Professional Web Applications Themes

UFS vs HFS+ - Mac Applications & Software

OK, I recently installed osx on a G4 cube, and since I wanted a maximally clean install, I started by building a new file system on the disk. I came to the question about UFS or HFS+, and, after a little reading, chose the obviously superior UFS. But now I find out that various Mac programs, in spite of bearing labels like "built for osx", do not grok UFS. Also, it appears that new machines currently are built with HFS+ on the hard drive instead of UFS. Clearly, it must be difficult for programs to change over to UFS support, ...

  1. #1

    Default UFS vs HFS+

    OK, I recently installed osx on a G4 cube, and since I wanted a
    maximally clean install, I started by building a new file system
    on the disk. I came to the question about UFS or HFS+, and, after
    a little reading, chose the obviously superior UFS. But now I find
    out that various Mac programs, in spite of bearing labels like
    "built for osx", do not grok UFS. Also, it appears that new machines
    currently are built with HFS+ on the hard drive instead of UFS.

    Clearly, it must be difficult for programs to change over to UFS
    support, but I am having a hard time imagining why. Shouldn't
    part of "built for osx" mean complete support of the UFS?

    Greg Shenaut
    gkshenaut@ucdavis.edu Guest

  2. #2

    Default Re: UFS vs HFS+

    [email]gkshenautucdavis.edu[/email] wrote:
    > OK, I recently installed osx on a G4 cube, and since I wanted a
    > maximally clean install, I started by building a new file system
    > on the disk. I came to the question about UFS or HFS+, and, after
    > a little reading, chose the obviously superior UFS. But now I find
    > out that various Mac programs, in spite of bearing labels like
    > "built for osx", do not grok UFS. Also, it appears that new machines
    > currently are built with HFS+ on the hard drive instead of UFS.
    UFS is inferior in every way to HFS+ *unless* you require case
    sensitivity in a file system. Don't use UFS. In fact, wipe your drive
    and start over. It is worth doing.
    > Clearly, it must be difficult for programs to change over to UFS
    > support, but I am having a hard time imagining why. Shouldn't
    > part of "built for osx" mean complete support of the UFS?
    Meh. There's no way I'm ever installing UFS on my system to test
    against. It's too slow, too inefficient, and as you pointed out nothing
    works with it.

    Steven Fisher Guest

  3. #3

    Default Re: UFS vs HFS+

    [email]gkshenautucdavis.edu[/email]:
    > OK, I recently installed osx on a G4 cube, and since I wanted a
    > maximally clean install, I started by building a new file system
    > on the disk. I came to the question about UFS or HFS+, and, after
    > a little reading, chose the obviously superior UFS. But now I find
    > out that various Mac programs, in spite of bearing labels like
    > "built for osx", do not grok UFS. Also, it appears that new machines
    > currently are built with HFS+ on the hard drive instead of UFS.
    > Clearly, it must be difficult for programs to change over to UFS
    > support, but I am having a hard time imagining why. Shouldn't
    > part of "built for osx" mean complete support of the UFS?
    *****

    You're asking this question a bit late; if you had searched USENET
    archives or asked here before you chose UFS you probably would not have
    done so. For the great majority of Mac users (something like 99.9999
    percent) UFS is not at all superior to HFS+ because it renders the
    system unusable for that 99.9999 percent -- including you, apparently.
    So hang onto your towel (you do know where it is, right?) and grok
    this: Reformat the drive as HFS+, reinstall everything, and enjoy
    yourself.

    Davoud

    --
    usenet *at* davidillig dawt com
    Davoud Guest

  4. #4

    Default Re: UFS vs HFS+

    Steven Fisher <sdfisherspamcop.net> wrote (Wed, 25 Jun 2003 15:43:48 GMT):
    > UFS is inferior in every way to HFS+ *unless* you require case
    > sensitivity in a file system.
    Well, duh. Given a choice, why would anyone live without case
    sensitivity in directory names? Why would anyone put up with
    multiple roots? To me, the greatest advantage of OSX was that it
    incorporated big chunks of FreeBSD, definitely including the Unix
    filesystem.

    One of the cool things about UNIX style systems is the concept of
    the "virtual file system" (VFS), which, unless I'm really missing
    something, could allow an underlying UFS filesystem to use HSF+ or
    any other filename semantics from the point of view of applications.
    I wonder why this hasn't been implemented?

