aptos-devel Mailing List for Aptos

Hey

   Just started building a new desktop computer after having problems
with the old one.  New is running Centos 7 x86.

   Just a note for anyone hitting the same pothole.  When installing
postp  I had to use the command "sudo python setup.py install" without
the "--prefix=/usr/local"  because with that I got a error when running
the command "apt test.apt" at the end of the compile.

  The --prefix=/usr/local appears to be what is used with ubuntu

  This install of apt360-1.0.3.0 went fine but get
gsl-1.15-13.el7.x86_64  &   gsl-devel-1.15-13.el7.x86_64 so the tabcyl
will work.


Thanks Brent
Jim

denis@denis-acerU:~/aptos/vapt-0.2.6/vapt$ sudo python2.7 vapt.py
No handlers could be found for logger "OpenGL.Tk"
Traceback (most recent call last):
  File "vapt.py", line 23, in <module>
    from aptview import APTView
  File "/home/denis/aptos/vapt-0.2.6/vapt/aptview.py", line 13, in <module>
    from bounds import BoundBox
  File "/home/denis/aptos/vapt-0.2.6/vapt/bounds.py", line 11, in <module>
    from matrix import Matrix3d
  File "/home/denis/aptos/vapt-0.2.6/vapt/matrix.py", line 9, in <module>
    from Numeric import pi, shape, identity, clip
ImportError: No module named Numeric
denis@denis-acerU:~/aptos/vapt-0.2.6/vapt$

Good Afternoon everybody.
I am trying to get VAPT working on an UBUNTU 18 box but seem to be having
issues.

Here is what I'm getting :

denis@denis-acerU:~/aptos/vapt-0.2.6/vapt$ sudo python vapt.py
[sudo] password for denis:
Traceback (most recent call last):
  File "vapt.py", line 15, in <module>
    from bwidget import NoteBook
  File "/home/denis/aptos/vapt-0.2.6/vapt/bwidget.py", line 53, in <module>
    import Tkinter, types, os, sys, new
  File "/usr/local/lib/python2.6/lib-tk/Tkinter.py", line 39, in <module>
    import _tkinter # If this fails your Python may not be configured for Tk
ImportError: No module named _tkinter
denis@denis-acerU:~/aptos/vapt-0.2.6/vapt$

I seem to have been here before but can't remember what we did to get over
it.

Any help is appreciated.
Regards
Denis

Hi Jim, et.al:
I pushed a change to the git repository for two new posts for postp. The 
first is called "JUSTPT" and outputs points instead of G2/G3 moves. The 
second is called "LIST" and just prints out a human-readable version of 
the cl file.

Let me know if "JUSTPT" is what you were looking for, Jim.  It might at 
least give you some ideas. If anyone has any questions, please ask.
Brent.

Brent

   I would like to have the apt360 output to list out G1  with no G2 or 
G3.  Is this possible with a post command?  Or is there a post processor 
to do that?

Thanks
Jim

Right, the postprocessor words (a/k/a 2000 class CLFILE records) are - for the 
most part - ignored by the APT processor itself, other than syntax errors and 
the MACHIN statement.

So any documentation for a keyword like INSERT will be documented in ones 
supplemental postprocessor documentation.

Attached is an example postprocessor documentation file someone provided me with 
a while back. Believe it's for a Fanuc controller.

But no guarantee this doc will match whatever post your using.

On 3/13/2016 3:57 PM, David wrote:
> Greetings Phil,
>
> Thanks for the response.
>
> The manual in the aptos is very helpful. However, look up the command "INSERT".
> It does not list the syntax to insert the commands I need.
>
> Cheers,
>
> Dave
>
>
>
> ------------------------------------------------------------------------------
> Transform Data into Opportunity.
> Accelerate data analysis in your applications with
> Intel Data Analytics Acceleration Library.
> Click to learn more.
> http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140
>
>
>
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel

APT-AC is no longer provided by IBM.  I think they dropped it several years ago and no longer support it. 

A company I am contracting at had it as part of their production process up through 2014.  I run a couple jobs every few months and they still work. 





> On Mar 13, 2016, at 2:25 PM, JEFFREY AUDIA <jef...@ms...> wrote:
> 
> That's right, it's "CONTUR" I made a typo.
> 
> I remember at Boeing there was a supplemental manual that described the additional feature-set including CONTUR and CIRARC, and many others. They had recently installed that extended package when I was there in 92 and I got to try them out including the God-sent CIRARC. I think that's what it was called. It allowed you to define only the "segment" of a circle or cylnder that you wanted and not have to deal with the annoying problem of the other part of the circle. I applied this feature and other ones from that package to new 777 parts machined on Boeing's giant spar-mills. Good times.
> 
> I never new there were 3 IBM-370 versions. 
> 
> And what does BobT mean by "I ran an APT-AC check two days ago..." Where is it available ? what does he mean ? IBM still sells it ?
> 
> jeff
> ------------------------------------------------------------------------------
> Transform Data into Opportunity.
> Accelerate data analysis in your applications with
> Intel Data Analytics Acceleration Library.
> Click to learn more.
> http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel

That's right, it's "CONTUR" I made a typo.

I remember at Boeing there was a supplemental manual that described the additional feature-set including CONTUR and CIRARC, and many others. They had recently installed that extended package when I was there in 92 and I got to try them out including the God-sent CIRARC. I think that's what it was called. It allowed you to define only the "segment" of a circle or cylnder that you wanted and not have to deal with the annoying problem of the other part of the circle. I applied this feature and other ones from that package to new 777 parts machined on Boeing's giant spar-mills. Good times.

I never new there were 3 IBM-370 versions. 

And what does BobT mean by "I ran an APT-AC check two days ago..." Where is it available ? what does he mean ? IBM still sells it ?

jeff
 		 	   		  

Greetings Phil,

Thanks for the response.

The manual in the aptos is very helpful. However, look up the command 
"INSERT". It does not list the syntax to insert the commands I need.

Cheers,

Dave

Under the APT360 project, there should be a 'doc' subfolder.

In 'doc' should be a subfolder named 'manual'.

In the 'manual' subfolder you should find a bunch of files, one named 'index.html'.

