I would say that "inertia" is PL/X's raison d'etre (even though that statement 
is probably controversial within the internal IBM Z development community).

I will acknowledge that PL/X is excellent at integrating HLASM code. GCC style 
inlining isn't terrible for including HLASM code but it is more painful.

I think the reason that PL/X was initially used is because it was the best we 
had at the time. The reason it is STILL used is because it's too 
dangerous/difficult to switch.

I really don't know why it was never truly externalized (that decision predates 
my time at IBM) but I could speculate that someone thought it gave IBM some 
competitive advantage. I don't see it that way, but again, just speculation.

Eric Rossman

-----Original Message-----
From: IBM Mainframe Discussion List <[email protected]> On Behalf Of 
Kirk Wolf
Sent: Monday, October 2, 2023 2:55 PM
To: [email protected]
Subject: [EXTERNAL] Re: PL/X Open Source and PL/I - Helping to save the world 
and cut CPU Cycles and electricity

Eric, 

I'm curious - wouldn't you say that PL/X integration with assembler and 
assembler macros is it's raison d'etre?   Even though I've done all sorts of 
integration of assembler with C/C++  (the GCC-style inlining, xplink assembler 
leaf routines, EDCDSECT conversion of DSECTs, etc, etc), which all work, they 
are still painful compared with PL/X and assembler.

Kirk Wolf
Dovetailed Technologies
http:// <http://dovetail.com >coztoolkit.com

On Mon, Oct 2, 2023, at 12:02 PM, Eric D Rossman wrote:
> I write PL/X daily. PL/X, on the average, is not really better than C 
> in terms of what you describe except when the string's length is known 
> in advance (which is hard or impossible in many circumstances
> 
> Don't get me wrong, it has a number of strengths as compared to C, but it 
> also is too close to "the metal" in some ways which would hamper it.
> 
> As for copying byte at a time, that is not a function of C (i.e. not 
> specified in the standard). It's usually a function of the complier 
> (sometimes deferred to the runtime libraries) and many of them can use 
> benefits of the instructions built into the hardware to speed things up as 
> well as general purpose things like copying DWORDs at a time with small 
> unrolled loops on either end to handle "extra"
> 
> Eric Rossman
> 
> -----Original Message-----
> From: IBM Mainframe Discussion List <[email protected]> On 
> Behalf Of Clem Clarke
> Sent: Monday, October 2, 2023 9:03 AM
> To: [email protected]
> Subject: [EXTERNAL] Re: PL/X Open Source and PL/I - Helping to save 
> the world and cut CPU Cycles and electricity
> 
> What would it take for IBM to Open Source the Windows and Linux version of 
> PL/I and PL/X?
> 
> Why?  To potentially make the Internet faster and safer.  How?
> 
> We know that C searches for a byte with a binary zero to find how long a 
> string is.  This takes time. And then it take time to copy a string 
> elsewhere, especially if it is done a byte at a time (often true, depending 
> on the C compiler - some do a word at a time).
> 
> PL/I, Pascal and even Assembler know how long a string is.  They don't have 
> to waste cycle looking for the length of a string. Most of the time, they 
> know how long the receiving string is, and won't go past the end, as C will.
> 
> IBM still has the "authority" to do this.  And it morally should.
> 
> Just do it, IBM.  Help save the planet.
> 
> Clem Clarke
> 
> 
> 
> 
> 
> Wayne Bickerdike wrote:
> > So many acronyms.
> > I've Been Married
> > I've Been Moved
> > It's Better Manual
> > I Broke Microcode
> >
> > etc..
> >
> > On Mon, Oct 2, 2023 at 4:17 AM David Spiegel < 
> > [email protected]> wrote:
> >
> >> Hi Peter,
> >> I was generalizing the problem. Allowing access to PL/xxxx wouyld 
> >> also solve the lack of PDFs.
> >>
> >> This reminds me of a joke.
> >> Q: What does IBM Stand for?
> >> A: Ich Bin M'shugoh
> >>
> >> Regards,
> >> David
> >>
> >> On 2023-09-30 08:18, Peter Relson wrote:
> >>>> There is another solution
> >>> What are you thinking the "problem" is for which you mention a
> >> "solution"? The first post I saw was asking about PDF's, not about 
> >> access to PL/X. Was there a post that did not show up in the daily 
> >> digest? The "access-to-PL/X ship" sailed long ago.
> >>> Peter Relson
> >>> z/OS Core Technology Design
> >>>
> >>>
> >>> ------------------------------------------------------------------
> >>> --
> >>> -- For IBM-MAIN subscribe / signoff / archive access instructions, 
> >>> send email to [email protected] with the message: INFO 
> >>> IBM-MAIN
> >> -------------------------------------------------------------------
> >> --
> >> - For IBM-MAIN subscribe / signoff / archive access instructions, 
> >> send email to [email protected] with the message: INFO 
> >> IBM-MAIN
> >>
> >
> 
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions, send 
> email to [email protected] with the message: INFO IBM-MAIN
> 
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions, send 
> email to [email protected] with the message: INFO IBM-MAIN
> 

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions, send email to 
[email protected] with the message: INFO IBM-MAIN

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to