TEXT   22

readme.txt

Guest on 7th May 2021 04:51:53 PM

  1.  
  2.    @PRODUCT_LONG  @{r74}@{d}@{_}
  3.  
  4.  
  5.    INTRODUCTION
  6.  
  7.       To meet  the data  archiving  needs, ARJ Software  Russia delivers a
  8.       range of its own  products based  on the code of  original ARJ. From
  9.       the  beginning,  our  goal  was  to  retain  the  functionality  and
  10.       compatibility with  the  original ARJ  for DOS, but we  also provide
  11.       features  and  enhancements  that  are a  must for certain  tasks or
  12.       certain platforms where no ARJ has been before.
  13.  
  14.  
  15.    NEW FEATURES AND DIFFERENCES FROM THE ORIGINAL ARJ
  16.  
  17.       ("-" indicates a  missing feature, "*" - a different  operation and
  18.        "+" stands for a feature unique for our implementation)
  19.  
  20.       -  There may be a  significant  performance drawback  when accessing
  21.          files on volumes with no write-back cache.
  22.  
  23.       *  The "-hm" options  are  compatible  with their ARJ implementation
  24.          hence  a  temporary  swap  file  is  created  if  the  file  list
  25.          exceeds  3000  files.  To keep  the  entire list  in  memory, use
  26.          -hm65000. This  is  useful if you are running  a non-DOS version,
  27.          have  an   adequate  amount  of   RAM  and   therefore   are  not
  28.          constrained with RAM occupied by file lists.
  29.  
  30.       *  The ARJ display  program, ARJ$DISP, has been  renamed to ARJDISP.
  31.          If you are using a custom  display module, you  have to rename it
  32.          as well.
  33.  
  34.       *  "-t1g"  can  really  be  used  as  a   shortcut  for "-t1gf",  as
  35.          documented but not implemented in the original ARJ.
  36.  
  37.       *  Parameters  accepting numeric  quantities (such as "-v" or "-2i")
  38.          tolerate  both  decimal and  hexadecimal values. To denote a  hex
  39.          value, place  "0x" in  front  of it, as  in "-2i0x1000" (equal to
  40.          "-2i4096").
  41.  
  42.          The  multipliers 'K'  and  'M' have  been  supplemented  with 'G'
  43.          (giga-) and three currently reserved quantities: 'T' (tera-), 'P'
  44.          (peta-) and 'E' (exa-). All  modifiers imply  a decimal power, so
  45.          "-v1G" is equivalent to "-v1000000000" or "-v1000M".
  46.  
  47.          These two changes may require a review of the existing ARJ setup,
  48.          as marshalling certain switches together ("-jd0x1") may no longer
  49.          have  the  desired  effect ("-jd0", "-x1"). This does  not affect
  50.          the parameters like "-m4" or "-a1" where the digit is a modifier,
  51.          not a free-form value, and thus will go unnoticed for most of the
  52.          practical configurations.
  53.  
  54.       *  Comments specified with "-jz" or "-z" will retain  their original
  55.          layout in the archive, without replacing their last character for
  56.          newline.
  57.  
  58.       +  "ARJ a -d1" will delete  files without asking for  permission, as
  59.          "ARJ m". "ARJ a -d2" will  truncate  files, rather than  deleting
  60.          them, which is usually suggested for keeping hard links.
  61.  
  62.       +  The "-_" option can be used to convert filenames into lower case.
  63.          When  adding  files, the  filenames  stored  in the  archive  are
  64.          folded  down  into  lowercase. When  extracting files, a  similar
  65.          conversion is carried out for filenames being restored. No checks
  66.          for duplicate filenames are made. On case-sensitive file systems,
  67.          the "-jt1" switch is not operable in conjunction with "-_".
  68.  
  69.       +  The "-h#" option  has been  improved to  allow  custom  date/time
  70.          formats.  A  custom  format  is  specified  by  putting a  format
  71.          sequence   right   after  the "-h#".  The  following   characters
  72.          represent date/time macros:
  73.  