If you can open this 'index.html' file with a browser, you should see 'APT360 
Programmers Manual', your wish granted ??? :-)

Phil(f) in WI

On 3/13/2016 11:53 AM, David wrote:
> Greetings All,
>
> After last using APT 20 years ago, I am surprised at how much I have remembered.
> However, the book I purchased, Numerical Control Programming in APT, does not
> list the syntax for all the APT commands.
>
> I do remember having a small grey APT language guide when I was at Boeing.
> Sadly, I cannot locate this booklet at present.
>
> So, to the heart of my message.
>
> Does anyone have a spare booklet they would part with? Or, is there an on-line
> reference guide to all the APT commands?
>
> With kind regards,
>
> David
>
> South Prairie, WA 98385
> Formerly a member of A-3620, Tooling NC Programming
>
>
>
> ------------------------------------------------------------------------------
> Transform Data into Opportunity.
> Accelerate data analysis in your applications with
> Intel Data Analytics Acceleration Library.
> Click to learn more.
> http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140
>
>
>
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel

Greetings All,

After last using APT 20 years ago, I am surprised at how much I have 
remembered. However, the book I purchased, Numerical Control Programming 
in APT, does not list the syntax for all the APT commands.

I do remember having a small grey APT language guide when I was at 
Boeing. Sadly, I cannot locate this booklet at present.

So, to the heart of my message.

Does anyone have a spare booklet they would part with? Or, is there an 
on-line reference guide to all the APT commands?

With kind regards,

David

South Prairie, WA 98385
Formerly a member of A-3620, Tooling NC Programming

Did IBM ever support / release APT-AC on AIX or any other operating systems 
besides the 370 mainframe ???

If it is / was available for AIX was it 100% the same functionality as the 370 
mainframe version ???

Phil(f) in WI

-------- Original Message --------
Subject: 	Re: [Aptos-devel] APT-AC ???
Date: 	Fri, 11 Mar 2016 07:23:43 -0800
From: 	GrtGmail <g.r...@gm...>
To: 	motokoji143 <mot...@ho...>



I checked the APT-AC manual and the word is CONTUR.

Btw,  there were three versions of IBM APT for the 370 mainframes.

APT-AC   Advanced contouring
APT-IC     Intermediate contouring
APT-BC    Basic contouring

I ran an APT-AC check two days ago and it is still available and working even though it was dropped from production status about two years ago and hasn't had any maintenance releases from our APT support group for over 10 years.


BobT


>  On Mar 10, 2016, at 11:38 AM, motokoji143<mot...@ho...>  wrote:
>
>  Kim,
>
>  If you do not already have a copy, you might try locating a copy of Irvin Kral's
>  book: "Numerical Control Programming in APT". The copy I have is copyright 1986,
>  ISBN 0-13-626599-5. Per the preface, all of the examples were run with IBM
>  System370/APT-AC, Version 1, Mod 2 and 3.
>
>  Checking the index, the COUTUR word (sp?) Jeff's post mentioned is not mentioned.
>
>  There are several copies - new and used - available via amazon.com.
>
>  fwiw, Phil(f) in WI
>
>>  On 3/5/2016 2:15 PM, Kim Kirwan wrote:
>>  OK, thanks, this is good info, an IBM APT product that was in common
>>  use at aerospace companies at the time. Does this relate in any way
>>  to the "other" APT lists in existence?
>>
>>  It was my understanding that the Apt4ssv3 list (on Assembla.com)
>>  was for rewriting the APT4 v3 Fortran source code, which was said to
>>  be available without copyright restrictions. And that the apt4ssx8
>>  list (on freelists.org) was for rewriting the APT4 v8 Fortran source
>>  code, which was said to be available, but was still covered by some
>>  copyright (maybe NASA?) but that someone was working on getting them
>>  to release it into the public domain.  However, there have been no
>>  substantive posts on either list for 1 year, and 3 years, respectively.
>>
>>  The delay (as I understood it) was to wait and see if the APT4 v8 code
>>  could be freed up, and if so, to start with that. But if v8 could not
>>  be freed up, then to start with APT4 v3.
>>
>>  What is the origin of APT360, would that be APT3? Any possibility of
>>  getting either of the two APT4 projects moving again? Or, the APT4
>>  improvements (if any?) merged with APT360? Or, something else?
>>
>>  Any information, ideas, suggestions, much appreciated.
>>
>>  Kim
>>
>>
>>>  On 03/05/2016 09:48 AM, motokoji143 wrote:
>>>  APT-AC was IBM's successor product to the APT360 product / code base.
>>>  The 'AC' stands for Advanced Contouring' (IBM product code: 5740-M53).
>>>
>>>  This was the common APT one used if one worked at an aerospace company such as
>>>  Boeing or General Dynamics (at least in 1980), though most of the aerospace
>>>  companies also had add-ins of their own.
>>>
>>>  As Brent noted it had additional geometric constructs. Also, it migrated TRACUT,
>>>  INDEX and COPY in Section 3 so the postprocessor no longer needed to provide
>>>  functionality (though the 2000 records are still in the CLFILE.) If I recall
>>>  right, it also provided the ability to REFSYS a toolpath, not just geometry.
>>>
>>>  Googling: "APT-AC" site:ibm.com
>>>
>>>  Will generate links to documents about various version specific software
>>>  withdrawals on various platforms IBM provided APT on.
>>>
>>>  One of the documents mentions support moving to remarketers.
>>>
>>>  Hope this helps.
>>>
>>>  Phil(f) in WI
>>>
>>>
>>>>  On 3/5/2016 12:59 AM, Brent Muller wrote:
>>>>  Hi David- I think Apt-AC was a successor to apt360 (apt360 is my working name,
>>>>  not IBM's). The reason I think this is that the code in the book is based on
>>>>  Apt-AC, and it is very similar to apt360, except that it has extra definitions
>>>>  and features. For example, Apt-AC has a POLYCONIC definition. In the code for
>>>>  apt360, interestingly, there is some code for the POLYCONIC, but it is coded to
>>>>  give an error for that definition.
>>>>
>>>>  I tried googling "Apt-AC IBM". I found one document is rather interesting-It
>>>>  isn't necessarily related to Apt-AC, but I found it enlightening, since it
>>>>  clears up some of the history of the software:
>>>>
>>>>  https://law.resource.org/pub/us/case/reporter/F2/643/643.F2d.1339.76-2680.html
>>>>
>>>>  It's a lawsuit brought by a smaller company developing an Apt program competing
>>>>  with IBM's. If you can skip over the legalese, there is some interesting history
>>>>  there. I'm going to cc the mailing list, for the history buffs :)
>>>>
>>>>
>>>>  On Fri, Mar 4, 2016 at 5:12 AM, David<bal...@co...
>>>>  <mailto:bal...@co...>>   wrote:
>>>>
>>>>      Greetings Brent,
>>>>
>>>>      Many thanks for the instruction sheet.
>>>>
>>>>      APT360 is probably what I used at Boeing.
>>>>
>>>>      What is APT-AC? I could not find any information via Google.
>>>>
>>>>      Cheers,
>>>>
>>>>      Dave
>>>>
>>>>
>>>>
>>>>
>>>>  ------------------------------------------------------------------------------
>>>>
>>>>
>>>>
>>>>  _______________________________________________
>>>>  Aptos-devel mailing list
>>>>  Apt...@li...
>>>>  https://lists.sourceforge.net/lists/listinfo/aptos-devel
>>>  ------------------------------------------------------------------------------
>>>  _______________________________________________
>>>  Aptos-devel mailing list
>>>  Apt...@li...
>>>  https://lists.sourceforge.net/lists/listinfo/aptos-devel
>>
>>  ------------------------------------------------------------------------------
>>  _______________________________________________
>>  Aptos-devel mailing list
>>  Apt...@li...
>>  https://lists.sourceforge.net/lists/listinfo/aptos-devel
>
>
>  ------------------------------------------------------------------------------
>  Transform Data into Opportunity.
>  Accelerate data analysis in your applications with
>  Intel Data Analytics Acceleration Library.
>  Click to learn more.
>  http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140
>  _______________________________________________
>  Aptos-devel mailing list
>  Apt...@li...
>  https://lists.sourceforge.net/lists/listinfo/aptos-devel

