out that the name he was using are not the normally used names (or correct) for those file systems and trying to make sure of what exactly we was talking about.

    10. Re:ironic fact by shawnce · · Score: 1

      Ah using the end user visible names... working so long below that level of things I forget some of the tools used that naming (used to seeing partition names, etc.).

    11. Re:ironic fact by Power+Everywhere · · Score: 1

      So sorry to intrude upon your eliteness with silly user-level nomenclature.

    12. Re:ironic fact by shawnce · · Score: 1

      Taking my comments the wrong way...

    13. Re:ironic fact by otomo_1001 · · Score: 1

      As an old BeOS user (R3 to the end) and new MacOSX Panther user. I can easily say MacOS X has a long way to go in the filesystem regard. (I refuse to speculate on unreleased products like Tiger)
      OT: How many more cats can they go through before they run out? I mean leopard, cheetah, what next siamese? :)

      I so miss my file finds on metadata. I had all my MP3's indexed with metadata. Soundplay (awesome mp3 player) would then just pull the playlists from the find, which updated immediately with any new files added.

      But the real fun thing was email in BeOS. It was so easy to find anything, a regular filesystem find was all you needed to search where something came from, when, etc... I had a very elaborate system setup using saved finds to not sort anything, everything was done by finds. /getting nostalgic and almost want to buy a old box to install R5 on just to relive the memories.

  13. Re:Ill shout Troll... by nazsco · · Score: 1, Funny
    Ill shout Troll..
    ...to anyone proclaiming this to be old news!
    No, you will not, sice you already posted in this thread.
  14. BeOS won't die? by Bruce+Perens · · Score: 1, Interesting
    BeOS won't die?

    What if there was a company that said it was using BeOS, and really wasn't, but just wanted people to think so until they were ready to release their product?

    Bruce

    1. Re:BeOS won't die? by Tin+Foil+Hat · · Score: 1

      Sorry? Where are you going with this? Why would anyone want to do that?

      --
      No matter how many of my rights are taken away, somehow I still don't feel safe. -Frigid Monkey
    2. Re:BeOS won't die? by Anonymous Coward · · Score: 0

      What are you talking about, stupid?

      And would anyone care?

    3. Re:BeOS won't die? by Bruce+Perens · · Score: 1
      I can't say any more, but you can connect the dots if you think about it for a while. Surely people here can help you.

      Bruce

    4. Re:BeOS won't die? by Anonymous Coward · · Score: 0

      Now you've got me all curious... Any hints?

    5. Re:BeOS won't die? by Bruce+Perens · · Score: 1
      You have all you are going to get out of me.

      Bruce

    6. Re:BeOS won't die? by Fuzzy · · Score: 1

      You mean like these guy?:

      http://www.yellowtab.com/products/

      You can never have too many OSs....

    7. Re:BeOS won't die? by CoolMoDee · · Score: 1

      maybe Zeta's Yellowtab? They claim to be using BeOS and such.

      --
      Jisho - A Japanese English German Russian French Dictionary for the rest of us.
    8. Re:BeOS won't die? by Anonymous Coward · · Score: 0

      To whom are you reffering Bruce? SONY, Palm, Gobe, YellowTAB, Edirol, Tascam?

      If you are reffering to SONY/PalmONE regarding PalmOS, please remember, everyone knows that PalmOS 6 is _not_ BeOS (the ex-Be engineers working at PalmSource didn't use BeOS code at all (just some ideas) and that's not a secret).

      I know of no one else who uses BeOS for high profile products, so I don't get what you are saying here. And why these cheap shots at BeOS? Jeez Bruce, you really need to get a life, you are really not as smart as you would like people to think you are.

    9. Re:BeOS won't die? by no+reason+to+be+here · · Score: 1

      Their project is definitely using BeOS, as far as I can tell; I haven't had a chance to use it up close and personal, but the screenshot look too much like screen shots of what Be, Inc. had developed in the last few months of their existence.

    10. Re:BeOS won't die? by Anonymous Coward · · Score: 0

      Sorry, but you can't be going anywhere with this. Palm has their new OS out and it's moved away from anything having to do with their BeOS IP, Zeta couldn't possibly be about to "not use" BeOS, Apple doesn't have any need for them, Gobe isn't claiming to support them anymore...

