On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:it might be a foot gun.
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The
major feature of this release is support for self-defining structures,
declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete
information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work, its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND (technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an object? Of course you included it to be compatible with IBM, but why is it necessary to begin with? It looks like
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The
major feature of this release is support for self-defining structures, declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete information.
http://www.iron-spring.com
On 1/27/23 4:33 PM, wat fiv wrote:like it might be a foot gun.
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The
major feature of this release is support for self-defining structures,
declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete >> information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work, its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND (technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an object? Of course you included it to be compatible with IBM, but why is it necessary to begin with? It looks
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the structure, so the REFER is necessary to make HBOUND and LBOUND work in
the first place. Languages with substantial differences from PL/I (like, say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it stands.
--
John W. Kennedy
Algernon Burbage, Lord Roderick, Father Martin, Bishop Baldwin,
King Pellinore, Captain Bailey, Merlin -- A Kingdom for a Stage!
On 1/27/23 4:33 PM, wat fiv wrote:
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The
major feature of this release is support for self-defining structures,
declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete
information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work,
its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND
(technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an
object? Of course you included it to be compatible with IBM, but why is
it necessary to begin with? It looks like it might be a foot gun.
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the structure, so the REFER is necessary to make HBOUND and LBOUND work in
the first place. Languages with substantial differences from PL/I (like,
say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it stands.
John W Kennedy <john.w.kennedy@gmail.com> wrote:
On 1/27/23 4:33 PM, wat fiv wrote:
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The
major feature of this release is support for self-defining structures, >>>> declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete >>>> information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work,
its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND
(technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an
object? Of course you included it to be compatible with IBM, but why is
it necessary to begin with? It looks like it might be a foot gun.
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the
structure, so the REFER is necessary to make HBOUND and LBOUND work in
the first place. Languages with substantial differences from PL/I (like,
say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it
stands.
Good explanation! I always thought of it (as far as I thought of it at all) as a more complicated version of COBOL’s “OCCURS DEPENDING ON.”
By the way, please hold off on downloading or using this feature for a
while, since a serious bug just surfaced that I will be working on anon.
On 1/28/23 9:18 AM, Peter Flass wrote:
John W Kennedy <john.w.kennedy@gmail.com> wrote:
On 1/27/23 4:33 PM, wat fiv wrote:
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The >>>>> major feature of this release is support for self-defining structures, >>>>> declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete >>>>> information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work,
its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND >>>> (technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an
object? Of course you included it to be compatible with IBM, but why is >>>> it necessary to begin with? It looks like it might be a foot gun.
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the
structure, so the REFER is necessary to make HBOUND and LBOUND work in
the first place. Languages with substantial differences from PL/I (like, >>> say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it
stands.
Good explanation! I always thought of it (as far as I thought of it at all) >> as a more complicated version of COBOL’s “OCCURS DEPENDING ON.”
Essentially, that’s what it is. Internally, a COBOL READ (without INTO)
is equivalent to PL/I READ SET, and a COBOL WRITE (without FROM) is functionally equivalent to PL/I LOCATE, but the first LOCATE is executed
for free after OPEN, each WRITE is followed by a free LOCATE, and CLOSE performs an undocumented un-LOCATE (in OS/360 and its descendants, using
an undocumented un-PUT function in QSAM, QISAM, etc.).
By the way, please hold off on downloading or using this feature for a
while, since a serious bug just surfaced that I will be working on anon.
Don’t mean nuttin’ to me unless you create a version for macOS (or at least for Darwin).
John W Kennedy <john.w....@gmail.com> wrote:linux subsystem for windows?
On 1/28/23 9:18 AM, Peter Flass wrote:
John W Kennedy <john.w....@gmail.com> wrote:
On 1/27/23 4:33 PM, wat fiv wrote:
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The >>>>> major feature of this release is support for self-defining structures, >>>>> declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete
information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work, >>>> its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND >>>> (technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an
object? Of course you included it to be compatible with IBM, but why is >>>> it necessary to begin with? It looks like it might be a foot gun.