  74.           Y = year,       M = month,      D = day
  75.           h = hour,       m = minute,     s = second
  76.           N = day of year
  77.  
  78.           (note that these are case-sensitive)
  79.  
  80.          All  other  characters,  as  well as  those going  beyond  format
  81.          limits (4  digits for year, 2 digits for all  other fields),  are
  82.          treated as delimiters. Examples:
  83.  
  84.          ARJ a project- -h#YYYYMMDD              (project-19991022.arj)
  85.          ARJ a backup- -h#MM-DD_hh-mm-ss    (backup-10-22_23-57-16.arj)
  86.          ARJ a specs -h#YY                                (specs99.arj)
  87.          ARJ a logs_ -h#NNN                              (logs_295.arj)
  88.          ARJ a test -h#YYYYYYYY                      (testYYYY1999.arj)
  89.  
  90.       +  The "-ha" has been improved. Now, when  used  in conjuction  with
  91.          an archiving  command, it does  not mark read-only  files as such
  92.          in archive.  This simplifies  archiving from  CD-ROM media  where
  93.          virtualized read-only  attribute  is  forced  by  respective IFS
  94.          driver.
  95.  
  96.       +  The  "-2a"  option  is  implemented  in  ARJ  due  to  a  popular
  97.          demand from  FidoNet system  operators.  Basically it  acts  like
  98.          "-jo",  with   the   exception  that  file  names,  and  not  the
  99.          extensions,  are  "serialized". Consider  having  a  file  called
  100.          "FILE995N.TXT" and  an archive that contains  the  same  file. If
  101.          extracted  with  the "-2a"  option,  the  file  will  be  written
  102.          to  "FILE9950.TXT", if  you extract it  again,  it'll  be  called
  103.          "FILE9951.TXT",  and   so   on   up   to   "FILE9959.TXT",   then
  104.          "FILE9960.TXT". And  after  "FILE9999.TXT" ARJ  will  start  with
  105.          "FIL00000.TXT". This option allows you  to extract  one  file  to
  106.          100000000  unique  names.  It's  essential  to  system  operators
  107.          since multiple mail  packets with  the same  name may  come  from
  108.          different systems.
  109.  
  110.          NOTES:
  111.            1. It'll be wise to  include  this  option in the  script  that
  112.               unpacks the ARCmail  packets and  NOT  in ARJ_SW environment
  113.               variable. This  option is  a  security  measure  for systems
  114.               running in  unattended  mode, and  will  only confuse you if
  115.               enabled by default.
  116.            2. There's a security  hole: a file  called  "9999.XXX" or  so,
  117.               will not  be  overwritten.  However, all  subsequent  writes
  118.               will be  redirected  into  file  "0000.XXX". So, files  with
  119.               9s in the beginning have  less  chances of being  preserved.
  120.               Hopefully such situation is unlikely for FidoNet systems.
  121.            3. There is  another  option, "-jo1", to  serialize  filenames,
  122.               however  its   operation   is  different.  The  volume  must
  123.               support  long  filenames  in   order  to  use  this  option,
  124.               moreover,  it's  not   suitable  for  dealing  with  FidoNet
  125.               ARCmail.
  126.  
  127.       +  "-2d" enforces  the header  compatibility mode. In this mode, the
  128.          archive   header   format   corresponds  to   the   original  ARJ
  129.          specification, besides this, "MS-DOS" is  stamped as the host OS,
  130.          to prevent the "Binary file from a different OS" warning messages
  131.          when unpacking  the archive  in DOS. "-2d1" retains  the enhanced
  132.          header format, but makes the archive comment display correctly in
  133.          DOS.
  134.  
  135.       +  "-2f" can be  used  to  apply  the archive  comment to the  first
  136.          volume only, and to strip it out for subsequent volumes.
  137.  