      Maybe if you said a company wasn't currently claiming to use Be Technology but was going to start, you'd have mystery, but this weird troll by a typically intelligent slashdotter makes no damn sense, since the *only* people actually claiming to use BeOS right now are Zeta, who don't exactly have a choice about "moving away" from that platform.

  15. Thanks For The Reminder by jacoby · · Score: 2, Funny

    I lent a copy to a friend a while ago. I just asked for it back.

    When's the last time you had a PDF book as bathroom reading?

    1. Re:Thanks For The Reminder by Anonymous Coward · · Score: 0

      When's the last time you had a PDF book as bathroom reading?

      Why just the other day I was reading a pdf on my Clie whilst on the loo.

    2. Re:Thanks For The Reminder by jacoby · · Score: 1

      I never did like the PDF reader on my iPaq. Perhaps, if it hadn't died, I'd have learned to love it.

  16. -1 Flamebait by Anonymous Coward · · Score: 0

    You cannot put HFS on a CD and expect it to work right.

    Linux has had a GUI for nine years.

    1. Re:-1 Flamebait by Anonymous Coward · · Score: 0

      HFS is the only filesystem that I have tried and failed to mount on my linux box. Where the hell am I going to find a Mac with a zip drive?

    2. Re:-1 Flamebait by Anonymous Coward · · Score: 0
      HFS is the only filesystem that I have tried and failed to mount on my linux box. Where the hell am I going to find a Mac with a zip drive?

      Huh? Pretty all Mac towers have a zip drive. Meanwhile, I have a USB zip drive for my TiBook -- you can borrow it if you want.

    3. Re:-1 Flamebait by Anonymous Coward · · Score: 0

      Cool, you mind if I come by and pick it up? Where do you live? I promise I will return it when done.

  17. BitTorrent by gspr · · Score: 4, Informative

    It may be redundant - if it is, just moderate it as such, but here is a Torrent, so that we don't completely destroy the nice BiggestPOS' mirror.
    I'll seed it for an hour or so.

  18. Interesting by 0xC0FFEE · · Score: 3, Insightful

    This seems like quite an interesting read. But what I wonder is if BeFS is encumbered by any patent... If so, who own them now, do they plan on enforcing it, etc. Would be oh so cool to have an open source BeFS implementation.

    1. Re:Interesting by Anonymous Coward · · Score: 0

      Yeah, if it can be made open source, that would be as near as you can get from an OSS _innovation_.

    2. Re:Interesting by stratjakt · · Score: 1

      You mean like OpenBeFS, of the OpenBeOS project?

      As far as BeFS plugging into linux any time soon, I'd doubt it. Linux has enough problems with ACLs, let alone metadata indexing, etc, etc.

      --
      I don't need no instructions to know how to rock!!!!
    3. Re:Interesting by 0xC0FFEE · · Score: 1

      The XOR algorithm applied to a cursor is (was) encumbered by a patent. Doesn't mean that because there's a free implementation (OpenBeFS) that it is clear...

    4. Re:Interesting by burns210 · · Score: 1

      Palm(palmsource, the OS company) owns all things Be. They would have control over any patents that Be had.

    5. Re:Interesting by Anonymous Coward · · Score: 0

      Be was filing for a replicant patent. they were (as far as I remember) not awarded the patent before Palm purchased the IP. (~) They did not have a patent on the BeFS.

      alphaseinor

    6. Re:Interesting by n3m6 · · Score: 1

      That begs the question. What are the patents on filesystems held by BeOS? Since Be has been bought out by Palm, wouldn't these patents be just another thorn in the sight for people who work on FSs. Wouldn't it be more morally right to just let go of those patents if Palm( who works on a totally different computer environment), does not require these patents.

  19. I love BeOS by helf · · Score: 1, Insightful

    I actually use it as my main OS. I still have windows 2k loaded on a second partition incase i want to play some games. But I use BeOS for 99% of my computer needs.

    1. Re:I love BeOS by Radon+Knight · · Score: 1
      I actually use it as my main OS.

      So, uh, what program do you use?

      Sorry, that was a cheap shot...

