"Mary Degallo" <[log in to unmask]> wrote in message
news:f0bf00c.0204260812.7579f4d1@posting.google.com...
> hi Everybody!
>
>    thanx for those participating in the discussion and providing valuable
>    help to newbies like me.
>
>    ---------------------------------------
>    FILENAME known at RunTime
>    FILENAME is NOT known at COMPILE time.
>    ---------------------------------------
>    C, C++, Java provide a file name that can be specified at RunTime.
>    Where as the way the FD and File Sections in COBOL are,
>     we have to specify the file name at COMPILE time.
>    We have to design our applications around this "feature/limitation".
>    Has anybody, worked around this, using some other "tricks".
>    JCL is one way, but we donot want to get into JCL stuff.
>    How does one do it in a more elegant way?

Hmm. If you don't want to specify the file name at compile time, and don't
want to specify the file name at execution time (via JCL), you could ask the
operator.

>
>
>
>    ----------------------------------------
>    HP3000 buffer sizes etc,...
>    ----------------------------------------
>    We have to read this data from a THIRD PARTY software objects.
>    And these Objects can run into 1 or 2 or 4 or somtimes even 7 Megs of
Data.
>    Do I have to define such a HUGE working Storage Area, to read the
>    Object data Into?
>
>    I could define a 1K or 2K kind of Buffers and tested my code on AS400.
>
>    How will the scenario change when it becomes, a 1 MEG buffer?
>     1.   What are the issues involved.
>     2.   Can we do it?
>
>    Some geeks in my office are playing "C++, JAVA vs  COBOL"  politics  in
>     issues like these.    Well its not new anyways.      :)

I have no idea on the third-party software objects. But you've got to stamp
out this Java vs. COBOL shit before it gets out of hand. I recommend killing
the Java people.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *