• Re: Eliza in COBOL

    From J Leslie Turriff@jlturriff@mail.com to comp.lang.cobol on Tue Sep 25 00:26:44 2018
    From Newsgroup: comp.lang.cobol

    Arnold Trembley wrote:

    If anyone remembers ELIZA in Basic, now it's been translated into COBOL.

    https://en.wikipedia.org/wiki/ELIZA

    Here's the BASIC source code:
    http://hps.elte.hu/~gk/Eliza/ELIZA.BAS

    Here's the COBOL source code:

    https://sourceforge.net/p/open-cobol/contrib/HEAD/tree/trunk/samples/eliza/

    And here's an installable version for Windows: http://www.arnoldtrembley.com/Eliza-setup-rename-7z-to-exe.7z

    Kind regards,

    Eliza was interesting, but I really wish I could find a working
    copy of Racter...
    --
    JLT
    --- Synchronet 3.20a-Linux NewsLink 1.114
  • From Clark F Morris@cfmpublic@ns.sympatico.ca to comp.lang.cobol on Wed Sep 26 19:42:50 2018
    From Newsgroup: comp.lang.cobol

    READ INTO was required in many IBM shops because addressability to the
    read buffer was dropped at end of file rather than at CLOSE. This was corrected in COBOL for MVS and VM or COBOL II V4. Also finding the
    record in the buffer in a dump was difficult until LE (Language
    Environment) was instituted in the mid 1990s. WRITE FROM combined
    with APPLY WRITE-ONLY optimized the creation of buffers for variable
    length record files by allowing the maximum length block to be
    written. It also allowed not having to build the entire output
    record. Again locating the record to be written in a dump was
    difficult prior to LE unless you had a proprietary product such as
    ABEND-AID. These considerations applied to OS360 and MVS and may have
    applied in other environments (and may still apply).

    Direct and keyed access files were inherent in the IBM DOS360 / VSE
    and OS360 with BDAM (Basic Direct Access Method), ISAM (Indexed
    Sequential Access Method) and VSAM (Virtual Sequential Access Method)
    being a part of the operating system. COBOL did not exploit all of
    the facilities of these file access facilities. Most of the tools
    such as the SORT, appropriate IBM utilities and third party languages
    such as DYL280 and Easytrieve provided means to access files created
    using these facilities. So far as I know neither Windows nor any of
    the UNIX / LINUX families have native support for either direct or
    indexed access. Interestingly database access also seems to be
    proprietary or at least an add-on to the operating environment for all operating systems and there are definite differences in the varying
    flavors of SQL let alone the other technologies.

    In short, if an organization is going to have standards probably the
    reason for each standard should be stated with the specific standard
    so that when the environment changes through upgrades or migrations
    (Mainframe to distributed and vice versa for example), the standards
    can be updated to reflect the new environment.

    Clark Morris
    --- Synchronet 3.20a-Linux NewsLink 1.114
  • From Robert Wessel@robertwessel2@yahoo.com to comp.lang.cobol on Thu Sep 27 10:33:50 2018
    From Newsgroup: comp.lang.cobol

    On Wed, 26 Sep 2018 19:42:50 -0300, Clark F Morris
    <cfmpublic@ns.sympatico.ca> wrote:

    Direct and keyed access files were inherent in the IBM DOS360 / VSE
    and OS360 with BDAM (Basic Direct Access Method), ISAM (Indexed
    Sequential Access Method) and VSAM (Virtual Sequential Access Method)
    being a part of the operating system. COBOL did not exploit all of
    the facilities of these file access facilities. Most of the tools
    such as the SORT, appropriate IBM utilities and third party languages
    such as DYL280 and Easytrieve provided means to access files created
    using these facilities. So far as I know neither Windows nor any of
    the UNIX / LINUX families have native support for either direct or
    indexed access.


    Direct-type files are trivial if you assume binary byte-stream files,
    as supported by most OSs, including Windows and *nix.

    It's true that indexed files have never really been a standard feature
    of those OS's.


    Interestingly database access also seems to be
    proprietary or at least an add-on to the operating environment for all >operating systems and there are definite differences in the varying
    flavors of SQL let alone the other technologies.


    While I agree that a database *should* be standard, and it should be
    easier to define small databases for an application's use, it's pretty straight-forward, and free, to install a DB on either *nix or Windows.
    --- Synchronet 3.20a-Linux NewsLink 1.114
  • From Arnold Trembley@arnold.trembley@att.net to comp.lang.cobol on Thu Sep 27 20:59:12 2018
    From Newsgroup: comp.lang.cobol

    On 9/25/2018 12:26 AM, J Leslie Turriff wrote:
    Arnold Trembley wrote:

    If anyone remembers ELIZA in Basic, now it's been translated into COBOL.

    https://en.wikipedia.org/wiki/ELIZA

    Here's the BASIC source code:
    http://hps.elte.hu/~gk/Eliza/ELIZA.BAS

    Here's the COBOL source code:

    https://sourceforge.net/p/open-cobol/contrib/HEAD/tree/trunk/samples/eliza/

    And here's an installable version for Windows:
    http://www.arnoldtrembley.com/Eliza-setup-rename-7z-to-exe.7z

    Kind regards,

    Eliza was interesting, but I really wish I could find a working
    copy of Racter...


    Here are some possibilities...

    https://github.com/fergusq/racter

    https://www.cs.cmu.edu/Groups/AI/areas/classics/racter/0.html

    https://www.cs.cmu.edu/Groups/AI/areas/classics/racter/

    https://news.ycombinator.com/item?id=7161285

    http://boston.conman.org/2008/06/18.2

    Kind regards!
    --
    https://www.arnoldtrembley.com/

    ---
    This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

    --- Synchronet 3.20a-Linux NewsLink 1.114