  138.       +  "-2i" is akin  to "-jx" but acts on  the .ARJ being processed. It
  139.          skips unconditionally the given  number of bytes at the beginning
  140.          of the archive. Its primary uses are  to recover severely damaged
  141.          archives  or  extract  ARJ   files  contained   within  some  raw
  142.          file system. Only the first  archive being  processed is affected
  143.          by this switch; subsequent archives (e.g. multivolume) assume -2i
  144.          of zero.
  145.  
  146.       +  "-2k" option forces 2-digit display of year in lists. This can be
  147.          helpful if the 3-digit  year format used for dates beyond 2100 is
  148.          confusing.
  149.  
  150.          Alternatively, "-2k1" uses a non-ambiguous  format  that is  both
  151.          easy to read and  information-packed.  The dates are  represented
  152.          by two digits if  the year  is 1970 to 2069, and in three  digits
  153.          if it's 2070 or beyond.
  154.  
  155.          Examples:
  156.                          15.07.1990  15.07.2040  15.07.2090  15.07.2103
  157.          Default:          90-07-15    40-07-15    90-07-15   103-07-15
  158.          -2k:              90-07-15    40-07-15    90-07-15    03-07-15
  159.          -2k1:             90-07-15    40-07-15   090-07-15   103-07-15
  160.  
  161.       +  "-2r" tells ARJ to store  directory  attributes first, then store
  162.          its contents. This is the order  that was used  by default in ARJ
  163.          prior to 2.76. It is useful  when the archive is  to be extracted
  164.          in an older version of ARJ to avoid directory  overwrite prompts.
  165.          Upon extraction, it forces ARJ to ask if directory attributes are
  166.          to  be  overwritten (by default, ARJ  will  always overwrite  the
  167.          directory attributes without asking for confirmation).
  168.  
  169.       +  With no ARJ_SW  specified, ARJ  looks for  a file  named  ARJ.CFG
  170.          in its home  directory. If found, this file  will  be parsed  and
  171.          used  as  a  standard  ARJ  configuration  file (see  manual  for
  172.          details). For UNIX platforms, this  has been changed to search in
  173.          certain standard locations instead of home directory, see the ARJ
  174.          for UNIX notes for further reference.
  175.  
  176.       +  REARJ v 2.42.05 and higher accepts the "T" modifier in REARJ.CFG,
  177.          which means that it should take care to delete the output archive
  178.          itself if rearchiving fails.
  179.  
  180.  
  181.    EXTENDED ATTRIBUTES HANDLING
  182.  
  183.       Beginning  with  version  2.62.10,  the  extended  attributes  (also
  184.       referenced to  as  EAs) can  be  backed  up  and   restored  without
  185.       needing  any external  utilities. This is  achieved  by  compressing
  186.       and  storing EAs as a part  of file  header. ARJ  supports SAA-style
  187.       EAs under OS/2 and Windows NT.
  188.  
  189.       Restrictions on EA support:
  190.  
  191.       *  The multivolume restart  feature (-jn) will not  work  if EAs are
  192.          enabled. You'll  have to  disable EAs  with -2e  prior  to  using
  193.          -jn, or to recreate the archive if the EAs are precious.
  194.  
  195.       *  Hollow mode archives do not support EAs.
  196.  
  197.       *  Under Windows NT, extended attributes cannot be overwritten. That
  198.          is, if the EA data is appended to  a file which  already contains
  199.          EAs at the time of unarchiving, the file will retain its original
  200.          EAs.
  201.  
  202.       A set of new options has been introduced to let  the user control EA
  203.       handling:
  204.  
  205.       *  "-2c" restricts  EA  handling  to  critical  EAs  only.  Archived
  206.          non-critical  EAs  will not  be  restored.  When  an  archive  is
  207.          created, only critical EAs will be saved.
  208.  
  209.       *  "-2e" specifies EA inclusion  filter. With no  parameters  given,
  210.          it disables EA  handling at  all. Otherwise, an  expression  that
  211.          follows  it  is  interpreted  as  a   wildcard  that   limits  EA
  212.          inclusion to a  particular EAs. Multiple  options can be  entered
  213.          to represent a set of EA names but list files are not allowed.