    2. Re:I love BeOS by harikiri · · Score: 2

      In 2000 I think, I actually spent a fair amount buying a copy of BeOS 5.0 and Gobe Office. It was a fantastic operating system - booted fast, had a bash terminal, gcc, etc, and everything seemed to work well.

      I think the main issue that Be ran into is that their vision was a little confuzzled. Anyone else who followed Be remember when they decided they wanted to drop the desktop market and move into set-top boxes? Perhaps if they'd managed to bring some of the big audio/video software companies on board and sold it as a one-stop solution for editing (because of their next-to-zero latency they advertised), they may have had a better chance. :(

      It's the last OS I payed for, and I still have a very scratched CD with 5.0 on it around here somewhere. :-)

      -- Fanboy!

      --
      Man watching 6 MSCE's around a sun box, looks alot like the opening scene's of 2001:space odyssey...
    3. Re:I love BeOS by CRYPTOFREQ · · Score: 1

      i actually bought the BeOS when it was being sold and absolutely loved it!! but i guess i enjoyed windoze too much that i eventually went back to it...do you use the gobe suite for your "office" needs i suppose? how about email? i'm going to buy a dedicated computer to run beos now...i'm sick of windoze and feel like cutting the tether.....any links to make things easier would be greatly appreciated....

    4. Re:I love BeOS by constant+imp · · Score: 1
      On one of my boxes (AMD AthXP 2000, ABIT board, Promise RAID disabled) I use a 'Developer's Edition.' It works great.

      There are workarounds for installing Be's final v5.0.3 on later AMD systems, but I gave up trying when the DE installed so easily. One of these days I'll figure it out....

      Zeta installed fine on the same box (sep partition), but I am having stability issues. There are supposedly 3 service packs since my version RC1, but I haven't been able to log in to download them.

    5. Re:I love BeOS by alphaseinor · · Score: 1

      First, I would download BeOS Max

      You may need to get the latest driver for your video card Nvidia or ATI

      After you do that, just go to "start" --> Preferences --> Display to change your graphics settings.

      I would also upgrade your sound card driver if you have a AC-97 Intel AMD

      BeOS has a built in E-mail system, the e-mail is saved as a attributed file in /boot/home/mail/in To configure, just go to the "start" menu --> Preferences --> e-mail you will notice a new icon next to the clock (looks like a US mailbox) you can access your e-mail from there.

      After setting it up download BeShare for all of your support needs (there are lots of files there as well)

      I would get the bleeding edge version of Firefox for a web browser

      if you are feeling adventureous, try finding and installing Bone7a (look for bone7a_install.zip on google or b7a.zip on beshare). This is a more stable network stack integrated into the BeOS Kernel.

      If BeOS isn't your thing, Windows or Linux are always out there.

  20. Re: only relational databases by Anonymous Coward · · Score: 0
    These days, the idea of using any other database model seems ludicrous.

    The network database model is STILL attractive when excessive speed is required (which is NOT most of the time).

    Keep an open mind, but look for good ideas in the junk heap.

  21. Like a warm hug. by hedon_elite · · Score: 2, Insightful

    Amiga is like being curled up in a rocking chair with your grandmother when you're 5 years old. Total comfort and contentment. Btw, the Amiga website just got Slashdotted. I bet its been a long time since that happened. Ray Akey's gonna be p*ssed =)

    1. Re:Like a warm hug. by Anonymous Coward · · Score: 0


      Ray Akey's gonna be p*ssed =)

      passed ?
      possessed ?
      possessedly ?
      possessedness ?
      prepossessed ?
      professed ?
      professedly ?
      prowessed ?


      Please clarify.

  22. Re:Holy Crap. I'm a Be-diot. by doctor1 · · Score: 1

    I do love my dusty BeBox too!!! I miss watching the dual LED towers pulse as it easily played multiple looping videos, played a music file, and the many other tricks that it did simultaneously. I'm hoping to get yellowtab.com's Zeta installed on my SuperMicro X5DAE with dual XEON 2.2GHz soon.

    --
    Astronauts in weightlessness of pixilated space, exchange graffiti with a disembodied race. - Rush
  23. But look at where he is now... by T-Kir · · Score: 1

    The reason he is at Apple is that he was brought in because of his work on the BeOS FS, so a lot of the ideas and inner workings (AFAIK) are being put into OSX.