Kim,

If you do not already have a copy, you might try locating a copy of Irvin Kral's 
book: "Numerical Control Programming in APT". The copy I have is copyright 1986, 
ISBN 0-13-626599-5. Per the preface, all of the examples were run with IBM 
System370/APT-AC, Version 1, Mod 2 and 3.

Checking the index, the COUTUR word (sp?) Jeff's post mentioned is not mentioned.

There are several copies - new and used - available via amazon.com.

fwiw, Phil(f) in WI

On 3/5/2016 2:15 PM, Kim Kirwan wrote:
> OK, thanks, this is good info, an IBM APT product that was in common
> use at aerospace companies at the time. Does this relate in any way
> to the "other" APT lists in existence?
>
> It was my understanding that the Apt4ssv3 list (on Assembla.com)
> was for rewriting the APT4 v3 Fortran source code, which was said to
> be available without copyright restrictions. And that the apt4ssx8
> list (on freelists.org) was for rewriting the APT4 v8 Fortran source
> code, which was said to be available, but was still covered by some
> copyright (maybe NASA?) but that someone was working on getting them
> to release it into the public domain.  However, there have been no
> substantive posts on either list for 1 year, and 3 years, respectively.
>
> The delay (as I understood it) was to wait and see if the APT4 v8 code
> could be freed up, and if so, to start with that. But if v8 could not
> be freed up, then to start with APT4 v3.
>
> What is the origin of APT360, would that be APT3? Any possibility of
> getting either of the two APT4 projects moving again? Or, the APT4
> improvements (if any?) merged with APT360? Or, something else?
>
> Any information, ideas, suggestions, much appreciated.
>
> Kim
>
>
> On 03/05/2016 09:48 AM, motokoji143 wrote:
>> APT-AC was IBM's successor product to the APT360 product / code base.
>> The 'AC' stands for Advanced Contouring' (IBM product code: 5740-M53).
>>
>> This was the common APT one used if one worked at an aerospace company such as
>> Boeing or General Dynamics (at least in 1980), though most of the aerospace
>> companies also had add-ins of their own.
>>
>> As Brent noted it had additional geometric constructs. Also, it migrated TRACUT,
>> INDEX and COPY in Section 3 so the postprocessor no longer needed to provide
>> functionality (though the 2000 records are still in the CLFILE.) If I recall
>> right, it also provided the ability to REFSYS a toolpath, not just geometry.
>>
>> Googling: "APT-AC" site:ibm.com
>>
>> Will generate links to documents about various version specific software
>> withdrawals on various platforms IBM provided APT on.
>>
>> One of the documents mentions support moving to remarketers.
>>
>> Hope this helps.
>>
>> Phil(f) in WI
>>
>>
>> On 3/5/2016 12:59 AM, Brent Muller wrote:
>>> Hi David- I think Apt-AC was a successor to apt360 (apt360 is my working name,
>>> not IBM's). The reason I think this is that the code in the book is based on
>>> Apt-AC, and it is very similar to apt360, except that it has extra definitions
>>> and features. For example, Apt-AC has a POLYCONIC definition. In the code for
>>> apt360, interestingly, there is some code for the POLYCONIC, but it is coded to
>>> give an error for that definition.
>>>
>>> I tried googling "Apt-AC IBM". I found one document is rather interesting-It
>>> isn't necessarily related to Apt-AC, but I found it enlightening, since it
>>> clears up some of the history of the software:
>>>
>>> https://law.resource.org/pub/us/case/reporter/F2/643/643.F2d.1339.76-2680.html
>>>
>>> It's a lawsuit brought by a smaller company developing an Apt program competing
>>> with IBM's. If you can skip over the legalese, there is some interesting history
>>> there. I'm going to cc the mailing list, for the history buffs :)
>>>
>>>
>>> On Fri, Mar 4, 2016 at 5:12 AM, David<bal...@co...
>>> <mailto:bal...@co...>>  wrote:
>>>
>>>      Greetings Brent,
>>>
>>>      Many thanks for the instruction sheet.
>>>
>>>      APT360 is probably what I used at Boeing.
>>>
>>>      What is APT-AC? I could not find any information via Google.
>>>
>>>      Cheers,
>>>
>>>      Dave
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>>
>>>
>>>
>>> _______________________________________________
>>> Aptos-devel mailing list
>>> Apt...@li...
>>> https://lists.sourceforge.net/lists/listinfo/aptos-devel
>> ------------------------------------------------------------------------------
>> _______________________________________________
>> Aptos-devel mailing list
>> Apt...@li...
>> https://lists.sourceforge.net/lists/listinfo/aptos-devel
>>
>>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel
>
>

On 03/05/2016 01:15 PM, Kim Kirwan wrote:
> OK, thanks, this is good info, an IBM APT product that was in common
> use at aerospace companies at the time. Does this relate in any way
> to the "other" APT lists in existence?
>
I guess the answer would be "no", if you mean the other projects that 
are old APT code bases that we hobbyists have access to. It looks, from 
what Jeff Audia mentioned, that IBM took the APT-AC code into the price 
stratosphere, and it is probably heavily copyrighted.

