The Museum of HP Calculators

HP Forum Archive 20

[ Return to Index | Top of Index ]

Sandmath-II and 41Z - questions to Angel
Message #1 Posted by Vincent Weber on 4 Oct 2011, 3:34 a.m.

Hi Angel if you are reading,

I am getting confused over the versions and usage of your wonderful modules...

For Sandmath-II, is it a replacement for the 8K Sanndmath module ? it seems that no 12K version is available any more on TOS, does it mean we need to use the ALGEBRA module instead ?

For 41Z, I downloaded the latest version, and I am getting confused over the key assignements. When I execute ZK?YN and press Y, the "SIGMA+" key, which used to prompt "Z" for Z commandes, now displays "-HP 41Z". Is it normal ?

Also, be it with V41 or i41CX+, CAT 2 does not display the whole 41Z catalog (only the ZMATH and ZVECTOR "subdirs")... unless the CCD module is plugged in, in which case everything is ok... ?

it would be nice if you could post a summary about with are the current versions of all your modules, so that no confusion exists ;)

Many thanks and best regards,

Vincent

      
Re: Sandmath-II and 41Z - questions to Angel
Message #2 Posted by 聲gel Martin on 4 Oct 2011, 4:23 a.m.,
in response to message #1 by Vincent Weber

Hi Vincent, yes I'm reading :-)

First off, you're not alone in this confusion - I myself need to have a roadmap to keep track of all the different versions and releases of these two modules - in particular the SandMath saga - so I can't imagine what's like for the casual user.

So let me try to do a summary recap. The 41Z is easier, let's tackle it first.

Because of the different addiditons and enhancements made to it, the FAT function order has slightly changed from some older revisions to the current one (9H). This explains why your key assignments can be mixed up, if they were made using a previous version of the module. The internal "ZY?YN" function is supposed to be alright whichever revision, in other words consistent with the changes made. Revision 9H assigns "ZKBRD" to the SIGMA+ key, but it's possible that the one you're using has some glitches.

I typically use only ZKBRD and not the complete USER keyboard, as that gives me better compatibility with the standard functions (i.e. for real numbers) and with the SadnMath launchers - Also it doesn't take up so many KA registers.

As to the CAT2 listings - you're correct, because of the length of the "header" functions being less than 7 they don't show up in the standard CATALOG - only the CCD-style. I did it this way to allow direct execution of the functions using "XEQ, ALPHA, fcn_name". Since I always have the AOSX module plugged in (the 4k subset of the CCD) it's a non-issue to me, but I could change it of course.

BTW, revision 9H is not released yet . I just finished making big changes to the launchers code, replacing an old pedestrian scheme with a MCODE-table based new one. This has resulted in huge space savings, more than 256 bytes freed up so I could "squeeze" in the wonderful ZZETA program - adapted from JM Baillard's original one.

I'm leaving the SandMath for later, this is getting kinda long. Thanks again for your interest in these modules.

Best, 簍

            
Re: Sandmath-II and 41Z - questions to Angel
Message #3 Posted by Vincent Weber on 4 Oct 2011, 4:44 a.m.,
in response to message #2 by 聲gel Martin

Hey Angel, Many thanks for your fast and detailed answer ! I shall wait for the Sandmath saga then ;) Just a few clarifications about 41Z: I am using 9F, the latest available on TOS. Starting fresh with a clear memory, using ZK?YN does NOT assign ZBRD properly. But the previous version (supplied with V41) does ! Also, with 9F, clearing memory trying leaves the calc in a state when nothing can be done - compley bricked until your remove the module, on both V41 and i41CX+. Is it normal ?

Thanks and cheers,

Vincent

                  
Re: Sandmath-II and 41Z - questions to Angel
Message #4 Posted by 聲gel Martin on 4 Oct 2011, 6:55 a.m.,
in response to message #3 by Vincent Weber

Quote:
clearing memory trying leaves the calc in a state when nothing can be done - compley bricked until your remove the module, on both V41 and i41CX+. Is it normal ?

Well, it's a bug - so it's not "normal". I changed the code to make use of the CPU M register not realizing that the [RESIZE] routine overwrites it. A previous version was using register Q(9), but I consolidated the two interrupts (Memory_Lost and CALC_ON) and that's when it got broken.

Anyway it's also fixed in version 9H, if you tell your email I'll be glad to send it to you.