    Actually, the one program that messed up so far I haven't decided
    I need to keep. With a few exceptions, I'll probably be using mostly
    unixy programs from gnudarwin & fink. That said, I am definitely
    pondering the filesystem issue, and may end up doing a reinstall.

    Greg Shenaut
    gkshenaut@ucdavis.edu Guest

  5. #5

    Default Re: UFS vs HFS+

    [email]gkshenautucdavis.edu[/email] wrote:
    > Steven Fisher <sdfisherspamcop.net> wrote (Wed, 25 Jun 2003 15:43:48 GMT):
    >
    >>UFS is inferior in every way to HFS+ *unless* you require case
    >>sensitivity in a file system.
    >
    >
    > Well, duh. Given a choice, why would anyone live without case
    > sensitivity in directory names? Why would anyone put up with
    > multiple roots? To me, the greatest advantage of OSX was that it
    > incorporated big chunks of FreeBSD, definitely including the Unix
    > filesystem.
    Because to most people, a case preserving file system is superior.
    Certainly, I prefer it. I'll never work with a case sensitive file
    system unless I have to... and the only way I'd have to is to work with
    source trees/applications that have both Foo.c and foo.c and require
    them to be different.

    But this is a lame argument. Everyone who has an opinion on it has a
    strong opinion on it, right? :)

    Steven Fisher Guest

  6. #6

    Default Re: UFS vs HFS+

    In article <bdci2l$b62$2woodrow.ucdavis.edu>, <gkshenautucdavis.edu>
    wrote:
    > Well, duh. Given a choice, why would anyone live without case
    > sensitivity in directory names?
    You've got that backward. Given a choice, why would anyone use a
    filesystem on which Foo and foo were different files?

    Computers are supposed to make life easier, not harder. Having to
    remember whether the file you want is named ReadMe or readme makes life
    harder. That is something the computer should remember.

    The case-sensitivity of UNIX filesystems is the result of an OS that
    was designed by programmers for programmers. Programmers are used to
    the idea that "Q" and "q" are different letters and are willing to put
    up with having to make case distinctions themselves in exchange for
    marginally more efficiency in the filesystem code. Over time, they
    convinced themselves that case-sensitivity itself was the feature,
    rather than the marginal improvement in efficiency being the feature.

    UNIX is like that in a lot of ways, really. Do things in the simplest
    way that works and then convince yourself that it's a feature.

    --
    Jerry Kindall, Seattle, WA [url]http://www.jerrykindall.com/[/url]

    If replying to this message by e-mail, send plain text only.
    Replies with files or HTML will be deleted by my spam filters.
    Jerry Kindall Guest

  7. #7

    Default Re: UFS vs HFS+

    In article <bdci2l$b62$2woodrow.ucdavis.edu>, [email]gkshenautucdavis.edu[/email]
    wrote:
    > Steven Fisher <sdfisherspamcop.net> wrote (Wed, 25 Jun 2003 15:43:48 GMT):
    > > UFS is inferior in every way to HFS+ *unless* you require case
    > > sensitivity in a file system.
    >
    > Well, duh. Given a choice, why would anyone live without case
    > sensitivity in directory names?
    Because English is not case-sensitive.

    Our language has rules for deciding whether letters should be written in
    upper- or lowercase, but it's still the same word regardless of whether
    it's written as "button", "Button", "BUTTON", or "BuTtOn". Most people
    want to see an alphabetical sort that places "button" between "Anchor"
    and "Cotton". If they're searching for a file, they probably remember
    what words were in the name, but might not remember what case was used.
    Wayne C. Morris Guest

  8. #8

    Default Re: UFS vs HFS+

    <gkshenautucdavis.edu> wrote:
    > Well, duh. Given a choice, why would anyone live without case
    > sensitivity in directory names?
    Case-insensitive filenames are idiotic, yes, but there is one good reason
    to deal with them: so that you can use HFS+ rather than UFS. The case
    sensitivity is pretty much the only advantage you'll get with UFS, and
    the enormous loss in performance and functionality isn't worth it.