> It was my understanding that the Apt4ssv3 list (on Assembla.com)
> was for rewriting the APT4 v3 Fortran source code, which was said to
> be available without copyright restrictions. And that the apt4ssx8
> list (on freelists.org) was for rewriting the APT4 v8 Fortran source
> code, which was said to be available, but was still covered by some
> copyright (maybe NASA?) but that someone was working on getting them
> to release it into the public domain.  However, there have been no
> substantive posts on either list for 1 year, and 3 years, respectively.
>

Those projects were worked on, but I think the implicit consensus was 
that no one felt like dedicating their hobby time to working on it (I'm 
guessing). The Apt4ssv3 code that I was given was started on a couple of 
times- The first time, we hit a snag with the syntax code generating 
file that was undocumented. It was restarted with Andy Pruitt from 
scratch, and I got it running about 20% of the test files included in 
the code base to completion. The reason I stopped was that the way 
forward from there didn't look real promising. The files ran though 
syntax-wise, but I couldn't figure why or where it was failing without 
understanding the code way more than a simple perusal. For example,if 
the code fails with a segfault or something, you can spot the problem 
and work backwards. That wasn't happening-the program ran to completion 
and basically it just gave an error message due to some problem 
somewhere. Additionally, the clfile was binary, and would have to be 
documented to write posts for.

If someone would like to look at it, I can give them a copy of the code 
to look over. I don't think Andy has been working on it either.

> The delay (as I understood it) was to wait and see if the APT4 v8 code
> could be freed up, and if so, to start with that. But if v8 could not
> be freed up, then to start with APT4 v3.
>
> What is the origin of APT360, would that be APT3? Any possibility of
> getting either of the two APT4 projects moving again? Or, the APT4
> improvements (if any?) merged with APT360? Or, something else?
>
> Any information, ideas, suggestions, much appreciated.
>
> Kim
>
>
<old stuff snipped>
Here is the relevant parts of the legal brief giving a short history of 
the software:

> 3
>
> * Both IBM and Symbolic produce a type of computer software known as an
> automatically programmed tools (APT) processor. This software permits
> the use of a computer in the production of metal parts in aviation,
> automobile manufacturing, and similar manufacturing enterprises. When an
> engineering diagram of a part to be produced is completed manually, the
> diagram is given to a computer programmer who translates the diagram
> into computer language, producing the "part program." At this point the
> APT processor comes into play: the APT processor, itself a computer
> program, is used by a computer to calculate the location, speed, and
> movement the machine tool must use to cut the metal part required.
>
> 4
>
> The first APT processor was developed by the Massachusetts Institute of
> Technology under contract to the United States Air Force. A later
> version of this program, called APT III, was placed in the public
> domain. APT III was the basis of later refined versions of the APT
> processor developed by IBM, General Electric Corporation, Sperry Rand
> Corporation, and Control Data Corporation, for use with the various
> computers produced by each of these firms.
>
> 5
>
> Between 1967 and 1970 IBM developed and distributed four versions of its
> APT processor, all designed for use with its System/360 computer. These
> software programs were identified as NC 360, followed by a version
> number and a modification level. IBM furnished all versions of the NC
> 360 program free of charge, and provided free maintenance and
> modification levels. Maintenance consists of correcting errors in the
> program, and a modification level is a compilation and incorporation of
> error corrections made after the release of the prior modification level.
>



Brent.

Some more info about Apt-AC from Jeff- thanks, Jeff!


---------- Forwarded message ----------
From: JEFFREY AUDIA <jef...@ms...>
Date: Sat, Mar 5, 2016 at 9:36 AM
Subject: APT-AC comments
To: "bre...@gm..." <bre...@gm...>


How silly of my to have used Linkedin, it took me a second to realize your
email. It's been a while, how are you doing, Brent ?

I have been quiet for some time regarding Aptos for various reasons. below
are some comments I had about APT-AC since the topic came up.

I originally sent them via Linked in:
jeff
Hi Brent, I don't know how to post comments on Aptos nor do I have your
email address any more.
Regarding Apt-AC. This was actually IBM 370 APT-AC. It ran on the IBM 370.My
understanding was that It was a proprietary version of APT and not public
domain and thus IBM developed it way beyond what 360 was. An implementation
cost $100K or more.
AC means "Advanced Contouring" and the system contained many powerful
features including the concept of a "COUTUR" a contour which is a
continuous and possibly closed sequence of 2D geometric entities and
allowed for tool paths over it with a single command much like what people
are accustomed to in graphics. It had many other entities such as open
cylinders and circles that got rid of the pesky "other part" of the circle
that was not part of your geometry. AC also had many advanced multi-axis
tool control features far beyond what 360 could do. It is truly very
powerful and the apex of APT development. Too many features to list here.
I used IBM 370 APT-AC in many places during my career including Boeing
where I programmed the giants.
BOEING has an even more specialized version of AC with features peculiar to
their work. And they still use it.
I was involved in implementing AC-like features in a graphical APT CAM
product called ELCAM (migrating eventually to ADRA) which claimed 370 AC
compatibility. In the late 80's I served as an expert on the language and
APT in general and helped test and sell the product to customers.I should
still have the IBM 370 APT-AC manual as well as the APT encyclopedia
written by Charles Drayton who originally taught me APT and who died only
recently. Drayton did pioneering work in the early days on the APT
development committee and also taught APT language at Northern IL
University in Dekalb, IL where he was famous for having a rubber ear tacked
to the wall in his office; when you had an APT problem you were required to
address the ear first; and inevitably students would solve it before they
finished stating the question and not bother professor Drayton.
So IBM 370 APT-AC does exist although I doubt IBM has anything to do with
it any more. I suspect Boeing bought the rights to it long ago when sales
plummeted shortly after the Reagan-era War Machine ground to a halt and no
one needed that kind of multi-axis horsepower any more. (and I switched
careers !). Boeing still uses the system and they also use it in an
incredibly clever way to process the output of their Catia system into
machine code. Feel free to post these comments jeff

OK, thanks, this is good info, an IBM APT product that was in common 
use at aerospace companies at the time. Does this relate in any way 
to the "other" APT lists in existence? 

