JFS and UNICODE.SYS - OS2

This is a discussion on JFS and UNICODE.SYS - OS2 ; Hi all, IIRC it is necessary to load UNICODE.SYS before IFS=JFS.IFS. My question: what other SETs and files must be present to allow JFS to work properly with CP850, CP437 and CP1004? TIA and best regards Uwe...

+ Reply to Thread
Results 1 to 7 of 7

Thread: JFS and UNICODE.SYS

  1. JFS and UNICODE.SYS

    Hi all,

    IIRC it is necessary to load UNICODE.SYS before IFS=JFS.IFS.

    My question: what other SETs and files must be present to allow JFS to
    work properly with CP850, CP437 and CP1004?

    TIA and best regards
    Uwe

  2. Re: JFS and UNICODE.SYS

    [A complimentary Cc of this posting was sent to
    Uwe Pilgram
    ], who wrote in article :
    > Hi all,
    >
    > IIRC it is necessary to load UNICODE.SYS before IFS=JFS.IFS.
    >
    > My question: what other SETs and files must be present to allow JFS to
    > work properly with CP850, CP437 and CP1004?


    ??? You can have only two codepages simultaneously, right?

    With 2 codepages, AND unicode data directory present, things work fine
    here. I do not know whether data directories are REQUIRED for
    unicode.sys...

    Yours,
    Ilya

  3. Re: JFS and UNICODE.SYS

    Hi Ilya,

    the background of my question is, how UNICODE.SYS and UCONV.DLL locate
    codepage tables. My OS/2 system has a SET ULSPATH=C:\LANGUAGE\CODEPAGE
    statement. This directory contains many codepage definitions. This is
    definitively used by UCONV.DLL to fetch codepage tables. I wonder which
    subset of these code page tables is necessary to make JFS (and FAT32) happy.

    Unfortunately I don't understand your statement below, that two
    codepages (you surely mean the CP definitions in Config.Sys) and UNICODE
    data directory work fine. JFS abd FAT32 always store all names
    translated to unicode. The question is how they do that and which
    codepage tables they need.

    I would like to share the experience of the experts out there.

    Best regards
    Uwe





    Ilya Zakharevich wrote:
    > [A complimentary Cc of this posting was sent to
    > Uwe Pilgram
    > ], who wrote in article :
    >> Hi all,
    >>
    >> IIRC it is necessary to load UNICODE.SYS before IFS=JFS.IFS.
    >>
    >> My question: what other SETs and files must be present to allow JFS to
    >> work properly with CP850, CP437 and CP1004?

    >
    > ??? You can have only two codepages simultaneously, right?
    >
    > With 2 codepages, AND unicode data directory present, things work fine
    > here. I do not know whether data directories are REQUIRED for
    > unicode.sys...
    >
    > Yours,
    > Ilya


  4. Re: JFS and UNICODE.SYS

    [A complimentary Cc of this posting was sent to
    Uwe Pilgram
    ], who wrote in article :
    > Unfortunately I don't understand your statement below, that two
    > codepages (you surely mean the CP definitions in Config.Sys) and UNICODE
    > data directory work fine. JFS abd FAT32 always store all names
    > translated to unicode. The question is how they do that and which
    > codepage tables they need.


    I never tried to minimize the codepage directory. However, it is
    "obvious" that codepage files for not enabled codepages are not
    needed.

    Hope this helps,
    Ilya

  5. Re: JFS and UNICODE.SYS

    On Tue, 25 Dec 2007 19:35:32 UTC, Uwe Pilgram wrote:

    > IIRC it is necessary to load UNICODE.SYS before IFS=JFS.IFS.
    >
    > My question: what other SETs and files must be present to allow JFS to
    > work properly with CP850, CP437 and CP1004?


    You also need the \LANGUAGE\CODEPAGE directory on the current drive,
    containing the codepage files you need (e.g. IBM850, IBM437 and
    IBM1004).

    --
    Alex Taylor
    http://www.cs-club.org/~alex

    Please take off hat when replying.

  6. Re: JFS and UNICODE.SYS

    On Fri, 28 Dec 2007 16:13:50 UTC, Uwe Pilgram wrote:

    > the background of my question is, how UNICODE.SYS and UCONV.DLL locate
    > codepage tables. My OS/2 system has a SET ULSPATH=C:\LANGUAGE\CODEPAGE
    > statement. This directory contains many codepage definitions. This is
    > definitively used by UCONV.DLL to fetch codepage tables. I wonder which
    > subset of these code page tables is necessary to make JFS (and FAT32)
    > happy.


    You only need the two codepage files for the two you have defined
    in your CONFIG.SYS CODEPAGE statement. Nothing else has to be in
    that directory for JFS.IFS to work, at least.

    Do note that UNICODE.SYS seems to be hardcoded to look for the files
    in ?:\LANGUAGE\CODEPAGE, where ?: is the currently-active drive at the
    moment UNICODE.SYS is loaded. It doesn't use the ULSPATH variable
    (which in any case should be \LANGUAGE, not \LANGUAGE\CODEPAGE).

    --
    Alex Taylor
    http://www.cs-club.org/~alex

    Please take off hat when replying.

  7. Re: JFS and UNICODE.SYS

    Thanx Alex. Very helpful and informative.
    Regards
    Uwe

    ------------

    Alex Taylor wrote:
    > On Fri, 28 Dec 2007 16:13:50 UTC, Uwe Pilgram wrote:
    >
    >> the background of my question is, how UNICODE.SYS and UCONV.DLL locate
    >> codepage tables. My OS/2 system has a SET ULSPATH=C:\LANGUAGE\CODEPAGE
    >> statement. This directory contains many codepage definitions. This is
    >> definitively used by UCONV.DLL to fetch codepage tables. I wonder which
    >> subset of these code page tables is necessary to make JFS (and FAT32)
    >> happy.

    >
    > You only need the two codepage files for the two you have defined
    > in your CONFIG.SYS CODEPAGE statement. Nothing else has to be in
    > that directory for JFS.IFS to work, at least.
    >
    > Do note that UNICODE.SYS seems to be hardcoded to look for the files
    > in ?:\LANGUAGE\CODEPAGE, where ?: is the currently-active drive at the
    > moment UNICODE.SYS is loaded. It doesn't use the ULSPATH variable
    > (which in any case should be \LANGUAGE, not \LANGUAGE\CODEPAGE).
    >


+ Reply to Thread