  214.          Examples:
  215.  
  216.          ARJ a test
  217.  
  218.          In this example, all EAs will be preserved.
  219.  
  220.          ARJ a no_eas -2e
  221.          ARJ x no_eas -2e
  222.  
  223.          EAs will neither be packed nor restored.
  224.  
  225.          ARJ a documents -2e.LONGNAME
  226.  
  227.          In this case, only .LONGNAME EAs will be handled.
  228.  
  229.          ARJ a test -2e.CLASSINFO -2e.ICON*
  230.  
  231.          .CLASSINFO and .ICON* (i.e. .ICON, .ICON1, .ICONPOS) EAs  will be
  232.          be packed and restored.
  233.  
  234.          It's wise  to specify "-2e.*" when backing up  your OS/2  desktop
  235.          or  configuration  files.  The system  EAs  start with  dot (".")
  236.          while application EAs start with application name.
  237.  
  238.       *  "-2l" allows to  convert  .LONGNAME  extended  attributes  (these
  239.          represent icon titles used in  WPS) to file names, when possible.
  240.          This feature  simplifies moving document  files away from an OS/2
  241.          system installed on a FAT volume.
  242.  
  243.          If  the icon  title (and  so  the  extended  attribute)  contains
  244.          line breaks, wildcard characters  or other symbols, real filename
  245.          will be used instead and the .LONGNAME EA will be preserved.
  246.  
  247.          This option is ignored during extraction. "-2e" and "-2x" have no
  248.          effect  on  this  option  (but  .LONGNAME EAs are  not  saved  if
  249.          .LONGNAME EA handling is implicitly or explicitly disabled).
  250.  
  251.       *  "-2x" specifies an  exclusion  filter. It must be followed  by an
  252.          exclusion  EA name  specification. The  rules  are  the  same  as
  253.          with "-2e". Also, the  two options  may  work together, providing
  254.          both an inclusion and an exclusion rule. For example:
  255.  
  256.          ARJ a backup_ -r -p1 -h#2 -2e.* -2x.FED* c:\projects
  257.  
  258.          may be used to  create  regular back-ups of your  work directory,
  259.          including all system EAs  but  excluding  EAs  created  with FED
  260.          (Fast Editor  Lite, an  editor written  by Sergey I. Yevtushenko,
  261.          evsi@@naverex.kiev.ua) - that program  does not follow traditional
  262.          EA naming  conventions  and  uses  system-alike  EAs  for  anchor
  263.          position marks.
  264.  
  265.       Extended  attributes are  also  supported in ARJSFXV self-extractors
  266.       where they are stored using the  same  technology as with  usual ARJ
  267.       archives.
  268.  
  269.       The presence of EAs  is  indicated  by a "(EA: ...)" message when  a
  270.       file is packed. Note that this size  may differ from  the one  given
  271.       when the file  is unpacked - the  former is  the EA  structure  size
  272.       and the latter is the space allocated  for EA storage. The number of
  273.       EAs and the size of EA structure is  also displayed when the archive
  274.       is listed with "ARJ v" command.
  275.  
  276.  
  277.    FREQUENTLY ASKED QUESTIONS
  278.  
  279.       Q: Third-party applications can't handle ARJ for @PLATFORM archives!
  280.       A: Try to disable extended attributes (-2e), DTA/DTC  storage (-j$),
  281.          hard  links  (-2h),  and   enforce  the  DOS  mode  (-2d).   Many
  282.          applications   are  incapable  of  handling  new  archive  format
  283.          (although  this  format is  fully compliant  with the  documented
  284.          guidelines). Known  examples  of such  applications  include File
  285.          Commander v 2.11, Norton Commander v 5.00 and WinRAR v 2.60.
  286.  