    --
    Jeremy | [email]jeremyexit109.com[/email]
    Jeremy Guest

  9. #9

    Default Re: UFS vs HFS+

    Jerry Kindall <jerrykindallnospam.invalid> wrote (Wed, 25 Jun 2003 09:50:16 -0700):
    > Computers are supposed to make life easier, not harder. Having to
    > remember whether the file you want is named ReadMe or readme makes life
    > harder. That is something the computer should remember.
    I see it exactly the opposite way. For example, take Makefile
    versus makefile, README vs readme, INSTALL vs install, ToDo vs
    todo. If you have a sea of files, it's very helpful to be able to
    spot such files both because they are sorted first and because they
    stand out from the others. Another common use is to put the first
    letter of names of directories in upper case, for example in your
    home directory. Finally, the distinction is sometimes used to
    distinguish between a newer and older program, as in mail vs Mail,
    cc vs CC, and so on. If you have a bunch of files and some require
    distinctiveness, being able to use the upper- vs lowercase distinction
    is very useful.

    If you think about the history of punctuation, the same arguments
    could be made about not capitalizing the first letter in sentences
    or proper names, or in acronyms--clearly these distinctions are useful
    or they wouldn't exist.
    > The case-sensitivity of UNIX filesystems is the result of an OS that
    > was designed by programmers for programmers. Programmers are used to
    > the idea that "Q" and "q" are different letters and are willing to put
    > up with having to make case distinctions themselves in exchange for
    > marginally more efficiency in the filesystem code. Over time, they
    > convinced themselves that case-sensitivity itself was the feature,
    > rather than the marginal improvement in efficiency being the feature.
    Not quite--I think that most of us who were around back then found
    case sensitivity to be a great step forward. Remember, these were
    the days when system designers squeezed every bit until it screamed
    to increase space--many systems in those days used a reduced
    character set for filenames because they could squeeze 3 radix 40
    characters into 2 bytes, not because they didn't understand that
    A and a were different symbols. The aboriginal UNIX 14-byte ASCII
    filenames were a tremendously freeing innovation, not to mention
    the single-root filesystem itself.
    > UNIX is like that in a lot of ways, really. Do things in the simplest
    > way that works and then convince yourself that it's a feature.
    Well, like it or not, Mac OSX is UNIX. It will be interesting to
    see how long it takes for pre-OSX Mac users & developers to adapt.

    Greg Shenaut
    gkshenaut@ucdavis.edu Guest

  10. #10

    Default Re: UFS vs HFS+

    Davoud <starsky.net> wrote (Wed, 25 Jun 2003 13:14:15 -0400):
    > For most Mac users the greatest advantage of OSX is stability when
    > running their Mac (not UNIX) applications. If all you want is a UNIX
    > platform you'll end up wasting time and money using a Mac.
    I don't think so. Our whole lab is moving to Macs on the desktops
    precisely because we can continue to keep our current file formats,
    utilities, and so on, but can use the much more world-compatible Mac
    environment when necessary. We are moving from an all-FreeBSD lab
    to a Mac desktop, FreeBSD server lab, which may evolve into an
    all-Mac lab in time. Undoubtedly there will be other cases of
    culture shock like this one regarding the HFS+ filesystem, but
    I think we are on the right path. At the very least, it should
    allow us to bypass Windoze.

    Greg Shenaut
    gkshenaut@ucdavis.edu Guest

  11. #11

    Default Re: UFS vs HFS+

    Wayne C. Morris <wayne.morristhis.is.invalid> wrote (Wed, 25 Jun 2003 18:35:59 GMT):
    > Because English is not case-sensitive.
    > Our language has rules for deciding whether letters should be written in
    > upper- or lowercase, but it's still the same word regardless of whether
    > it's written as "button", "Button", "BUTTON", or "BuTtOn". Most people
    > want to see an alphabetical sort that places "button" between "Anchor"
    > and "Cotton". If they're searching for a file, they probably remember
    > what words were in the name, but might not remember what case was used.
    The HFS+ filesystem is not case sensitive because *English orthography*
    is not case sensitive? That seems rather odd given that surely people
    who speak other languages use Macintoshes. Given a string like kindergarten,
    say, should it be capitalized? Yes if it is German, no if English.

    It's simply necessary to be able to make those distinctions in file
    names, or so it seems to me.

    Greg Shenaut
    gkshenaut@ucdavis.edu Guest

  12. #12

    Default Re: UFS vs HFS+

    Jerry Kindall <jerrykindallnospam.invalid> writes:
    [...]
    > You've got that backward. Given a choice, why would anyone use a
    > filesystem on which Foo and foo were different files?
    [...]

    "Polish" and "polish" mean two different things.

    The first means 'of, or relating to, the country of Poland'; the
    second means 'to make shiny'.

