COMPLEX.QQ : necessary adjustments for use by APL2/PC 
Author Message
 COMPLEX.QQ : necessary adjustments for use by APL2/PC

COMPLEX.QQ stored at WATSERV1, the subject of my last two letters, follows a
convention for APL-ASCII transliteration. Most APL's conform to the majority of
its prescriptions, but some are not 100% compatible. This is what I found out
when I typed it into a TryAPL2 workspace. TryAPL2 has the same keyboard as the
complete APL2 and probably earlier IBM APL's.
         The only difference which poses the need for adjustment is the role of
the right-bracket ("]"). In COMPLEX.QQ it is the right-arrow ("branch"). But in
APL2, "]" is the squad, whereas the curly brace "{"  -- "[" shifted -- is the
right-arrow. Therefore I must go back and replace all the squads I keyed in
with a right-arrow, for proper program branching to occur. Back-quote ("`") is
not standard for Diamond but it is retained here as a statement separator.

         Awkward punctuation poses the second need for adjustment. The dieresis
on line 8 of CATAN is the first character in a comment line, so it will be re-
moved.
         The names of some variables have been underscored. Whether this makes
them lowercase or literally underlines them is immaterial. The point is it's
not possible on a PC. So I have to register them in straight uppercase like any
other one.
         There will be a few conflicts with pre-existing names. We have "ERROR"
the variable and "Error" the program. I will resolve the conflict by inserting
underbar in one of them: "E_RROR". I'll do the same for "PMULLER" or "Pmuller",
two versions of a program that does the same thing. But the second is peculiar-
ly different. It references programs defined earlier but by a semi-underscored
name which should be a misspelling. It uses the Delta as a statement separator.
Normally, the diamond (which works) is used. And it uses some system services
that are unknown: Print, PrintC, PrintInit, []TCNL.
         This brings me to the last problem. The program appears to be invoking
some functions unique to STSC-APL. Hopefully, I will replace them by an APL2
counterpart with your help.
         Non-critical ones are: []SYMB , []LWSSIZE . At first I thought the
global variable GDeltaPRTFMT was one too. Now I am inclined to think it is what
was intended by "GDeltaFORMAT" in the introductory text. The original spelling
was not adhered to. Correct me if it is in fact unique to STSC-APL.
         Important ones are:

         Print , PrintInit , PrintC , []TCNL , and []VR

         The first four occur pre{*filter*}ly in the "Cmuller" program, version
the second. "[]VR" occurs in the "MakeFunction" program and is very ambiguous.
It could be from the context a Quad function or a misprint.

         Will you write me if you know what is going on here? If you need more
data like lines of code, I will send it to you. I want to find APL2/PC substi-
tutes for these STSC functions.
         When I succeed in adapting this workspace to TryAPL2, I will save it
as a .zip file and give it to anyone who asks for it. This is under the proviso
that it doesn't infringe the "copyright" [sic] of the code's author Ross Jay
Bettinger.                            -- Warren Vogt



Mon, 01 Jan 1996 12:58:23 GMT  
 
 [ 1 post ] 

 Relevant Pages 

1. COMPLEX.QQ overstrikes

2. Conversion of COMPLEX.QQ for use on TryAPL2

3. Have you used APL2/PC ?

4. APL2/PC in DOS Window with Win95 using APL fonts

5. getting apl characters using ibm apl2 on pc

6. APL2/PC display font problem

7. DIF files & APL2/PC

8. APL2/PC display font problem

9. 32-bit APL2/PC under Windows

10. APL2/PC; ATF-file format

11. APL2/PC and DRDOS

12. memory managers & APL2/PC

 

 
Powered by phpBB® Forum Software