It was my understanding that the Apt4ssv3 list (on Assembla.com)  
was for rewriting the APT4 v3 Fortran source code, which was said to 
be available without copyright restrictions. And that the apt4ssx8 
list (on freelists.org) was for rewriting the APT4 v8 Fortran source 
code, which was said to be available, but was still covered by some 
copyright (maybe NASA?) but that someone was working on getting them 
to release it into the public domain.  However, there have been no 
substantive posts on either list for 1 year, and 3 years, respectively.

The delay (as I understood it) was to wait and see if the APT4 v8 code 
could be freed up, and if so, to start with that. But if v8 could not 
be freed up, then to start with APT4 v3. 

What is the origin of APT360, would that be APT3? Any possibility of 
getting either of the two APT4 projects moving again? Or, the APT4 
improvements (if any?) merged with APT360? Or, something else? 

Any information, ideas, suggestions, much appreciated. 

Kim


On 03/05/2016 09:48 AM, motokoji143 wrote:
> APT-AC was IBM's successor product to the APT360 product / code base.
> The 'AC' stands for Advanced Contouring' (IBM product code: 5740-M53).
> 
> This was the common APT one used if one worked at an aerospace company such as 
> Boeing or General Dynamics (at least in 1980), though most of the aerospace 
> companies also had add-ins of their own.
> 
> As Brent noted it had additional geometric constructs. Also, it migrated TRACUT, 
> INDEX and COPY in Section 3 so the postprocessor no longer needed to provide 
> functionality (though the 2000 records are still in the CLFILE.) If I recall 
> right, it also provided the ability to REFSYS a toolpath, not just geometry.
> 
> Googling: "APT-AC" site:ibm.com
> 
> Will generate links to documents about various version specific software 
> withdrawals on various platforms IBM provided APT on.
> 
> One of the documents mentions support moving to remarketers.
> 
> Hope this helps.
> 
> Phil(f) in WI
> 
> 
> On 3/5/2016 12:59 AM, Brent Muller wrote:
>> Hi David- I think Apt-AC was a successor to apt360 (apt360 is my working name,
>> not IBM's). The reason I think this is that the code in the book is based on
>> Apt-AC, and it is very similar to apt360, except that it has extra definitions
>> and features. For example, Apt-AC has a POLYCONIC definition. In the code for
>> apt360, interestingly, there is some code for the POLYCONIC, but it is coded to
>> give an error for that definition.
>>
>> I tried googling "Apt-AC IBM". I found one document is rather interesting-It
>> isn't necessarily related to Apt-AC, but I found it enlightening, since it
>> clears up some of the history of the software:
>>
>> https://law.resource.org/pub/us/case/reporter/F2/643/643.F2d.1339.76-2680.html
>>
>> It's a lawsuit brought by a smaller company developing an Apt program competing
>> with IBM's. If you can skip over the legalese, there is some interesting history
>> there. I'm going to cc the mailing list, for the history buffs :)
>>
>>
>> On Fri, Mar 4, 2016 at 5:12 AM, David <bal...@co...
>> <mailto:bal...@co...>> wrote:
>>
>>     Greetings Brent,
>>
>>     Many thanks for the instruction sheet.
>>
>>     APT360 is probably what I used at Boeing.
>>
>>     What is APT-AC? I could not find any information via Google.
>>
>>     Cheers,
>>
>>     Dave
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>>
>>
>>
>> _______________________________________________
>> Aptos-devel mailing list
>> Apt...@li...
>> https://lists.sourceforge.net/lists/listinfo/aptos-devel
> 
> ------------------------------------------------------------------------------
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel
> 
> 

APT-AC was IBM's successor product to the APT360 product / code base.
The 'AC' stands for Advanced Contouring' (IBM product code: 5740-M53).

This was the common APT one used if one worked at an aerospace company such as 
Boeing or General Dynamics (at least in 1980), though most of the aerospace 
companies also had add-ins of their own.

As Brent noted it had additional geometric constructs. Also, it migrated TRACUT, 
INDEX and COPY in Section 3 so the postprocessor no longer needed to provide 
functionality (though the 2000 records are still in the CLFILE.) If I recall 
right, it also provided the ability to REFSYS a toolpath, not just geometry.

Googling: "APT-AC" site:ibm.com

Will generate links to documents about various version specific software 
withdrawals on various platforms IBM provided APT on.

One of the documents mentions support moving to remarketers.

Hope this helps.

Phil(f) in WI


On 3/5/2016 12:59 AM, Brent Muller wrote:
> Hi David- I think Apt-AC was a successor to apt360 (apt360 is my working name,
> not IBM's). The reason I think this is that the code in the book is based on
> Apt-AC, and it is very similar to apt360, except that it has extra definitions
> and features. For example, Apt-AC has a POLYCONIC definition. In the code for
> apt360, interestingly, there is some code for the POLYCONIC, but it is coded to
> give an error for that definition.
>
> I tried googling "Apt-AC IBM". I found one document is rather interesting-It
> isn't necessarily related to Apt-AC, but I found it enlightening, since it
> clears up some of the history of the software:
>
> https://law.resource.org/pub/us/case/reporter/F2/643/643.F2d.1339.76-2680.html
>
> It's a lawsuit brought by a smaller company developing an Apt program competing
> with IBM's. If you can skip over the legalese, there is some interesting history
> there. I'm going to cc the mailing list, for the history buffs :)
>
>
> On Fri, Mar 4, 2016 at 5:12 AM, David <bal...@co...
> <mailto:bal...@co...>> wrote:
>
>     Greetings Brent,
>
>     Many thanks for the instruction sheet.
>
>     APT360 is probably what I used at Boeing.
>
>     What is APT-AC? I could not find any information via Google.
>
>     Cheers,
>
>     Dave
>
>
>
>
> ------------------------------------------------------------------------------
>
>
>
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel

Hi David- I think Apt-AC was a successor to apt360 (apt360 is my working
name, not IBM's). The reason I think this is that the code in the book is
based on Apt-AC, and it is very similar to apt360, except that it has extra
definitions and features. For example, Apt-AC has a POLYCONIC definition.
In the code for apt360, interestingly, there is some code for the
POLYCONIC, but it is coded to give an error for that definition.