    There is such a thing as proper spelling.

    --
    David Magda <dmagda at ee.ryerson.ca>, [url]http://www.magda.ca/[/url]
    Because the innovator has for enemies all those who have done well under
    the old conditions, and lukewarm defenders in those who may do well
    under the new. -- Niccolo Machiavelli, _The Prince_, Chapter VI
    David Magda Guest

  13. Moderated Post

    Default Re: UFS vs HFS+

    Removed by Administrator
    David Magda Guest
    Moderated Post

  14. #14

    Default Re: UFS vs HFS+

    In article <bdci2l$b62$2woodrow.ucdavis.edu>, <gkshenautucdavis.edu>
    wrote:
    > One of the cool things about UNIX style systems is the concept of
    > the "virtual file system" (VFS), which, unless I'm really missing
    > something, could allow an underlying UFS filesystem to use HSF+ or
    > any other filename semantics from the point of view of applications.
    > I wonder why this hasn't been implemented?
    That's not what VFS does. It's one way of implementing plug-in
    filesystems, in a somewhat object oriented way.

    As a gross generalization, it's similar to C++ vtables as a way of
    letting you call a particular method of an object, without having to
    know what kind of object it is.

    -Mark
    Mark Day Guest

  15. #15

    Default Re: UFS vs HFS+

    In article <86vfut3nz1.fsfnumber6.magda.ca>, David Magda
    <dmagda+trace030624ee.ryerson.ca> wrote:
    > Jerry Kindall <jerrykindallnospam.invalid> writes:
    > [...]
    > > You've got that backward. Given a choice, why would anyone use a
    > > filesystem on which Foo and foo were different files?
    > [...]
    >
    > "Polish" and "polish" mean two different things.
    >
    > The first means 'of, or relating to, the country of Poland'; the
    > second means 'to make shiny'.
    >
    > There is such a thing as proper spelling.
    "Polish" can also be spelled "Polish" at the beginning of a sentence.
    Or in the middle of the sentence as I did in the preceding sentence.
    In neither case do I intend the word to refer to anything Polish.

    If you received a letter addressed to "david magda", would you:

    a) mark it "Return to Sender - Recipient Unknown" and send it back?
    b) open it?

    Should we expect our computers to be more or less tolerant of ambiguity
    than we are?

    --
    Jerry Kindall, Seattle, WA [url]http://www.jerrykindall.com/[/url]

    If replying to this message by e-mail, send plain text only.
    Replies with files or HTML will be deleted by my spam filters.
    Jerry Kindall Guest

  16. #16

    Default Re: UFS vs HFS+

    In article <bdctal$gq2$2woodrow.ucdavis.edu>, <gkshenautucdavis.edu>
    wrote:
    > Jerry Kindall <jerrykindallnospam.invalid> wrote (Wed, 25 Jun 2003 09:50:16
    > -0700):
    > > Computers are supposed to make life easier, not harder. Having to
    > > remember whether the file you want is named ReadMe or readme makes life
    > > harder. That is something the computer should remember.
    >
    > I see it exactly the opposite way. For example, take Makefile
    > versus makefile, README vs readme, INSTALL vs install, ToDo vs
    > todo. If you have a sea of files, it's very helpful to be able to
    > spot such files both because they are sorted first and because they
    > stand out from the others. Another common use is to put the first
    > letter of names of directories in upper case, for example in your
    > home directory.
    Those are arguments for case preserving, but not necessarily case
    sensitive, names. Both HFS and HFS Plus are case preserving but case
    insensitive.

    If you create "README", then that's what it will look like. If you
    create "readme", that's what it will look like. You just can't have
    both in the same directory at the same time.

    -Mark
    Mark Day Guest

  17. #17

    Default Re: UFS vs HFS+

    On Wed, 25 Jun 2003 about 16:24 -0400 UTC David Magda wrote:
    > Jerry Kindall <jerrykindallnospam.invalid> writes:
    > [...]
    > > You've got that backward. Given a choice, why would anyone use a
    > > filesystem on which Foo and foo were different files?
    > [...]
    >
    > "Polish" and "polish" mean two different things.
    >
    > The first means 'of, or relating to, the country of Poland'; the
    > second means 'to make shiny'.
    Yes, as in

    "Polish that Apple!"
    > There is such a thing as proper spelling.
    And proper mood.

    robert delius royar Guest

  18. #18

    Default Re: UFS vs HFS+

    In article <86vfut3nz1.fsfnumber6.magda.ca>,
    David Magda <dmagda+trace030624ee.ryerson.ca> wrote:
    > There is such a thing as proper spelling.
    Which is, of course, why case preservation is important, but it is not
    at all an argument for case sensitivity. Among other things, polish is
    always spelled "Polish" at the start of the sentence.

