Alfred Posted February 4, 2015 Share Posted February 4, 2015 Hi JAC, I would have preferred github over sourceforge, the quality of sf is poor these days, it is a faint copy of its better days, it has so many advertising, tracking and even add-on-downloads. Too commercial for my taste. For collaborating, github is so much nicer. If you want to use github, why not put it up there then. I doubt this is going to be a hotbed of activity that it can't be ported. I don't used Mads either, so I'll just stick with my Mac/65 copy of the code. Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 4, 2015 Share Posted February 4, 2015 (edited) Syntax conversion is done. Nesting analysis is in process. Interesting to see how the original files were merged together somehow. Maybe for speed of compilation when loading from disk or memory concerns. Unbundling reveals nice structures. Edited February 4, 2015 by JAC! 2 Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 5, 2015 Share Posted February 5, 2015 (edited) >11452 lines of source assembled in 3 pass >16384 bytes written to the object file All compile time errors are fixed. First part of ordering the segments is completed, too. The $b000 bank is about 99% identical to the cartdump already. Just realized how late/early is is - damn, that was fun :-) Edited February 5, 2015 by JAC! 4 Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 5, 2015 Author Share Posted February 5, 2015 Hi JAC! That is so fantastic! Wow! 16384? Incredible! What a time! Christmas in February! And we are all part of it. Quote Link to comment Share on other sites More sharing options...
cas Posted February 5, 2015 Share Posted February 5, 2015 If you want to use github, why not put it up there then. I doubt this is going to be a hotbed of activity that it can't be ported. I don't used Mads either, so I'll just stick with my Mac/65 copy of the code. In my view collaboration (working in a team) on Github is easier. It is not so much about 'git' as a tool, which is also available from SF. I will use SF, as JAC is more familliar with it. I had my projects on SF too, but after SF has been sold multiple times, my feeling is that the new owners just try to make as much money as possible with advertising and tracking of users. Github has a better business model, they charge for non-open-source projects, and the page is advertising free. But the main benefit is the way how code can be forked into local repos and then merged again, that makes collaboration on a project easier. Quote Link to comment Share on other sites More sharing options...
danwinslow Posted February 5, 2015 Share Posted February 5, 2015 I agree about github. Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 6, 2015 Share Posted February 6, 2015 Just couldn't got to bed without playing around with this. Splitting the 11 published source files into the original 34 source files is complete. Reordering is complete. The include sequence now matches the byte order in a type 15 one chip OSS cart. Just a few padding bytes, bank id bytes, text stamps like "ces" and maybe 1 or 2 wrong labels left. Current state is committed to sourceforge. C:\Users\D025328\Documents\Eclipse\workspace.ext\net.sourceforge.atari-action\JAC>fc /b ACTION.xex Action36-15.rom Comparing files ACTION.xex and ACTION36-15.ROM 0000075D: 01 03 000007E8: 01 03 000007ED: 01 03 00000827: 01 03 000008C3: 01 03 000008D4: FF 00 000008D5: FF 00 000008D6: FF 00 000008D7: FF 00 00000A7D: FF 00 00000A7E: FF 00 00000A7F: FF 00 00000FF1: FF 00 00000FF2: FF 00 00000FF3: FF 00 00000FF4: FF 00 00000FF5: FF 00 00000FF6: FF 00 00000FF7: FF 00 00000FF8: FF 00 00000FF9: FF 00 00001FE4: FF 20 00001FE5: FF 28 00001FE6: FF 41 00001FE7: FF 43 00001FE8: FF 53 00001FE9: FF 29 00001FEA: FF 20 00001FEB: FF 20 00001FEC: FF 4E 00001FED: FF 6F 00001FEE: FF 76 00001FEF: FF 65 00001FF0: FF 6D 00001FF1: FF 62 00001FF2: FF 65 00001FF3: FF 72 00001FF4: FF 20 00001FF5: FF 34 00001FF6: FF 2C 00001FF7: FF 20 00001FF8: FF 31 00001FF9: FF 39 00001FFA: FF 38 00001FFB: FF 33 00001FFC: FF 20 00001FFD: FF 20 00001FFE: FF 00 00001FFF: 01 00 000020A2: A4 5C 000020A3: A0 B7 000023B2: B7 DF 000029DA: B7 DF 00002FFD: FF 00 00002FFE: FF 00 00002FFF: 00 09 000032CD: 01 03 000032FB: 00 0A 00003E15: 22 42 00003E19: 1F 3F 00003E1A: 00 20 00003FFA: FF 63 00003FFB: FF 65 00003FFC: FF 73 00003FFD: FF 00 00003FFE: FF 00 00003FFF: 09 01 See also: http://atari800.cvs.sourceforge.net/viewvc/atari800/atari800/DOC/cart.txt 9 Quote Link to comment Share on other sites More sharing options...
flashjazzcat Posted February 6, 2015 Share Posted February 6, 2015 Excellent, fast work! Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 6, 2015 Author Share Posted February 6, 2015 JAC! Your midichlorian-value even exeeds the one from Yoda... http://en.wikipedia.org/wiki/The_Force_(Star_Wars)#Midichlorians Thank you so much. Quote Link to comment Share on other sites More sharing options...