I tried googling "Apt-AC IBM". I found one document is rather
interesting-It isn't necessarily related to Apt-AC, but I found it
enlightening, since it clears up some of the history of the software:

https://law.resource.org/pub/us/case/reporter/F2/643/643.F2d.1339.76-2680.html

It's a lawsuit brought by a smaller company developing an Apt program
competing with IBM's. If you can skip over the legalese, there is some
interesting history there. I'm going to cc the mailing list, for the
history buffs :)


On Fri, Mar 4, 2016 at 5:12 AM, David <bal...@co...> wrote:

> Greetings Brent,
>
> Many thanks for the instruction sheet.
>
> APT360 is probably what I used at Boeing.
>
> What is APT-AC? I could not find any information via Google.
>
> Cheers,
>
> Dave
>

Hi Dave, and all, 

Normally I am not able to contribute much, but "What distribution 
do you use?" Is a question I can answer, and if it helps to get 
vapt (or something similar) back in shape, then great. 

As Brent mentioned on 02/22/2016 (I'm paraphrasing here), there is 
an advantage to staying pretty close to the distributions favored 
by the LinuxCNC developers. 

In the past, that was the Long-Term-Support (LTS) versions of Ubuntu, 
great from 6.06 (2006, June), to 8.04 (2008, April), to 10.04 (2010, 
April). When 12.04 came out, Ubuntu (and Gnome) had introduced some 
shenanigans that made users unhappy, including the Unity desktop, 
an Amazon icon, Ubuntu One, Gnome problems, etc., so Ubuntu users 
everywhere (not just LinuxCNC users) began to look around for an 
Ubuntu replacement. Distrowatch at that time showed Ubuntu taking 
a nose dive and Linux Mint coming up fast, so I tried it (for my 
multi-purpose desktops and laptops) and liked it a lot. I've been 
using Mint ever since. 

The LinuxCNC developers (as I recall, anyway) did not release any 
new ISOs in the 12.04 timeframe, but by the time 14.04 came around, 
they had moved to Debian (with XFCE desktop instead of Gnome). And 
that's fine for a CNC machine OS because you have Debian's commitment 
to free/open software in a "minimalist" install, while eliminating 
the problems related to Gnome. So I don't mind using the LinuxCNC 
all-in-one ISO on CNC-machine-only PCs. 

Debian is the parent of Ubuntu, and Ubuntu is the parent of Mint. 
They all use the .deb, apt-get, Synaptic, Aptitude packaging system. 
The problem with the Debian version is that Debian doesn't support 
Ubuntu PPA's (at least not that I know of), so it's a great 
inconvenience not to be able to subscribe to the "daily" builds 
for FreeCad, Kicad, etc. But PPAs work fine on Mint. 

Ubuntu came up with a "Gnome-classic-fallback" mode for 12.04 (and 
14.04) to get rid of Unity and make the start menu look more like 
it did in 10.04, which helped, but may have been too little too late 
for Ubuntu. 

Mint comes in several versions, depending on the desktop. Mint took 
Gnome 3 (ala 12.04) and forked it to become "Cinnamon", and took 
Gnome 2 (ala 10.04) and forked it to become "MATE" (MAH-tay, named 
after the South American tea). Right now it is relatively difficult 
to change from one desktop to another without re-installing, but 
after Ubuntu 16.04 comes out, Mint says that Mint 18 will allow 
users to change from one desktop to another with a couple of 
meta-package clicks in Synaptic. Nice! 

In the interests of full disclosure, Mint does have one 
shenanigan, it comes with mint-search-enhancer preinstalled, 
but it's easy to go into Synaptic and remove it. They say they 
get some funding from that. I wish they wouldn't do it, but, 
it's easy enough to get rid of. 

I'm still running 10.04 on my main desktop, and I need to update. 
So I probably will when Mint 18 (16.04) comes out. I think the 
regular LinuxCNC install script works on Mint, or if not, then 
someone has a LinuxCNC install script for Mint, because I recall 
trying to script-install LinuxCNC as an experiment and it worked. 

Sorry this got kind of long. To summarize: 
For CNC machines I use the LinuxCNC Debian 32 XFCE all-in-one ISO. 
(I have heard rumors that a 64-bit version will be added next time.)
For Desktops/Laptops I use Linux Mint 64 Cinnamon 17.3 (soon 18). 

Hope this helps. 

Kim


On 02/27/2016 07:52 PM, David wrote:
> Greetings All,
> 
> This trip down memory lane to APT360-ville has been most illuminating. I
> am so eager that I purchased an APT training book on Amazon.
> 
> The trouble I run into is compiling the apt360 software. So far, I have
> had no luck with SUSE 12.2. Today, I spent 4 hours downloading and
> installing every developmental library/application the software had to
> offer. I still get the error indicating the math library is not found.
> 
> I installed UBUNTU 14.04.3 on a spare machine and got apt360 and ppost
> to run. Sadly, vapt is written with libraries that UBUNTU 14.04.3 no
> longer offers.
> 
> So, before I flatten another machine, I would like to hear from other
> users the Linux distribution they use.
> 
> UBUNTU is okay; but, I did not see any opportunity to select the
> packages prior to loading. SUSE does offer me a list.
> 
> With kind regards,
> 
> Dave
> 
> South Prairie, WA 98385
> 
> 
> 
> ------------------------------------------------------------------------------
> Site24x7 APM Insight: Get Deep Visibility into Application Performance
> APM + Mobile APM + RUM: Monitor 3 App instances at just $35/Month
> Monitor end-to-end web transactions and take corrective actions now
> Troubleshoot faster and improve end-user experience. Signup Now!
> http://pubads.g.doubleclick.net/gampad/clk?id=272487151&iu=/4140
> 
> 
> 
> _______________________________________________
> Aptos-devel mailing list
> Apt...@li...
> https://lists.sourceforge.net/lists/listinfo/aptos-devel
> 

Greetings All,

This trip down memory lane to APT360-ville has been most illuminating. I 
am so eager that I purchased an APT training book on Amazon.