    I would hunt some links out but I can't be arsed, but El Reg did a very good interview with Dominic a couple of years ago that was quite a good read.

    --
    Are you local? There's nothing for you here!
  24. Re:Tell me honestly by azav · · Score: 0, Offtopic

    Oh, yes, sorry. I did want to say something.

    You are correct. Your post does make you look fat.

    --
    - Zav - Imagine a Beowulf cluster of insensitive clods...
  25. Other possibility by mirko · · Score: 3, Informative

    Another OS which proposed a very ergonomical approach to file system design and implementation was RiscOS.
    Check its Programmer Reference Manuals if you can find these.

    --
    Trolling using another account since 2005.
  26. BeOS Books by TheBadger · · Score: 1

    I have "Be Developers Guide" and "Be Advance Topics". I've also got "BeOS 3.1" and "CodeWarrior 1.5 for BeOS". I think I also have BeOS 4 and 4.5 buried somewhere.

    If anyone is interested in taking them off my hands. Post a message to the guestbook of my website.

    1. Re:BeOS Books by MyDixieWrecked · · Score: 1
      I've got BeOS DR1, DR2, R3, R4, and R4.5 discs as well as the BeOS bible. I used to run it on my 7600/132 back in the day, and a dual 100mhz Umax box. The 6 second boot was a great thing, even off a zip disk.

      i used BeOS before I ever even touched linux, and it gave me quite a headstart on BASH programming and many of the concepts behind linux (plain-text programming and config files, etc).

      I also loved the API for it. It was amazingly simple to write GUI applications in C++. On the same level as Cocoa, here in OSX.

      When it was rumored that apple would buy Be, I was wishing so hard for that. the BeFS and their method of handling threads would have been a godsend to the MacOS. As well as that great multiple-virtual-desktop thing.

      --



      ...spike
      Ewwwwww, coconut...
  27. BeFS lives by no+reason+to+be+here · · Score: 4, Interesting

    OpenBeOS's clone of the Be File System has been selected recently by the folks creating the SkyOS.

    The BFS replacement has been one of the fastest progressing parts of the OpenBeOS project. Dominic Giampaolo has actually commented to the team lead of OpenBFS, and complemented the team on the good work they accomplished.

    1. Re:BeFS lives by Anonymous Coward · · Score: 0

      You are forgiven, but only because you used a homonym...

      complimented is the proper spelling of the word you are trying to use.

    2. Re:BeFS lives by BasilBrush · · Score: 1

      Well give that Dominic had more or less written a how to book for writing a BFS clone, that's hardly surprising.

  28. Re:BeOS won't die? Perhaps Palm. by guidryp · · Score: 1

    Didn't Palm make some vague statements about ongoing support. It seems very unlikely at this point.

  29. web designers by Anonymous Coward · · Score: 0

    Has anyone got a link to the original slashdot post that describes web designers as scooter-riding, latte-drinking, roll-neck-wearing tits who peruse Kafka whilst clicking aimlessly around in Dreamweaver? It is comedy cold and I need to post it on our internal newsgroup.

    TIA.

  30. Straight up: your sig is gay beyond belief by Anonymous Coward · · Score: 0

    "Bling bling"??? Do people still say that?

    1. Re:Straight up: your sig is gay beyond belief by jred · · Score: 1, Funny

      Fo' shizzle...

      --

      jred
      I'm not a mechanic but I play one in my garage...
  31. Drop linux development effort for open BeOS by Anonymous Coward · · Score: 0

    The never ending GUI debate would end.

    X would pass into BSD is dying status.

    A large amount of legacy application / OS baggage would no longer be dragging down forward movement.

    1. Re:Drop linux development effort for open BeOS by donscarletti · · Score: 1
      I seem to recall that OpenBeOS has its fair share of debates. Last I heard there is currently one about package management.

      Also, creating another platform to solve the problem of having too many different platforms is hardly a clever solution is it?

      Thirdly, aiming for full compatibility with multiple known weaknesses is hardly the best way to remove the issue of legacy systems.

