|
SCOUG-Programming Mailing List ArchivesReturn to [ 15 | May | 2008 ]
157 This is a multi-part message in MIME format. ------=_NextPart_000_00B8_01C8B6A4.6AE78A30 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Having just upgraded my FileStar to the latest version, I don't have to = attend a meeting in which the speaker may or may not show up. For = several reasons I will not attend this month's meeting. I have not had = the time to devote to preparing material and presentation for the = Programming SIG. I have resuscitated my "old" IBM ThinkPad 760XD to use = in this endeavor. It charged up to 100%. Everything seems to be = working. I've ordered a wireless connector to ease file transfers and = downloads. As it already has APL and PL/I loaded, I will add the Forth package as = well as an OS/2 version of LISP. I also have the DOS-based Trilogy. = Perhaps I will add Prolog. That says while we will focus on a detailed = understanding of Forth as a TIL and certainly gain a working knowledge = of assembly language, we can explore some other subjects along the way. I did want to address again Greg's demand about BNF in SL/I. Whether we = call it the Backus-Normal or Backus-Naur Form Backus, the "father" of = Fortran, had to have a means of defining the language syntax for which = Fortran seemed ill-suited. Defining syntax has taken on a number of = forms or variations since then from that of LEX to that of the syntax = diagrams IBM uses in its LRMs. While the LEX statements may not look like C, it uses a C language = program to produce C output. In fact all parsers basically derive from = the same process. If we can write a parser, that which confirms the = syntax of a source language, in the same source language, then its = capability includes self-defining. If we can do it in C or PL/I, which = we can, then certainly we can do it in SL/I. The tricky part for any language syntax lies in its support of = recursion, where the thing we define we use in the definition. That = comes straight out of manufacturing bills-of-material (BOM) where we = have two basic entities, raw material and assemblies. While we cannot = decompose a raw material, we define an assembly as something composed of = raw materials or other assemblies. In this manner we define an assembly = recursively. Software adds two caveats to this. Software will allow an assembly to = invoke itself: true recursion. Software will also allow co-routines to = exist where routine A will invoke B which in turn will invoke A and so = on. As in recursion the invocation reaches some finite depth and then = stepwise returns to the invoker until reaching the level of the initial = invoker. In the normal manufacturing process this produces a "loop", a = non-allowed instance. Normally we protect the process by the = establishment of "low-level" codes which get checked during any = invocation of an assembly. For those of us embarrassed by writing = unwanted loops in programs you get a sense of the responsibility which = goes with the greater flexibility. ------=_NextPart_000_00B8_01C8B6A4.6AE78A30 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable charset=3Diso-8859-1"> Having just upgraded my FileStar to the latest = version, I=20 don't have to attend a meeting in which the speaker may or may not = show=20 up. For several reasons I will not attend this month's = meeting. I=20 have not had the time to devote to preparing material and = presentation for=20 the Programming SIG. I have resuscitated my "old" IBM = ThinkPad 760XD=20 to use in this endeavor. It charged up to 100%. Everything = seems to=20 be working. I've ordered a wireless connector to ease file = transfers and=20 downloads. As it already has APL and PL/I loaded, I will add = the Forth=20 package as well as an OS/2 version of LISP. I also have the = DOS-based=20 Trilogy. Perhaps I will add Prolog. That says while we will = focus on=20 a detailed understanding of Forth as a TIL and certainly gain a working=20 knowledge of assembly language, we can explore some other subjects = along=20 the way. I did want to address again Greg's demand about BNF = in=20 SL/I. Whether we call it the Backus-Normal or Backus-Naur Form = Backus, the=20 "father" of Fortran, had to have a means of defining the language syntax = for=20 which Fortran seemed ill-suited. Defining syntax has taken on a = number of=20 forms or variations since then from that of LEX to that of the syntax = diagrams=20 IBM uses in its LRMs. While the LEX statements may not look like C, it = uses a C=20 language program to produce C output. In fact all parsers = basically=20 derive from the same process. If we can write a parser, = that=20 which confirms the syntax of a source language, in the same source = language,=20 then its capability includes self-defining. If we can do it in C = or PL/I,=20 which we can, then certainly we can do it in SL/I. The tricky part for any language syntax lies in its = support of=20 recursion, where the thing we define we use in the definition. = That comes=20 straight out of manufacturing bills-of-material (BOM) where we have two = basic=20 entities, raw material and assemblies. While we cannot decompose a = raw=20 material, we define an assembly as something composed of raw materials = or other=20 assemblies. In this manner we define an assembly = recursively. Software adds two caveats to this. Software = will allow=20 an assembly to invoke itself: true recursion. Software will also = allow=20 co-routines to exist where routine A will invoke B which in turn = will=20 invoke A and so on. As in recursion the invocation reaches some = finite=20 depth and then stepwise returns to the invoker until reaching the level = of the=20 initial invoker. In the normal manufacturing process this produces a = "loop", a=20 non-allowed instance. Normally we protect the process by the = establishment=20 of "low-level" codes which get checked during any invocation of an=20 assembly. For those of us embarrassed by writing unwanted = loops in=20 programs you get a sense of the responsibility which goes with the = greater=20 flexibility. ------=_NextPart_000_00B8_01C8B6A4.6AE78A30-- ===================================================== To unsubscribe from this list, send an email message to "steward@scoug.com". In the body of the message, put the command "unsubscribe scoug-programming". For problems, contact the list owner at "postmaster@scoug.com". =====================================================
P.O. Box 26904 Santa Ana, CA 92799-6904, USA Copyright 2001 the Southern California OS/2 User Group. ALL RIGHTS RESERVED. SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners. |