Cheers, 簍

                        
Re: Sandmath-II and 41Z - questions to Angel
Message #5 Posted by Vincent Weber on 4 Oct 2011, 7:09 a.m.,
in response to message #4 by 聲gel Martin

Thanks Angel ! Actually you do have m'y email, since we had an chat by mail with Antonio this morning ;) thanks if you can send it ! Cheers, Vincent

      
Sandmath Modules Roadmap
Message #6 Posted by 聲gel Martin on 4 Oct 2011, 7:25 a.m.,
in response to message #1 by Vincent Weber

In the beginning the SandMath was available in two flavors, the 8k version and the 12k version - with the latter just adding a Matrix/Polynomial section on top of the 8k version, and thus a real superset.

The 12k footprint has some obvious issues that made me reconsider that arrangement. Besides, more Matrix routines were competing for the same space, and there was no available room in the 12k version to accomodate for them. The straw that broke the camel's back was my new "Matrix Editor" mode, which required substantial room to code and would never fit in the 12k - so it was time for a new approach.

That's when I rolled the 8k version into the "SandMath-II", as a direct replacement, and when I released the ALGEBRA module. This one is a self-contained ROM, that does not have any dependencies on the SandMath-II - so they can be used independently.

So far so good, but there are a few side frills that make the story more interesting :-).

Shortly after this (or it may have been simultaneously, not sure anymore) I decided to put together a Matrix add-on to the SandMath-II. This is available as the MATRIX ROM, a 4k module that only works in combination with the SandMath-II. Its contents is basically the same as the matrix-related stuff in the ALGEBRA module - but of course it lacks everyting else. The 4k-MATRIX ROM is a subset of the ALGEBRA Module.

But wait, the plot thickens:- As both the 8k-ALGEBRA and the 4k-MATRIX rely on the ADVANTAGE Pack (all those wonderful MCODE Array functions from the CCD being there), it occurred to me that the ultimate MATRIX module should be a page-switched, 12-k construct occupying just 8k, with (at least) the same contents as the MATRIX rom plus all the Matrix functionality of the Advantage Pack.

And that's the ADVANCED_MATRIX Pac - which BTW also retains both SOLVE and INTEG just to make things more complicated :-) In truth, this was motivated by the page-switching design, and also due to the fact that one of the 4k blocks doesn't have a FAT (obviously). SOLVE and INTEG are monster routines, filling up 3/4th.s of a block, so they lend themselves splendidly to this usage.

In case you wonder, the ADVANCED-MATRIX Pack does not require the SandMath-II module either, so there's some inevitable redundancy but the dependencies are removed. However it does not have the Polynomial programs of the ALGEBRA ROM, so alas, it's not a superset of it.

The obvious missing piece in all this jigsaw puzzle would be a 4k-Polynomial ROM, to use on top of the ADVANCED_MATRIX Pac to come up with a real superset of the ALGEBRA Module with the Advantage Pack power into it. Are we lost yet? I am, somebody please draw a sketch! ... "-)

Which ones to use? Your choice, depends on what you're more likely to do or need at a given point.

PS. In case you got this, let me throw another couple of logs to the bonfire:- Another two modules that are very closely related to the ones described above are the "BESSEL ROM" and the "POLYNOMIAL_ROM", done in collaboration with Jean-Marc Baillard. Those two are also self-contained, and treat the subjects with a deeper/wider reach. Not for the faint of heart, but a wonder for the adventurous.

Hope this helps, and again thanks for your interest in all those roms.

Best, 簍

            
Re: Sandmath Modules Roadmap
Message #7 Posted by Geir Isene on 4 Oct 2011, 8:33 a.m.,
in response to message #6 by 聲gel Martin

Would it help to move all the reusable code into it's own 4K MATHLIB module and then have the other modules as pure 4K modules calling MATHLIB when needed? More modular methinks.

                  
Re: Sandmath Modules Roadmap
Message #8 Posted by 聲gel Martin on 4 Oct 2011, 10:25 a.m.,
in response to message #7 by Geir Isene

Yes methinkstoo - albeit it's easier to say it than to do it. At the end of the day it's all about footprint and there's only 32k that can be actively mapped at a time. Add to that the other system constraints (limited FAT entries, XROM id# conflicts) and you'll have a nice challenge in your hands.

On the other hand, work to get port #4 as a silent library is already in progress... just not only with a MATH focus though.

