n233dk@tamuts.tamu.edu (Rick Grevelle) (10/12/90)
While it is understandable that many users would like too see HP release more information about the internals of their calculators, personally I'm rather ambivalent. To me, the HP48 is a Rubic's Cube; a puzzle of sorts. Releasing limited details about the 48's internals could enhance playing the GAME. Disclose too much information, and it could possibly ruin most of the fun. At first this may seem a somewhat strange and selfish attitude, but considering all of the arguments regarding the alleged unethical publishing practices of HP employees, it's really not. Undoubtedly, those individuals who would seriously limit, or completely prohibit, the flow of this valuable source of information are unquestionabley moon struck. Information of this caliber does indeed serve to enhance the GAME, as well as provide those more serious users with valuable new tools. Although I've yet to aquire Mr. Donnelly's Tool Kit, I fully intend to do so. Judging from the reviews seen here, not only do I feel that the software merits high praise, but also requires the continuous support of the user community if we are to see more such software. After all, individual members of the design team are not required to publish in order to stay employed. Mr. Donnelly and various other design team members who chose to invest their personal time and energy towards enhancing the very products they are producing, should be commended for their efforts. A case and point. It has now been so long that I can't actually recall exactly when it was I first purchased Dr. Wickes' book, "HP-28C/S Insights". Studying this excellently written document for uncountable hours enabled to me master the stack oriented programming language, RPL, utilizing KEY WORD commands and functions. While it took Alonzo Gariepy and his Processor Notes to provide me with the necessary tools to begin Saturn Assembler, it was Wickes' book which laid the foundation so crucially needed for development of an understanding of of RPL internally. Here I have chosen to illustrate the point using an alternate stack display routine. It exemplifies RPL programming on an internal level very much in the manner of HP. Programs such as this can be easily modified in order to call user written machine code routines. An excellent example that utilizes this technique has already been provided by Dr. Wickes himself when he posted the encoding schemes ->ASC and ASC->. Currently, I'm writing almost all of my routines on this level for the reasons that programs written utilizing KEY WORDS are often quite slow by comparison, and considerably less challenging. The routine displays a seven level stack leaving the menu bar intact. It's considerably faster than any of the other alternate stack display routines to date. Because I've only documented about a thousand SYSEVALs in the 48, there is obviously a great possiblility the routine could be improved to a point to where it would not be recognizable from what appears here. Documenting new SYSEVALs and making improvements to already existing routines such as these provides a great deal of pleasure. Should HP choose to make available to the public their colossal files documenting the 48's internals, they would in essence solve the puzzle and kill the thrill of reverse engineering the calculator. The RPL disassembly is presented in the same style as Eric Toonen's ROM map listing for the 28S. Only a brief explanation appears aside each address. Several of the addresses can NOT be SYSEVALed as they're only availble for use in routines written in this fashion. To package up the program I've included two choices. The first, CALTD, can be entered via the 48's keybaord and will in turn create the alternate display program ALTD on the stack. Make doubly sure to checksum CALTD before using it to create ALTD. I'll be glad to answer any questions. Rick Grevelle Alternate Stack Display Routine 02D90 begin_RPL 0403F short_integer_<8h> 073CE start_1_to_n-1 07221 current_loop_count 18DBF short->real 162AC real->string 15442 string_": " 05193 +_1string2string 0314C depth 07221 current_loop_count 03D83 boolean_>_1short2short 03188 dup 619AD branch_if_true 02D9D begin RPL 03244 drop 0403F short_integer_<8h> 07221 current_loop_increment 03DE0 -_1short2short 12429 display_string_131x8_1short2string 07334 next 391EE freeze_display 0312B end RPL 61993 branch_if_false 02D9D begin RPL 032C2 over 15978 ->string_1any 05193 +_1string2string 0403F short_integer_<8h> 07221 current_loop_count 03DE0 -_1short2short 12429 display_string_131x8_1short2string 03140 depth 04035 short_integer_<7h> 624BA min_1short2short 0339E rolld_1short2any...nany 07334 next 0400D short_integer_<3h> 142FB freeze_display_1short 0312B end RPL 0312B end RPL CALTD (BYTES: #A1B8h 243.5) %%HP: T(1)A(D)F(.); \<< # 1073CE0403F02D9Dh # 42162AC18DBF0722h # 2210314C05193154h # 19AD0318803D8307h # 403F0324402D9D6h # 41242903DE007221h # 930312B391EE0733h # 978032C202D9D619h # 72210403F0519315h # 314C1242903DE00h # 40339E624BA04035h # 2B142FB0400D0733h # 312B031h 1 12 START # 5193h SYSEVAL NEXT # 4003h SYSEVAL # 62B9Ch SYSEVAL \>> ALTD (Requires ASC->) %%HP: T(1)A(D)F(.); "D9D20F3040EC37012270FBD81CA2612445139150C41301227038D3088130DA91 6D9D2044230F3040122700ED309242143370EE193B213039916D9D202C230879 5139150F3040122700ED3092421C413053040AB426E933043370D0040BF241B2 130B2130C5D9"