    If you're storing countries and shoe supplies (for instance) in the same
    directory, you need to rethink how you've structured things.
    Steven Fisher Guest

  19. #19

    Default Re: UFS vs HFS+

    In article <bdcudd$gq2$4woodrow.ucdavis.edu>, [email]gkshenautucdavis.edu[/email]
    wrote:
    > The HFS+ filesystem is not case sensitive because *English orthography*
    > is not case sensitive? That seems rather odd given that surely people
    > who speak other languages use Macintoshes. Given a string like kindergarten,
    > say, should it be capitalized? Yes if it is German, no if English.
    >
    > It's simply necessary to be able to make those distinctions in file
    > names, or so it seems to me.
    This is a great argument for a CASE PRESERVING file system, but not a
    case sensitive one. Luckily, HFS+ is case preserving.
    Steven Fisher Guest

  20. #20

    Default Re: UFS vs HFS+

    In article <bdd484$kvs$1woodrow.ucdavis.edu>, [email]gkshenautucdavis.edu[/email]
    wrote:
    > Jeremy <jeremyexit109.com> wrote (Wed, 25 Jun 2003 18:51:53 -0000):
    > > Case-insensitive filenames are idiotic, yes, but there is one good reason
    > > to deal with them: so that you can use HFS+ rather than UFS. The case
    > > sensitivity is pretty much the only advantage you'll get with UFS, and
    > > the enormous loss in performance and functionality isn't worth it.
    >
    > If by "performance and functionality" you mean that some programs
    > won't work on UFS filesystems then I see what you mean; otherwise,
    > what do you mean? (This is a serious question not an attack; I'm
    > old with UFS but new with HFS+.)
    >
    > Greg Shenaut
    (editied repost of something I wrote earlier).

    * File name:
    255 Unicode characters in HFS+. 255 ASCII in UFS. (cf 20 ASCII in
    HFS, 11 ASCII in FAT16)

    Case preserving in HFS+ (like NTFS, FAT and others). cf case
    sensitive in UFS, ext2/3, ReiserFS and most Unix/Linux FS.

    * Metadata:
    E.g. User/group/system permissions. Creation/modify/access/backup
    date. Owner (user/group), file type/creator. inode. File ID, etc etc.
    HFS+ metatdata is a superset of UFS.

    * Multiple forks:
    I.e. resource and data. NTFS and Novell (NWFS, NSS) also offer this
    (HFS+ & NTFS allows an arbitrary number of forks, although only 2 are
    currently implemented in HFS+). NOT to be confused with
    Metadata. Mac OS X simulates this on UFS with some artful renaming. Mac
    OS X packages are aiming to obsolete this anyway.

    * File size:
    HFS+ is 64 bit. UFS is 32 bit, thus file size is TB in HFS+, GB in
    UFS. Depends on whether fields in the file structure are 64 bits or 32
    bits. Very difficult to fix without radically changing the file system.
    (You can play some games with allocation sizes to allow file systems
    larger than 2GB, but still have an individual file size limit of 2GB).

    * Journalling:
    Mac OS X supports journalling on HFS+ but not UFS. (ext3 and ReiserFS
    on Linux also do journalling). (Note, this is a feature of tha current
    Mac OS implementation, not an inherent feature of HFS+ or UFS).

    * File structure:
    HFS+ is a B-Tree based system---as was HFS. This is considered a Good
    Thing and is one of the claims to fame of ReiserFS on Linux. The new
    Novell NSS is also B-Tree, the old NWFS not. UFS is not B-Tree.

    * Implmentation:
    On Mac OS X, HFS+ seems to be better implemented than UFS, i.e.
    faster (though Mark Day may be able to correct me on this).

    In summary---use HFS+ on Mac OS X. The only thing UFS, in the current
    Mac OS X implementation, offers that HFS+ does not is CaSe SeNsItivitY.
    --
    Michael Newbery
    "Never trust anything that thinks for itself unless you can see
    where it keeps its brains"
    Michael Newbery Guest

Page 1 of 6 123 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139