      --
      When Argumentum ad Hominem falls short, try Argumentum ad Matrem
    2. Re:Drop linux development effort for open BeOS by Anonymous Coward · · Score: 0

      Firstly, the package management debate, AFAIK, has not been active for quite some time. People seem to have settled upon the basics of the system and prototypes have been in development for quite a while.

      Creating a new platform is sometimes necessary as its quite often not possible to cut away all of the cruft in a given system and be left with something useable, or the underlying methodology of a system may not have been thought out as clearly as needed. This way you can start out with a system that is known to be free of previous problems and does not have to suffer under the burden of support older applications.

  32. Sky Operating System by Anonymous Coward · · Score: 3, Informative

    The BFS lives on in a few other places, most notably in a rebuild by the OpenBeOS team. This file system is also being used by SkyOS for their new file system, SkyFS. They are already making use of the attributes in a number of ways.

  33. At Apple Now by wintahmoot · · Score: 2, Interesting

    One thing that didn't get mentioned is that the author of this book is at Apple today, and rumor is that he is working on Apple's version of WinFS, a metadata-driven filesystem.

    Be saw the potential of metadata on the filesystem level before everybody else did.

  34. Re:Can you actually read this? by Anonymous Coward · · Score: 1, Funny

    Yeah, actually. Some people find this stuff interesting.

    Clearly you don't. Oh, look - "Who Wants to be a Millionaire" is on! Why don't you run along and pretend to be smart, okay?

  35. wait a second... by veg_all · · Score: 1

    So first he doesn't get specific enough, but then he gets too specific? Solly, Cholly, but you're going to have to make up your mind what exactly it is you want!

    --
    grammar-lesson free since 1999. (rescinded - 2005)
  36. Alive? by moosesocks · · Score: 2, Interesting

    It's sort of hard for an OS to die which wasn't really alive to begin with.

    That being said, I remember using the last BeOS Personal Edition (the one that ran on top of FAT32).

    It was incredibly cool. There's definitely something to be said for an OS which boots almost instantaneously.

    The inability to print was somewhat of a drawback...

    --
    -- If you try to fail and succeed, which have you done? - Uli's moose
  37. Re:Holy Crap. I'm a Be-diot. by torpor · · Score: 1

    i'd just be happy to get some sort of linux/*bsd on it somehow ... i don't know what version of BeOS i've got on there now, but its not the 'latest-and-greatest', which i fear is only available to me through some sort of "Wayback Machinations" or other ... but if there were at least a linux kernel on it, then i could fit it into my other 'port/hack/compile/rewrite' style of sysadmin'stration...

    --
    ; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
  38. Re:CUM by Rev+Saxon · · Score: 1

    Why is it someone always feels the need to post a link to that in every article? I mean come on, you seen it once, you seen it a million times.

    --
    I am that much more enlightened and proportionally disillusioned
  39. You can do this by commodoresloat · · Score: 1
    1. Re:You can do this by torpor · · Score: 1

      yeah, tah, that rocks. i'll give it a go one spare weekend. i wonder if it still lets me use the blinkenlights... thats really about the only thing that truly endears me to the bebox... ppc603e's, yeah, okay, would've been good in 1995.

      i stand by my request for a bebox-like tiBook though. that'd rock ass.

      --
      ; -- the corruption of government starts with its secrets. a truly free people keep no secrets. --
  40. Explicate? by FireBreathingDog · · Score: 3, Funny
    His style is very straightforward - introduce idea, explicate idea, summarize idea.

    Isn't it ironic (dontcha think?) that 'explicate' appears in a sentence about straightforward writing? How about explain???

  41. MODERATORS by Anonymous Coward · · Score: 0

    While I agree with your assessment of the above post as offtopic and overrated, I think that you must have failed to notice that it was written by Bruce Perens. Yes the real Bruce Perens with UID 3872. It is well established that everything Bruce writes is +5 informative and/or inciteful. Please correct the moderation rating accordingly. TIA.

    1. Re:MODERATORS by Anonymous Coward · · Score: 0

      Who the f*ck cares if this was "the" Bruce Perens or not? The guy was offtopic, cheezy, vague and was even flamebaiting. He deserved what he got.

  42. You are correct by Anonymous Coward · · Score: 0

    I have actually written a filesystem that ran on a MySQL database, and I can assure you that it was very slow.