The trouble I run into is compiling the apt360 software. So far, I have 
had no luck with SUSE 12.2. Today, I spent 4 hours downloading and 
installing every developmental library/application the software had to 
offer. I still get the error indicating the math library is not found.

I installed UBUNTU 14.04.3 on a spare machine and got apt360 and ppost 
to run. Sadly, vapt is written with libraries that UBUNTU 14.04.3 no 
longer offers.

So, before I flatten another machine, I would like to hear from other 
users the Linux distribution they use.

UBUNTU is okay; but, I did not see any opportunity to select the 
packages prior to loading. SUSE does offer me a list.

With kind regards,

Dave

South Prairie, WA 98385

Hi Dave
The Suse software should work (maybe it's really old, which is maybe why
you are updating, I don't know). The big advantage with Ubuntu is that the
LinuxCNC people use it as their standard system, so there is that. Were you
able to find development packages for Suse?

The windows version is up on the Sourceforge site now. I will cc this to
the list for Phil Fuhlman (jsuber) to see. He has been doing the Windows
stuff. The file is an exec and some ancillary files, so it should just
work, and I think any Windows should be no problem. I could also compile a
static version and send you that, but you might have to do some manual
fiddling to get it to work.

Keep at it, and let me know if more problems come up.
Brent.

On Mon, Feb 22, 2016 at 3:38 PM, David <bal...@co...> wrote:

> Greetings Brent,
>
> Still struggling with the compiler.
>
> I downloaded the Ubuntu software; but, I need to buy some DVD's to make
> the boot disk.
>
> Meantime, you say the program has been compiled in Windoze? Would that the
> the obsolete XP version? If so, do I need to be a magician to do this in
> the comfort of my own home?
>
> Cheers,
>
> Dave Wilson
>

You might see if it is a permissions issue. Maybe try to compile as root.
If that works, you can do a make install and be done. It won't solve the
bigger issue of why it isn't compiling, but at least it will get you going.

If it ended up being a permissions issue, and you feel you might want to
compile something in the future, I would look at the /etc/ld.so.conf.d
directory to see if the comments in the ld.so.conf file are fibbing. If so,
maybe you are not a member of a group permission that has ld access or
something. I don't know a great deal about Linux admin, and nothing about
Suse- and they all set up things slightly differently, so I'm just guessing.

I would be interested in knowing about you and your Apt experiences, and am
sure others on the list would too. Let me know your results.
Brent.

On Fri, Feb 19, 2016 at 8:40 AM, David <bal...@co...> wrote:

> Greetings Brent,
>
> Many thanks for attending to my plea for help. I look forward to
> programming in APT360 like the "old days".
>
> Here is a clip of my findings...
>
> david@linux-bkje:~> find / -name "lib[cm]\.[aso]*" 2>/dev/null
> /lib/libc.so.6
> /lib/i686/nosegneg/libc.so.6
> /lib/i686/nosegneg/libm.so.6
> /lib/libm.so.6
> /usr/lib/libm.so
> /usr/lib/libc.so
> david@linux-bkje:~> uname -m
> i686
> david@linux-bkje:~>
>
> david@linux-bkje:~> cat /etc/ld.so.conf
> /usr/local/lib
> include /etc/ld.so.conf.d/*.conf
> # /lib64, /lib, /usr/lib64 and /usr/lib gets added
> # automatically by ldconfig after parsing this file.
> # So, they do not need to be listed.
>
> I suspect I have a 32 bit machine from the i686?
>
> Many thanks for porting this software to the Linux community.
>
> With kind regards,
>
> David
>

Hi David-
The compile is failing during the final step, linking, where all the
individual object files are connected with the libraries on the computer.
The important error messages are the  "cannot find -lm" and "cannot find
-lc" lines. These are the math and C libraries, which should be on every
Linux machine.

There are a couple things that come to mind that might be causing the
problem:

1: You have a 64 bit machine, and it is trying to find 32bit libraries
2. The libraries are in a non-standard place on SuSe

If you are on a 64 bit machine, the current version won't work anyway, but
fear not: I have a 64 bit release ready, I just have to post it to
Sourceforge. You may know already, but ff you run the command:

uname -m

you should find out whether you have a 64bit or 32bit machine. On my 32 bit
machine, I get "i686" and on my 64bit machine I get "x86_64".

To find out where the libraries are, I can run the following command:

find / -name "lib[cm]\.[aso]*" 2>/dev/null

This should tell you where your libraries are, and you can adjust the
configure options (see "./configure --help") to tell it where to find them.
Also look in /etc/ld.so.conf for the places that the gcc compiler is
looking for system libraries.

I don't have time tonight to do anything else, but I should have the 64 bit
version up by this weekend. Let me know what you find out, and we will take
it from there.

Brent-



On Thu, Feb 18, 2016 at 4:28 PM, Brent Muller <bm...@si...> wrote:

>
>
>
> -------- Original Message --------
> Subject:        [Aptos-devel] Compiling Help
> Date:   Thu, 18 Feb 2016 10:02:35 -0800
> From:   David <bal...@co...>
> To:     apt...@li...
>
>
>
> Greetings,
>
> Having a spot of bother compiling the APT program.
>
> Using SUSE 12.2
>
> With the apt360-1.0.2 source, ./configure seems to have run okay. During
> the "make" portion of the process, I get the following message at the
> end of the file...
>
>
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld:
> Warning: size of symbol `a1stab_' changed from 184 in ACLAS5.o to 192 in
> BLOCK5.o
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld:
> Warning: alignment 4 of symbol `adfprl_' in ADFPRL.ASM.o is smaller than
> 32 in ADFPRE.o
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld:
> cannot find -lm
> /usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld:
> cannot find -lc
> collect2: error: ld returned 1 exit status
> make[2]: *** [apt360] Error 1
> make[2]: Leaving directory `/home/david/aptos/apt360-1.0.2/src'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/home/david/aptos/apt360-1.0.2'
> make: *** [all] Error 2
>
>
> I am fairly good at following instructions; however, I am not a true
> programmer when it comes to source code.
>
> Anyone offer some hints as to what is missing?
>
> Cheers,
>
> David
>
> South Prairie, WA 98385
> USA
>
>
>

Greetings,

Having a spot of bother compiling the APT program.