                        
Re: Sandmath Modules Roadmap
Message #9 Posted by Eric Smith on 4 Oct 2011, 11:11 a.m.,
in response to message #8 by 聲gel Martin

Remember that using block 4 will be incompatible with the HP-IL module when the printer enable switch on the latter is turned off. The switch moves the printer ROM from page 6 to page 4.

                              
Re: Sandmath Modules Roadmap
Message #10 Posted by Geir Isene on 4 Oct 2011, 12:57 p.m.,
in response to message #9 by Eric Smith

Why turn it off, though?

                                    
Re: Sandmath Modules Roadmap
Message #11 Posted by 聲gel Martin on 4 Oct 2011, 2:21 p.m.,
in response to message #10 by Geir Isene

because you may:

a) use the IR printer, and the IR module takes page #6
b) not have any printer and want to spare the CAT2 searches
c) none of the above :)

                                          
Re: Sandmath Modules Roadmap
Message #12 Posted by Eric Smith on 4 Oct 2011, 6:01 p.m.,
in response to message #11 by 聲gel Martin

Or because you have the 82143A printer.

            
Re: Sandmath Modules Roadmap
Message #13 Posted by Vincent Weber on 4 Oct 2011, 9:04 a.m.,
in response to message #6 by 聲gel Martin

Thanks Angel for the detailed explanation ! Just to add to the confusion: I noticed you also did a curve fitting module. How is it compared to the ADVANTAGE CFIT or with the one in AEC ? i noticed that you left the SILOOP routine in your advanced matrix pac and that you have some linear r嶲ression routines in Sandmath, how is all this related ? Thanks and cheers Vincent

                  
Re: Sandmath Modules Roadmap
Message #14 Posted by 聲gel Martin on 4 Oct 2011, 9:59 a.m.,
in response to message #13 by Vincent Weber

The CurveFitting Module is a compendium of all the related techniques from the other modules, as follows:

1. includes the AECrom functions and programs - impressive!
2. includes the Advantage Pack CFIT
3. includes W. Kolb's seminal CF program
4. throws in the Standard pack CFIT just for kicks
5. plus a selection of statistical and probability functions, among which you've seen the Linear regression (also in the SandMath)

Pfeeww....

PS. SILOOP and SIRTN must remain in the ADVANCED_MATRIX as they're an integral component of SOLVE / INTEG. A real wonderful implementation, worth spending some time studying...

                        
Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #15 Posted by gene wright on 4 Oct 2011, 10:12 a.m.,
in response to message #14 by 聲gel Martin

Angel...

So many modules, so confusing to most of us. :-)

Favor?

Please give us a short list of the modules you presently have on TOS along with a short explanation of what is in them. 2 sentences perhaps. Please indicate which roms are "obsolete"

I think it would also help if you could indicate which roms are in the 41CL and what the 4 letter rom abbreviation is in the 41CL.

This is not a criticism! I LIKE your rom enthusiasm but it is difficult to keep them straight! :-)

And, any chance we can get a curvefit .rom format? Our 41CL units are waiting. ;-)

Edited: 4 Oct 2011, 10:28 a.m.

                              
Re: Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #16 Posted by 聲gel Martin on 4 Oct 2011, 10:48 a.m.,
in response to message #15 by gene wright

No worries, here's the list in all its alphabetical glory: (have you used the CLLIB function already?)

