Thinking about the ability to use BOTH Java and NetRexx Syntax & SEMANTICS

[ThSITC]

Hello all,

    again, publically thinking (sorry, I'm a lonely man)

I'm thinking (and implementing) my own NetRexx extensions (formerly
called Rey (Rexx for Java), still called Rey)
in this way, that:

    1.) Rey does even accept Java Syntax.
    2.) Thus, a trained Java Prrogrammer willl be able to use his well
known syntax (and semantics)
    3.) Thus, Rey is a PREPROCESSER accepting Java and/or NetRexx statements

As far as I can see, there (currently) *are NO conflicts!

*OR*

  is SIGNAL xxxxx *in NetRexx* *anything else* than THROW xxxx in Java?

Please do advise! :-)
Thomas

PS: More important, it would/will be necessary to have the abiliy to use
*any* NUMERIC
implementation (e.g. Mike's BigDecimal vs. Java's BigDecimal vs.
Hardware zOS vs.
current class Rexx implementation) b choice (OPTION BIGDECIMAL)

After having written so much (too much :-() I will again turn into my
*quiet Mode*
(most difficult to me, you see::::) finishing my proposed, partly new,
implementation
of NetRexx :-(

  .

--
Thomas Schneider (Founder of www.thsitc.com) Member of the Rexx Languge
Asscociation (www.rexxla.org) Member of the NetRexx Developer's Team
(www.netrexx.org)

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[alansam]

On 8 November 2011 10:51, Thomas Schneider <[hidden email]> wrote:

Hello all,

PS: More important, it would/will be necessary to have the abiliy to use *any* NUMERIC
implementation (e.g. Mike's BigDecimal vs. Java's BigDecimal vs. Hardware zOS vs.
current class Rexx implementation) b choice (OPTION BIGDECIMAL)

Thomas,

You've mentioned this several times and in the past I've ignored it but I am curious to know how you intend to enforce the hardware support for BigDecimal on System z within NetRexx.  (BTW As the support is built into the processor it's not just for zOS.  It will work on any operating system running System z so zVM and Linux (VM guest or native) will be able to avail itself of the feature too.)

As IBM's System z JVMs already use the new instructions if they're available, how do you propose improving on that?

Alan.

--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

[ThSITC]

Hello Alan,
    1.) When I would be asked (probably by Rene) I would
        a) Separate Mike's Numeric Code from the Current class Rexx to again RexxNum (as he had implemented
originally, as can be seen from his original comments in Rexxx.Nrx)
        b) Separate *all Collection issues* to a separate class (probably RexxColl)
       Compile it again, and see if that works

    2.) When YES: implement a new option BIGDECIMAL0,1,2,3,4,5,6 as already proposed to KENAI.
I   3.) Modify the remaining class Rexx (or the new class RexxNum) to handle any/all Numerics
    4.) Most probably, RexxNum has simply to be an ABSTRACT Class, defining Numerics,
and the calling the proper implemented BigDecimal implementation
    5.) I did meanwhile imlement class Packed (IBM Packed Decimals) which, as Mike said,
when we did last time converse about this, is part of the C Version of his BigDecimal, but
not yet of his Java implementation.
    6.) Document the stuff.
    7) Write test-cases, and collect testcases (first: old NetRexx RexxNum, Mike's BigDecimal, Java's BigDecimal)
         7.a) first release to org.netrexx for tests and QA (Rob Wilson)
 .  8) ReREAD the IBM API's (zOS & Power Systems) for hardware implementation)
    9) Modification of the sources, again, and re-test.


Alan, I did change the SUBJECT of this message. Hope you do understand.

What I *do think* is that Mike did an excellent Job, both designing NetRexx and implementing it,
but we all together will have to update both the language and the implementation to
the state of the art (Java 1.7, which I'm running) very soon, so that we are not loosing
the edge (if that is the correct english word).

Thomas.
=======================================================================================
 
Am 08.11.2011 20:11, schrieb Alan Sampson:

On 8 November 2011 10:51, Thomas Schneider <[hidden email]> wrote:

Hello all,

PS: More important, it would/will be necessary to have the abiliy to use *any* NUMERIC
implementation (e.g. Mike's BigDecimal vs. Java's BigDecimal vs. Hardware zOS vs.
current class Rexx implementation) b choice (OPTION BIGDECIMAL)

Thomas,

You've mentioned this several times and in the past I've ignored it but I am curious to know how you intend to enforce the hardware support for BigDecimal on System z within NetRexx.  (BTW As the support is built into the processor it's not just for zOS.  It will work on any operating system running System z so zVM and Linux (VM guest or native) will be able to avail itself of the feature too.)

As IBM's System z JVMs already use the new instructions if they're available, how do you propose improving on that?

Alan.

--
Can't tweet, won't tweet!

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/