  287.       Q: Extended attribute  sizes reported  by  InfoZIP and  ARJ  differ.
  288.          What's the cause?
  289.       A: As we have  stated earlier, ARJ reports  the size of its internal
  290.          EA storage structure as the  EA size when archiving files. Across
  291.          various platforms   (OS/2-16, OS/2-32, NT)  there are  various EA
  292.          structures.  The system is   questioned for  the  actual  EA size
  293.          during archiving.
  294.  
  295.       Q: How can I back up my OS/2 Workplace Shell folders, preserving the
  296.          icons?
  297.       A: Since  folders are  represented  with directories, you'll have to
  298.          enable directory storage with -a1 or -hbfd.
  299.  
  300.       Q: The EAs have vanished after I used ARJ/DOS to update an archive.
  301.       A: Current versions  of ARJ/DOS and ARJ32, as well as ARJ/2 prior to
  302.          2.62.10, strip  the  extended  headers when  any kind  of archive
  303.          update occurs.
  304.  
  305.       Q: I want  to  create  single-volume  self-extracting  archive  that
  306.          supports EAs but ARJ/2 uses ARJSFX instead of ARJSFXV.
  307.       A: You need to  force  use of  ARJSFXV/2. The best way for  it is to
  308.          specify an arbitrary large value for volume size, e.g. -va.
  309.  
  310.       Q: How can I create an installer for my OS/2 product with ARJ?
  311.       A: ARJSFX/2 is  able  to run OS/2  commands after unpacking archive.
  312.          Try this: create a  script you want  to to be  invoked  after the
  313.          installation completes.
  314.  
  315.          e.g., INSTALL2.CMD:
  316.          =======
  317.           /* REXX */
  318.  
  319.           if RxFuncQuery('SysLoadFuncs') then do
  320.            call RxFuncAdd 'SysLoadFuncs', 'RexxUtil', 'SysLoadFuncs'
  321.            call SysLoadFuncs
  322.           end
  323.  
  324.           say "Installation has completed, creating desktop object..."
  325.           call SysCreateObject ....
  326.          ========
  327.          and so on...
  328.  
  329.          Now create an archive comment with the first line as follows:
  330.          )) \InstallDir\ -b -x -y -!INSTALL2.CMD
  331.  
  332.          Call it, for example, CMT.ANS. Now create an ARJSFX archive:
  333.  
  334.          ARJ a PACKAGE.EXE -je -r -a -jm -zCMT.ANS -xCMT.ANS
  335.  
  336.          You'll make  a  self-extracting  archive  with  an  automatically
  337.          invoked installation program.
  338.  
  339.       Q: How can I distinguish between ARJ/ARJZ/ARJ32 archives?
  340.       A: Use the ARJ V  command. The "Host OS" field contains  the type of
  341.          host OS. The  "Revision"  field  may  be used  to  determine  the
  342.          archiver version that added the file:
  343.  
  344.                  1 = ARJ versions earlier than 0.14,
  345.                      ARJZ compatibility mode (-md is less than 26624)
  346.                  2 = ARJ v 0.14...0.20
  347.                  3 = ARJ v 1.00...2.22
  348.                  4 = ARJ v 2.30, X1
  349.                  5 = ARJ v 2.39a, 2.39b
  350.                  6 = ARJ v 2.39c...2.41
  351.                  7 = ARJ v 2.42a...2.50a
  352.                  8 = ARJ v 2.55...2.61, ARJ/2 v 2.61
  353.                  9 = ARJ v 2.62, ARJ/2 v 2.62, ARJ32 v 3.00
  354.                 10 = ARJ v 2.70 and higher, ARJ/2 v 2.62.10 and higher
  355.                 11 = ARJ with UNIX support (2.77/3.10 and higher)
  356.                 50 = ARJZ with maximum distance up to 32K
  357.                 51 = ARJZ with maximum distance up to 64K
  358.                100 = ARJ32 v 3.00b...3.01
  359.                101 = ARJ32 v 3.02 and higher
  360.  
  361.          Notes:
  362.  
  363.          1. ARJ versions  that created the  Revision 1 header have  used a
  364.             different method 4 compression. Support for it  was dropped in
  365.             versions 1.xx. Such  archives may  be incorrectly processed by
  366.             @PRODUCT.
  367.  
  368.          2. ARJ v 2.76.07 and higher can  read the  newer UNIX time format
  369.             of  ARJ  v  3.10/2.77. The  intention was  to make  the stable
  370.             versions compatible with it.
  371.  
  372.  
  373.    INFORMATION FOR DEVELOPERS
  374.  
  375.       The new UNIX time  format  can be identified  by "Host OS" equal  to
  376.       "UNIX" or "NeXT", and "arj_nbr" greater than or equal to 11. In this
  377.       case, all of the time fields in the  corresponding header are in the
  378.       UNIX  time  format, i.e. 4-byte  value specifying number  of seconds
  379.       passed since 01/01/1970, 00:00:00 UTC.
  380.  
  381.       By other means, the header format  is 100% compatible  with standard
  382.       ARJ format, but we  utilize the  extended  header fields. Here  is a
  383.       brief overview of this  technology. The extended header layout is as
  384.       follows (all values are little-endian):
  385.  
  386.       Bytes  Description
  387.       -----  -----------
  388.           1  Extended header ID.
  389.           1  Continuation  flag. If set to  0, marks the  end of  block
  390.              chain so the header data can be concatenated and processed.
  391.              Also it provides a way of checking for trashed blocks.
  392.           ?  Header data.
  393.  
  394.       A standard CRC32 of the whole header, including the ID but excluding
  395.       header size field,  is  appended to  it. It is strongly  recommended
  396.       that the CRC is verified before any further processing occurs.
  397.  
  398.       ID 0x45 ('E') == Extended attributes
  399.       ------------------------------------
  400.  
  401.       This is only valid if  the arj_nbr (header revision number) is 10 or
  402.       greater.
  403.  
  404.       The packed  EA  block is a  complex  structure  that  can span  over
  405.       multiple  volumes. In  case of  such  spanning,  separate  parts  of
  406.       the block  are stored  in  separate extended  headers  on  different
  407.       volumes and they must be joined together when the last block is read
  408.       (it's recognized   by  EXTFILE_FLAG being clear). The layout of  the
  409.       packed block follows:
  410.  
  411.       Bytes  Description
  412.       -----  -----------
  413.           1  Compression method (0...4, may differ from the  one found  in
  414.              the file header)
  415.           2  Unpacked EA data size in bytes
  416.           4  CRC32 of unpacked EA data
  417.           ?  Raw packed EA data
  418.  
  419.       The raw  packed data  may  be decompressed  using the  original  ARJ
  420.       algorithms. In the case when the file is encrypted, the packed block
  421.       is also  encrypted (but the  garble  routine is  reinitialized  when
  422.       compression  of the EA occurs). The password modifier is the same as
  423.       for the  first file section. After  decompression, the  following EA
  424.       structure will exist:
  425.  
  426.       Bytes  Description
  427.       -----  -----------
  428.           2  Total number of EA records
  429.           ?  Extended attribute records
  430.  
  431.       The extended attribute records are merged altogether. They should be
  432.       processed sequentially. A single record  represents a single EA, and
  433.       no EA can be represented twice. The layout is as follows:
  434.  
  435.       Bytes  Description
  436.       -----  -----------
  437.           1  fEA byte (may indicate a critical EA)
  438.           1  Size of extended attribute name
  439.           2  Size of extended attribute value
  440.           ?  Extended attribute name (not ASCIIZ)
  441.           ?  Extended attribute value (binary data)
  442.  
  443.       Even if the file is  a text  one, the EAs must be  handled as binary
  444.       data during  compression and  extraction. When EAs are spanned  over
  445.       multiple volumes, neither  the packed  block header  is repeated nor
  446.       the compression  is  restarted (actually, the whole block  including
  447.       its header is created in memory and later split to volumes).