#	ID	Size	Name	        Author
1	A41P	12k	Advantage Pac	HP Co.
2	AADV	4k	Advantage Apps	J-F Garnier
3	ADV1	16k	Adventure_1	Angel Martin
4	ADV2	12k	Adventure_2	Angel Martin
5	AEC3	8K	AECROM 13-digit	Angel Martin
6	AECR	8k	AECROM	        Red Shift
7	AFDE	8k	AFDC1	        GunZen
8	AFDF	8k	AFDC2	        GunZen
9	AFIN	4k	Auto Finance	GMAC
10	ALGG	8k	Algebra ROM	Angel Martin
11	ALGY	4k	Astrology	??
12	ALPH	4k	ALPHA ROM	A. Martin & D. Wilder
13	AOSX	4k	AMC OS/X	Angel Martin
14	ASM4	4k	Assembler4	
15	ASMB	4k	Assembler3	
16	ASTT	16k	ASTRO-2010 	Jean-Marc Baillard
17	AUTO	4k	Auto-Start/Dup	HP Co.
18	AV1Q	4k	AV1 ROM	        Beechcraft
19	AVIA	4k	Aviation Pac	HP Co.
20	B52B	8k	B-52 ROM	Boeing 
21	BCMW	4k	BCMW ROM	
22	BESL	8k	Bessel ROM	A. Martin & JM Baillard
23	BLDR	8k	BLD ROM	W. Doug Wilder
24	BLND	4k	Bufferland ROM	Angel Martin
25	CCDP	8k	CCD Plus	Angel Martin
26	CCDR	8k	CCD Module	W&W GmbH
27	CCDX	4k	CCD OS/X	Raymond del Tondo
28	CHEM	4k	Chemistry User	
29	CHES	8k	Chess/Rubik's	A. Martin & C. Roetlgen
30	CIRC	4K	Circuit AnalysisHP Co.
31	CLIN	4K	Clinical Lab	HP Co.
32	CLUT	4k	CL Utils 1H	Angel Martin
33	CURV	8k	Curve-Fitting	Angel Martin
34	CVPK	8k	Cv-Pack ROM	
35	DA4C	4k	DisAssembler4C	W. Doug Wilder
36	DACQ	8k	Data Acq Pac	HP Co.
37	DASM	4k	DisAssembler4D	W. Doug Wilder
38	DAVA	4K	David Assembler 2C	David van Leeuwen
39	DEVI	8k	HP-IL Development	HP Co.
40	DIIL	4k	HP-IL Diagnostics	HP Co.
41	DMND	4k	Diamond ROM	
42	DYRK	4k	Dyerka ROM 	David Yerka
43	E41S	8k	ES41 Module	
44	ESML	4k	ES MLDL 7B	
45	EXIO	4k	Extended I/O 	HP Co.
46	EXTI	4k	Extended-IL ROM	Ken Enery
47	FACC	4k	300889_FACC	??
48	FINA	4k	Financial Pac	HP Co.
49	FRTH(*)	8k	FORTH Module	Serge Vaudenay
50	FUNS	8k	FunStuff Module	Angel Martin
51	GAME	4k	Games Pac	HP Co.
52	GMAS	4k	Auto Finance-2	GMAC
53	GMAT	8k	Auto Finance-3	GMAC
54	HCMP	4k	HydraComp ROM	Paul Monroe
55	HEPR	4k	HEPAX RAM	VM Electronics
56	HEPX	16k	HEPAX Module	VM Electronics
57	HOME	4k	Home Mngmt. Pac	HP Co.
58	ICDO	4k	Icode ROM	
59	IDC1	8k	ML-ICD	        BCMC 1987
60	IDC2	4k	BG/UG IDC	BCMC 1985
61	JMAT	8k	JMB Math	Jean-Marc Baillard
62	JMTX	8k	JMB Matrix	Jean-Marc Baillard
63	ILBF	4k	IL-Buffer	Angel Martn
64	KC135	12k?	Weight & Balance Comp.	??
65	L119	8k	AFDC-1E-003	Zengun
66	LAIT(*)	4k	LaitRAM XQ2	LaitRam Corp.
67	LAND	4k	Land Navigation	Warren Furlow
68	LBLS	4k	Labels ROM	W. Doug Wilder
69	MADV	12k	Modified Advtg	Angel Martn
70	MATH	4k	Math Pac	HP Co.
71	MCHN	4k	Machine Pac	HP Co.
72	MDP1	8k	AFDC-1F ROM	Zengun
73	MDP2	8k	AFDC-1F ROM	Zengun
74	MELB	4k	Melbourne ROM	PPC Members
75	MFLB	4k	Mainframe Labl  David van Leeuwen
76	MILE	8k	Military Eng ROM	??
77	MLRM	4K	ML ROM	        Frits Ferwerda
78	MPLB	8k?	Multi-Prec. Lib	Peter Platzer
79	MTRX	4k	MATRIX ROM	Angel Martin
80	MTST	4k	MC Test ROM	??
81	MUEC	8k	Muecke ROM	Mcke Software GmbH
82	NAVI	8k	Navigation Pac	HP Co.
83	NCHP	4k	NoVoCHAP	G. Isene & A. Martin
84	NFCR	4k	NFC ROM	        Nelson F. Crowe
85	NPAC	8k	NavPac ROM	??
86	NVCM	8k	NaVCOM 2	??
87	OILW	8k	OilWell Module	Jim Daly
88	P3BC	16k	Aviation P3B/C	??
89	PANA	8k	PANAME ROM	S. Bariziene & JJ Dhenin
90	PARI	4k	PARIO ROM	Nelson F. Crowe
91	PCOD	4k	Proto-Coder 1A	Nelson F. Crowe
92	PETR	8k	Petroleum Pac	HP Co.
93	PFST	4k	ProfiSet	Winfried Maschke
94	PLOT	8k	Plotter Module	HP Co.
95	PMLB	4k	PPC Melb ROM	PPC Members
96	POLY	8k	Plolynomial ROM	A. Martin & JM Baillard
97	PPCM	8k	PPC ROM	        PPC Members
98	PRIQ	8k	PRIDE ROM	
99	PS0F	4k	Profiset Subset	Winfried Maschke
100	QUAT	8k	Quaternions ROM	Jean-Marc Baillard
101	RAMP	4k	RAMPage Module	Angel Martin
102	REAL	8k	Real State Pac	HP Co.
103	ROAM	4k	ROAM Module	Wilson Holes
104	ROMS	4k	SV's ROM	Serge Vaudenay
105	SANA	12k	SandMath-12k	Angel Martin
106	SBOX	8k	SandBox	        Angel Martin
107	SEAK	4k	SeaKing MK5	Navy Air
108	SECY	4k	Securities Pac	HP Co.
109	SGSG	4k	Gas Module	SGS Redwood
110	SIMM	16k	SIM Module	
111	SKWD	4k	Skwid's BarCode	Ken Emery
112	SMCH	8k	Speed Machine	Alameda Mngmt. Corp.
113	SMPL	4k	Simplex Module	Phillipe J. Roussel
114	SMTS	8k	SandMath-8k	Angel Martin
115	SND2	8k	SandMath-II	Angel Martin
116	SPEC	4k	Spectral An.    Jean-Marc Baillard
117	SRVC(*)	4k	Service ROM	HP Co.
118	STAN	4k	Standard Pac	HP Co.
119	STAT	4k	Statistics Pac	HP Co.
120	STRE	4k	Stress Analysis	HP Co.
121	STRU	8k	Structural An.	HP Co.
122	SUPR	8k	SUP-R-ROM	James W. Vick
123	SURV	4k	Surveying Pac	HP Co.
124	THER	4k	Thermal Pac	HP Co.
125	TOMS	4k	Tom's ROM	Thomas A. Bruns
126	TOOL	4k	ToolBox-II	Angel Martin
127	TREK	4k	Start Trek	Angel Martin
128	TRIH	4k	83Trinh	        Phil Trinh
129	UNIT	4k	Unit Conversion	Angel Martin
130	USPS	8k	Mail Delivery	USPS
131	YFNS	4k	Alternate YFNS	Monte Dalrymple
132	YFNZ	4k	Main YFNS	Monte Dalrymple
133	Z41Z	8k	41Z Module	Angel Martin
134	ZENR	4k	Zenrom	        Zengrange Ltd.
135	ZEPR	4k	Programmer	Zengrange Ltd.