  43. HERE IS ANOTHER GOOD MIRROR!!!! by danielblair · · Score: 3, Interesting
    Here is a link to another GOOD MIRROR! Use it!

    You are all welcome!

    http://www.realcoders.org

    Have Fun Everyone!

    -Danny

    joecamel (at) realcoders (dot) org

    --
    -- Daniel R. Blair Senior Software Architect/Unix & FreeBSD Guru/DJ w: http://unixcoders.org t: @freebsd_hacker
  44. Re: Relational databases are the only answer by some+guy+I+know · · Score: 1
    These days, the idea of using any other database model seems ludicrous.
    ISDN still has its place.
    --
    Those who sacrifice security to condemn liberty deserve to repeat history or something. - Benjamin Santayana
  45. just remember... by hak1du · · Score: 4, Interesting

    Access control lists, user-defined metadata, indexing at the file system level, and all that are not new ideas; they go back to the 1960's. Be didn't invent them, and neither did Microsoft (with WinFS).

    All that complexity comes at a price. UNIX was a reaction against putting so many features into the kernel, and, in my opinion, the UNIX arguments against putting those kinds of features into the kernel are as valid today as they were 30 years ago.

    Unfortunately, the book gives very little historical perspective. It seems to simply assume that "more features" translates into "more advanced". From a quick perusal, division of functionality between kernel and user space seems to be not covered. File versioning at the file system level, another important feature, does not seem to be covered. Historically important file systems and functionality, like those found on VMS, IBM mainframes, and database-based file systems are hardly covered at all.

    This book may give you a good idea of what kind of thinking went into the design of the BeOS file system, but it doesn't even come close to a book on file system design in general. And even as a book on the BeOS file system design, it tells you as much about what the creator of the file system didn't think about as what he did think about.

    1. Re:just remember... by shic · · Score: 1

      Can you suggest a better reference?

      I'm looking for one - not trying to be argumentative... I couldn't care less about Be's FS, but I am interested in concepts surrounding file system design in general.

    2. Re:just remember... by Anonymous Coward · · Score: 0

      Since I am teaching a FS class and your post is very
      informed I'd appreciate if you could contact me -

      I've searched the net and the libraries up and down,
      but there's not much around in FS history and or
      general FS theory I can find.

      Would you mind contacting me?

      tpo at "sourcepole" and then "ch".

      Thanks,
      *t

    3. Re:just remember... by hak1du · · Score: 1

      To both responses: sorry, I don't know of a single reference specific to file systems, nor even of a coherent review or survey comparing the different approaches or philosophies.

      Without any claim of completeness, some papers that you may not have thought about reading are the following... You can get quite a bit of information about the UNIX design philosophy out of the early UNIX research papers (the V7 papers collection) and the more recent Plan 9 papers. Krishnamurthy's book "Practical Reusable UNIX Software" talks a lot about what can be done providing advanced file system functionality in user mode. There are the papers on n-DFS, and lots of papers on file versioning, copy-on-write, and version control in file systems. Then there are Korn's papers on UWIN, which also talk about file system issues.

      People should probably also look at systems like Pick and AS/400, which have very different views of data storage within the operating system from "traditional" file systems and give you some idea of where more database-inspired system designs lead.

    4. Re:just remember... by shic · · Score: 1

      Thanks for your reply... I'm familiar with some of the material you mention... though (at least in the stuff I've read) there seems to be little discussion of what I consider the central 'hard' problem. I can find a lot of information about file system features, however I'm far more interested in a detailed analysis of state-of-the-art placement policy. I am extremely interested in strategies to avoid fragmentation and to enhance locality of reference between associated files (or other logical units of data) in order to devise systems where performance does not degrade over time. Is there published research into the effects of various placement strategies on long-term locality of references / fragmentation? Do some placement strategies always result in inherently stable systems where the emergent layout has desirable locality of reference, or are all reasonably simple strategies inherently unstable as volumes fill?

  46. Syllable too by Anonymous Coward · · Score: 0

    The AtheOS File System by Kurt Skauen is a 64-bit clone of the BeFS, he even used the Practical File System Design book as a guide on how to make the file system. Syllable is using this file system.

  47. floating-point use? by Anonymous Coward · · Score: 0

    Does it explain why early PPC builds of BeOS used floating-point doubles in the file system APIs for things like file size?