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the >>> structure, so the REFER is necessary to make HBOUND and LBOUND work in >>> the first place. Languages with substantial differences from PL/I (like, >>> say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it >>> stands.
Good explanation! I always thought of it (as far as I thought of it at all)
as a more complicated version of COBOL’s “OCCURS DEPENDING ON.”
Essentially, that’s what it is. Internally, a COBOL READ (without INTO) is equivalent to PL/I READ SET, and a COBOL WRITE (without FROM) is functionally equivalent to PL/I LOCATE, but the first LOCATE is executed for free after OPEN, each WRITE is followed by a free LOCATE, and CLOSE performs an undocumented un-LOCATE (in OS/360 and its descendants, using an undocumented un-PUT function in QSAM, QISAM, etc.).
By the way, please hold off on downloading or using this feature for a
while, since a serious bug just surfaced that I will be working on anon.
Don’t mean nuttin’ to me unless you create a version for macOS (or at least for Darwin).
Windoze next :-(
--
Pete
bearlyabus...@gmail.com schrieb am Samstag, 28. Jänner 2023 um 23:40:38 UTC+1:
John W Kennedy <john.w....@gmail.com> wrote:linux subsystem for windows?
On 1/28/23 9:18 AM, Peter Flass wrote:Windoze next :-(
John W Kennedy <john.w....@gmail.com> wrote:
On 1/27/23 4:33 PM, wat fiv wrote:
On Friday, January 27, 2023 at 9:46:53 AM UTC-5, bearlyabus...@gmail.com wrote:
Version 1.0.0 of the Iron Spring PL/I compiler is now available. The >>>>>>> major feature of this release is support for self-defining structures, >>>>>>> declared with the REFER option. It also includes other minor
enhancements and bug fixes. Consult the documentation for more complete >>>>>>> information.
http://www.iron-spring.com
Congratulations on making version 1.0.0! And thank you for your work, >>>>>> its a shame that IBM doesn't do this anymore.
I've never understood the point of refer; can't one use just use HBOUND >>>>>> (technically, HBOUND-LBOUND+1) or MAXLENGTH to find the size of an >>>>>> object? Of course you included it to be compatible with IBM, but why is >>>>>> it necessary to begin with? It looks like it might be a foot gun.
Since the variable is BASED, it has no dope vector or descriptor
associated with it. It has only the naked data contained within in the >>>>> structure, so the REFER is necessary to make HBOUND and LBOUND work in >>>>> the first place. Languages with substantial differences from PL/I (like, >>>>> say, Swift) might find a way to work around that, but READ SET
absolutely demands that the structure be entirely self-describing as it >>>>> stands.
Good explanation! I always thought of it (as far as I thought of it at all)
as a more complicated version of COBOL’s “OCCURS DEPENDING ON.”
Essentially, that’s what it is. Internally, a COBOL READ (without INTO) >>> is equivalent to PL/I READ SET, and a COBOL WRITE (without FROM) is
functionally equivalent to PL/I LOCATE, but the first LOCATE is executed >>> for free after OPEN, each WRITE is followed by a free LOCATE, and CLOSE
performs an undocumented un-LOCATE (in OS/360 and its descendants, using >>> an undocumented un-PUT function in QSAM, QISAM, etc.).
By the way, please hold off on downloading or using this feature for a >>>> while, since a serious bug just surfaced that I will be working on anon. >>>Don’t mean nuttin’ to me unless you create a version for macOS (or at >>> least for Darwin).
--
Pete
br johann
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 374 |
| Nodes: | 16 (2 / 14) |
| Uptime: | 23:35:49 |
| Calls: | 7,968 |
| Calls today: | 2 |
| Files: | 13,014 |
| Messages: | 5,816,516 |