                                    
Re: Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #17 Posted by gene wright on 4 Oct 2011, 10:52 a.m.,
in response to message #16 by 聲gel Martin

Great, now which of yours in this list are obsolete?

(My CLUTILs version did not have that function - hopefully in a week or two it will)

                                          
Re: Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #18 Posted by 聲gel Martin on 4 Oct 2011, 11:56 a.m.,
in response to message #17 by gene wright

Quote:
Great, now which of yours in this list are obsolete?

none!

They're all there, actively supported at no cost :-)

                                    
Re: Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #19 Posted by Monte Dalrymple on 4 Oct 2011, 12:27 p.m.,
in response to message #16 by 聲gel Martin

There are a couple of errors in this table.

The mnemonic for the Mainframes Labels is "MLBL"

The mnemonic for the Multiprecision Library is "MLTI"
and it is a 4k ROM.

Remember that mnemonics must use unique first and last
characters to be parsed correctly.

                              
Re: Angel, favor ? And the Curvefit is a .mod not a .rom !
Message #20 Posted by Monte Dalrymple on 4 Oct 2011, 12:21 p.m.,
in response to message #15 by gene wright

Gene, your machine includes the curvefit ROM. It's the upper
page of the image loaded with the CURV mnemonic.


[ Return to Index | Top of Index ]

Go back to the main exhibit hall