    What's a fractional byte of a file? What's NaN or Inf mean?

    The Metrowerks compiler of the day didn't have 64-bit long long working just yet, but that glue isn't all that hard to write.

    Did Dominic simply not understand floating-point?

    1. Re:floating-point use? by Anonymous Coward · · Score: 0

      but that glue isn't all that hard to write.

      Maybe he realized that "simple" glue can get you in alot of trouble down the road and decided to just use something that was supported and yes, ignore the fractions.

  48. Re:Future is relational databases WHY??? by peragrin · · Score: 1

    Why would you want an relational database for your file system. maybe it's just me but I remember everypoint of every piece of data. My mind may use a relational system to keep everything in place, but my phyiscal world is seperated into specific locations for each object.

    When I go for a file, a DVD or one of the swords I know where it will be and what it will contain. Everything has a place and everything in it's place.

    If you pull all that apart, and interconnect it then it doesn't have it's place anymore. You can't be certian you actually removed the item as well. Because of the interconnections it will be possible that you may not be able to truely delete anything.

    Think about Longhorn's new virus, it travels along the relations between items overwriting them but keeping the meta data the same. You will never be able to find the virus, you would literally have to follow it around your hard drive erasing files as you go. No this virus doesn't exsit, but is now a possiblity.

    --
    i thought once I was found, but it was only a dream.
  49. Re:Future is relational databases WHY??? by leandrod · · Score: 1
    > maybe it's just me but I remember everypoint of every piece of data

    It is just you. Or you're not quite truthful.

    > If you pull all that apart, and interconnect it then it doesn't have it's place anymore

    There is no interconnection. In fact, the relational model does away with pointers (and relationships) and substitutes relations instead.

    The point of the relational model is not doing away with hierarchies as ways of presenting and organising data at the user or physical levels, but as ways of managing it at the logical one.

    --
    Leandro Guimarães Faria Corcete DUTRA
    DA, DBA, SysAdmin, Data Modeller
    GNU Project, Debian GNU/Lin
  50. Tests... by EvanED · · Score: 1

    Test

    If that's a link to /., spiffy! :-p

  51. Re:Sorry by Anonymous Coward · · Score: 0

    Well RedHat agrees with you. CONFIG_REISERFS_CHECK is still turned on for Fedora Core 2.

  52. Re:Future is relational databases WHY??? by CableModemSniper · · Score: 1

    What do you mean by relations? You keep saying relational DBs are inherently fast and good and damn if they aren't perfect but they aren't SQL and they don't have relationships they have relations. Now I understand if you feel like you shouldn't have to explain what a relation is, but at least provide a link somewhere. Do you mean relation in teh mathematical sense? (Ie a directed graph or a function or an equilvalence relation, etc.)

    Of course, you're probably trolling, with the 5 name name and the need to advertise your profession on slashdot. In which case I fell for it, but I couldn't just read these things you wrote as statement of fact without bothering to provide any explanation or evidence.

    --
    Why not fork?
  53. Re:Future is relational databases WHY??? by leandrod · · Score: 1
    > What do you mean by relations?

    Like in Mathematics. Really this is not the place to teach that, I recommend Chris J Date's books as listed at DBDebunk.

    > you're probably trolling

    Am not.

    > with the 5 name name

    Old Portuguese tradition, plus my mother didn't want me to have troubles with sharing the same name. Really.

    > the need to advertise your profession on slashdot

    Why is that a problem?

    > I couldn't just read these things you wrote as statement of fact without bothering to provide any explanation or evidence.

    I won't do anyone's homework.

    --
    Leandro Guimarães Faria Corcete DUTRA
    DA, DBA, SysAdmin, Data Modeller
    GNU Project, Debian GNU/Lin
  54. Funny Be IRC quotes by Gunfighter · · Score: 2, Funny

    A few years back, one of the members of my Quake clan was a programmer who preferred BeOS as his platform of choice for development and other everyday tasks. He eventually went to work for Be and we didn't hear from him much after that. Nevertheless, we always gave him hell about his BeOS preference. Here are a few choice quotes from our IRC logs:

    This first one is particularly applicable as it pertains to the "uncorruptable" BeOS filesystem.