  448.  
  449.       ID 0x4F ('O') == Owner information
  450.       ----------------------------------
  451.  
  452.       Contains the owner  information, and possibly, group information, in
  453.       character form. Valid if arj_nbr is 11 or greater.
  454.  
  455.       Bytes  Description
  456.       -----  -----------
  457.           1  Owner's name length.
  458.           ?  Owner's name (non-ASCIIZ)
  459.              OR
  460.              Owner's name (ASCIIz), followed by
  461.           ?  Group name (non-ASCIIz)
  462.  
  463.       ID 0x55 ('U') == UNIX special files
  464.       -----------------------------------
  465.  
  466.       The UNIX special file entry is  a complex  freeform structure, valid
  467.       if arj_nbr is 11 or greater:
  468.  
  469.       Bytes  Description
  470.       -----  -----------
  471.           1  Special file descriptor:
  472.              Bits 0...4 = data size (0 allowed, 31 has a special meaning)
  473.              Bits 5...7 = type:
  474.                           000 = pipe
  475.                           001 = hard link
  476.                           002 = symbolic link
  477.                           003 = block device
  478.                           004 = character device
  479.           ?  If the size field in descriptor  contained 31, then two bytes
  480.              here contain the size, otherwise there is no area between the
  481.              descriptor and data.
  482.           ?  Raw data. Size is reported by the descriptor or the dedicated
  483.              size field. Format:
  484.  
  485.              Pipe:      Empty (size is zero).
  486.              Link:  Target file (non-ASCIIZ).
  487.                Device:  The dev_t  structure  in the  host OS' format  and
  488.                         endia order.
  489.  
  490.       ID 0x6F ('o') == Owner information (numeric)
  491.       --------------------------------------------
  492.  
  493.       Contains the owner's UID and GID. Valid if arj_nbr is 11 or greater.
  494.  
  495.       Bytes  Description
  496.       -----  -----------
  497.           1  Data length (must be 8 for the current implementation).
  498.           4  Owner's UID (little-endian).
  499.           4  Owner's GID (little-endian).
  500.  
  501.  
  502.    COMPATIBILITY ISSUES
  503.  
  504.       ARJ has  been  briefly tested  on the  following platforms  and file
  505.       systems:
  506.  
  507.       DOS (FAT)
  508.  
  509.         MS-DOS v 2.11, 3.20, 3.21, 4.01, 5.00, 6.00, 6.20, 6.22
  510.         PC DOS v 6.30, 7.00
  511.         Windows 95, 98
  512.         Windows NT Workstation v 3.51, 4.00
  513.  
  514.       OS/2 (FAT, HPFS, Ext2FS, JFS, NTFS)
  515.  
  516.         Microsoft OS/2 v 1.21
  517.         IBM OS/2 v 1.30, 2.00, 2.10, 3.00, 4.00, 4.50
  518.         OS/2 subsystem in Windows NT v 3.51, 4.00, Windows 2000
  519.  
  520.       Win32 (FAT, HPFS, NTFS)
  521.  
  522.         Windows 95, 98, ME
  523.         Windows NT v 3.51, 4.00, Windows 2000, Windows XP
  524.  
  525.       Linux (UMSDOS, HPFS, Ext2FS, Ext3FS, JFS)
  526.  
  527.         Linux v 2.2.13/19-20, 2.4.5/18-23, 2.6.3
  528.         glibc v 2.1.2, 2.1.3, 2.2.3
  529.  
  530.       FreeBSD (FAT, UFS, HPFS)
  531.  
  532.         FreeBSD v 3.4/STABLE
  533.  
  534.       QNX (QNX4FS, FAT)
  535.  
  536.         QNX v 6.2.1/PE Patch B
  537.  
  538.  
  539.       End of document

Raw Paste


Login or Register to edit or fork this paste. It's free.