Using SUSE 12.2

With the apt360-1.0.2 source, ./configure seems to have run okay. During 
the "make" portion of the process, I get the following message at the 
end of the file...


/usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld: 
Warning: size of symbol `a1stab_' changed from 184 in ACLAS5.o to 192 in 
BLOCK5.o
/usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld: 
Warning: alignment 4 of symbol `adfprl_' in ADFPRL.ASM.o is smaller than 
32 in ADFPRE.o
/usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld: 
cannot find -lm
/usr/lib/gcc/i586-suse-linux/4.7/../../../../i586-suse-linux/bin/ld: 
cannot find -lc
collect2: error: ld returned 1 exit status
make[2]: *** [apt360] Error 1
make[2]: Leaving directory `/home/david/aptos/apt360-1.0.2/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/home/david/aptos/apt360-1.0.2'
make: *** [all] Error 2


I am fairly good at following instructions; however, I am not a true 
programmer when it comes to source code.

Anyone offer some hints as to what is missing?

Cheers,

David

South Prairie, WA 98385
USA

Here is a discussion that Phil emailed me about that I thought the list
might be interested in.

Also, Jim brought up the intriguing possibility of using the READ/2
statement to read in dxf files, among other things.

Any comments would be appreciated.
Brent.
---------- Forwarded message ----------
From: Rusk County Manufacturing LLC <rc...@br...>
Date: Thu, Nov 19, 2015 at 5:32 PM
Subject: Re: APT360 XML Clfile ???
To: Brent Muller <bre...@gm...>


Fine with me to forward it to the list.

I'm ok with using other languages besides C for the external tools like the
posts.

I should try some Python, maybe writing a app that demonstrates how to
access / use an XML clfile would be a good project to start with.

I like keeping the posts 'out of the kernel' but being able to invoke them
with a late binding scheme could be an improvement versus launching a
separate process, at least on Windows.

I seem to recall running into a few instances where launching / tracking
the postprocessor process instance got out of sync but that could be me /
Windows. Implementing a late binding scheme in Windows is simple but am
clueless about Linux, it looks something like this (less error checking):

        // Load user's post dll. Convention would be 6 char max name from
MACHIN
        // stmt plus '.dll' file extension.

        HDLL hdll = LoadLibrary("postp.dll");

        // Get function pointer to user's 'standard' APT360 entry point
        // we'd have to agree on a 'standard' function name for the
        // postprocessor entry point.

        (function pointer type)* pfn = GetProcInstance(hdll,"PostPMain");

        pfn(arglist); // launch user's postprocessor code

        FreeLibrary(hdll);

As I recall, the Windows security officers dislike user processes launching
other processes ;-)

A bit of support code that would act as the EXE to launch the user's DLL
for post developers without having to have a full APT development setup is
also just a few lines of code in Windows.


On 11/19/2015 4:12 PM, Brent Muller wrote:

> I think it is a great idea. There wouldn't be any downsides to lowering the
> entry bar for newcomers. The author of vapt actually parsed the text std
> out to
> get geometry data for his program. I don't know if this was before or
> after I
> created the binary "print.tap" file for geometry.Regardless, the binary
> file
> wasn't documented, and required a person to sort through hundreds of lines
> of code.
>
> Speaking of print.tap, it might be a good idea to have a text print file
> also.
>
> Although it isn't a requirement, it would be nice to keep the dependencies
> low.
> Right now, the code compiles easily for most people. I don't know what
> libraries
> you plan on using, but that is also a barrier to entry for newcomers, to
> compile
> and install libraries. But if you have code already, don't worry about it.
>
> I have been thinking about moving the postp code into the main package.
> What
> would you think of that? Again, it would make it easier for people to get
> started.
>
> Have you programmed in Python before? I made the postp code in python
> because
> most of the LinuxCNC people are pretty familiar with Python, and I
> expected them
> to want to tweak the posts. I only mention this because it fits well with
> the
> theme of barrier to entry that you rightly point out, and something that is
> increasingly on my mind. I also think that Python dependencies would be
> easier
> for newcomers to install than C libraries.I know you are a good C
> programmer,
> and Python would be a piece of cake for you.pick up.
>
> Above all, I am only making suggestions. Any kind of help is greatly
> appreciated, whether it be in C or anything else, and you are welcome to do
> whatever you feel comfortable with. For me to put any kind of restrictions
> on
> helpers would only be detrimental, so keep it fun for yourself.
>
> Would you have any issues with me forwarding this to the mailing list? I
> keep
> thinking I should keep other people in the loop.
> Brent.
>
> On Tue, Nov 17, 2015 at 1:04 PM, Rusk County Manufacturing LLC
> <rc...@br... <mailto:rc...@br...>> wrote:
>
>     Hi Brent,
>
>     Traded some emails w/ Bob Eggleton about APT360.
>
>     He mentioned others would like to generate postprocessors but find
> working
>     with the native clfile format to be a barrier so I thought an XML
> format
>     might solve the problem.
>
>     Thought is to add the code to APT360 and execute it just before
> invoking the
>     user's postprocessor so as to generate a second clfile, albeit in XML
> format.
>
>     So attached an example in a ZIP file. I omitted the 1000 records
> entirely on
>     purpose, otherwise the barebones code only does a basic job of the
> 2000,
>     3000 and 5000 records, others just emit major and minor word values
> for now.
>
>     I'm emitting floating point values as decimal values here but for real
> I'd
>     prefer to emit them in hexidecimal format to preserve the full
> fidelity, but
>     would some have problems getting them to converted back ??? Easy
> enough to
>     do in 'C' but maybe not so much with something like QBasic ???
>
>     Anyway, am hoping you can find time to review and ponder and let me
> know
>     your thoughts. Thanks!
>
>     Phil
>
>
>