    <dEad{Ni}> but you have more problems with win95 than i have ever imagined anyone having
    <Tolen{Ni}> nah...you should see some of the people on my dorm floor...
    <Tolen{Ni}> one guy had to fdisk like 5 times last semester
    <Magaera{Ni}> hehe
    <Magaera{Ni}> You CAN'T corrupt the BeOS file system
    <Magaera{Ni}> Even by kicking out the power cord
    <Gunfighter{Ni}> you can't play Q2 on it either :P

    <Magaera{Ni}> potty stop - brb
    <Gunfighter{Ni}> overkill.. yellow card
    <Magaera{Ni}> what, you'd rather say i was going to "the little programmer's room" or something??
    <Deathwish{Ni}> I got take a BeOS

    <Magaera{Ni}> "BeOS combines the best features of all the major operating systems: the ease-of-use of the Macintosh, the power and flexibility of Linux, and Minesweeper from Windows."

    --
    -- Stu

    /. ID under 2,000. I feel old now.
  55. Mirror with PDF-GMail it. by Anonymous Coward · · Score: 0

    Or you could store it in your GMail account, and E-Mail it to everyone requesting a copy.

  56. Re:Future is relational databases WHY??? by Anonymous Coward · · Score: 0

    I won't do anyone's homework.

    But you are perfectly willing to have others do your homework, eh? hypocritical prick

  57. Reviewer Credentials? by galt2112 · · Score: 1

    I stopped reading as soon as I saw this:
    "As a filesystem and kernel ignoramus...."

    Why is a book about filesystem design & implementation being reviewed by someone who doesn't have an understanding of the topic?

  58. But We're Interested by SoopahMan · · Score: 1

    Maybe - but if you're reading this thread, especially this far down, the "interested in File Systems" requirement has been met ;o)

  59. I was reading about Be the otehr night.. by rofthorax · · Score: 1

    The reason it died.. Microsoft!! Who bought it? Palm ? I woudl suggest to most everyone with a palm pilot, in protest throw it under a steamroller because Palm sells devices with less processing power than a 386.. Palm figured it would be useful in cleverly designing an intuitive interface..

    Anyhow.. I only learned of this last night.. I'm not a Be user.. However it looks like a very interesting and simple to use operating system, and its still available for download..

    A newer incarnation or spinoff of Be, not unlike Handspring being a spinoff of palm, is something called Yellowtag's Zeta..

    Do a search for it.. evidently they are using the source code of BeOS 5 as a basis for a new environment, not developed by Be.. And there is projects for OpenBeOS.. But those who think of Os's like the Amiga dead, are just naysayers who are confident with what they have and figure others will be too.. I'm a Windows XP user, and I am quite confident with what I have, and everything is really easy, I can't imagine how have BeOS would make it easier.. But if I wanted something like MacOS without purchasing a mac, and have the creative artist centric environment that the Mac's do, what do I do? BeOS is the closest thing to the mac without getting Mac hardware.. You have to admit..

    Geeks have no problem using a GUI, but everyone else does.. So naysayers reconsider.. But I, like you naysayers, understand because I use Linux, Windows XP, I've used AmigaOS and MacOS, but have not used BeOS.. I prefer each OS for different things..
    Like I can't imagine having an IDE as tight as linux in any other OS. And XP is really nice for games and client applications.. MacOS is terrific for creative oriented work because of the detail given to the particular signficance of usability and the need for a precise result when doing graphcis work. The AmigaOS would be terrific for embedded systems that haven't the bulk of a MMU and other things that are required for a fault tollerant environment with imperfect applications, AmigaOS enforces object oriented design with assembly language, go figure. Each OS has a use and a strength.. And you can't expect one os to have everything.. Its like designing a house that suits everyone's needs.. The only way that could be done is if the house were to be rebuilt million times a second, once for each perceiver.. VMware comes close to offering the multi-faceted capability of a all being OS. A multi-OS.. But the drawback there is many OS's that do what others do on other machines, but much slower.. You will never get around the advantages of one OS over another..

    --
    Just say no to license servers!!
  60. Listen PPC Fanboy by Anonymous Coward · · Score: 0

    I'll buy a PPC machine when I can get it in ATX motherboard