fujidude Posted February 6, 2015 Share Posted February 6, 2015 Hi Carsten, since all other projects I'm participating in are on sourceforge and since it also supports git, that's my preferred environment. Yep, folks just need to not click on the ads etc. Should be trivial for sophisitcated computer users such as most of the people around here. Quote Link to comment Share on other sites More sharing options...
fujidude Posted February 6, 2015 Share Posted February 6, 2015 In my view collaboration (working in a team) on Github is easier. It is not so much about 'git' as a tool, which is also available from SF. I will use SF, as JAC is more familliar with it. I had my projects on SF too, but after SF has been sold multiple times, my feeling is that the new owners just try to make as much money as possible with advertising and tracking of users. Github has a better business model, they charge for non-open-source projects, and the page is advertising free. But the main benefit is the way how code can be forked into local repos and then merged again, that makes collaboration on a project easier. Some good points there, but it isn't the end of the world to use SF. Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 7, 2015 Share Posted February 7, 2015 @Luckbuck: Well, no. I'm just concentrating on such a small amount of reality that all my mental capacity goes into 6502. Ask my wife :-) Everything except bank ids and the serial number seems to be OK now. Comparing files ACTION.rom and ACTION-TYPE-15.ROM 00000819: 04 09 0000081E: 04 09 00000855: 04 09 0000085A: 04 09 00001FFF: 03 00 00002FFF: 00 09 000032FB: 00 0A 00003FFF: 04 01 The bank ids require some sleep, due to the total mix of types and dump versions (3/15/43). Just can't get it straight in my head now. Also not sure about the serial id at $32fb. That's $0000 in the source because it was set as part of production as it seems. 41 A2FA 00 00 serial .word 0 ; serial number of ROM 42 ; to be filled in before blowing ROM Quote Link to comment Share on other sites More sharing options...
UNIXcoffee928 Posted February 7, 2015 Share Posted February 7, 2015 Is there any chance that you can relieve us of this accursed OSS cart format, altogether? I have a Bit-3 80 column board in my Atari 800, and always wanted to take advantage of the Action! 80 column functionality, but the cartridges themselves have always been prohibitively expensive. Even nowadays, the ebay price is always high, and you never know what version cartridge it is. Is there any hope that you could convert this thing into something that could be used with SIO2PC? Quote Link to comment Share on other sites More sharing options...
+Stephen Posted February 7, 2015 Share Posted February 7, 2015 Is there any chance that you can relieve us of this accursed OSS cart format, altogether? I have a Bit-3 80 column board in my Atari 800, and always wanted to take advantage of the Action! 80 column functionality, but the cartridges themselves have always been prohibitively expensive. Even nowadays, the ebay price is always high, and you never know what version cartridge it is. Is there any hope that you could convert this thing into something that could be used with SIO2PC? Is there a schematic for Bit3 anywhere? Failing that, has anyone photographed the front and back of PCBs? Some time ago, I managed to source the main video chip from that device. I would love to clone it since I don't think I'll ever manage to get to buy one. Quote Link to comment Share on other sites More sharing options...
UNIXcoffee928 Posted February 7, 2015 Share Posted February 7, 2015 Is there a schematic for Bit3 anywhere? Failing that, has anyone photographed the front and back of PCBs? Some time ago, I managed to source the main video chip from that device. I would love to clone it since I don't think I'll ever manage to get to buy one. As far as I know it has not been emulated, either. I can take photos of it, for you, sometime this week. If you can physically reproduce it, then, perhaps, it's functionality can be added to the emulators, if you share the knowledge that you gain while reverse engineering it... that would help everybody. Editing Action code in 80 columns could really inspire a lot of people to code new stuff for the Atari. It's such a well thought out language, but working with it in 40 columns is far from ideal when doing well-structured procedural programming, because you can't tab as far as you would like to, and separate the code blocks nicely, etc. Quote Link to comment Share on other sites More sharing options...
AtariGeezer Posted February 7, 2015 Share Posted February 7, 2015 Is there a schematic for Bit3 anywhere? Failing that, has anyone photographed the front and back of PCBs? Some time ago, I managed to source the main video chip from that device. I would love to clone it since I don't think I'll ever manage to get to buy one. Pics of the board and ROM dump is still among many things on my ToDo list Still need to fix a cable for the 1080 monitor (Thought I had 2 1084's) and take pics of it in action... Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 8, 2015 Author Share Posted February 8, 2015 (edited) @unixcoffe928: why so complicated? Carts: http://www.atarimax.com/freeshippingsale/ 80: https://atariwiki.org/wiki/Wiki.jsp?page=The%20XEP80 As far as I know, the Action! cart is running with 80 columns. And(!) 80 is just the beginning! Further: Maybe this time with Mac/65 1.02 and in the future: OSS 6 in 1: with Bug /65 ? CU all out there in the galaxy. Let the new Atari age begin! Edited February 8, 2015 by luckybuck Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 8, 2015 Share Posted February 8, 2015 >Is there any chance that you can relieve us of this accursed OSS cart format, altogether? The binary will be available in multiple output formats, including the OSS formats so people can update their carts. 16k plain ROM is fairly straight, once the source is relocatable. Of course it comes at the expense of 8k less RAM for code & source. That's why OSS came with that special design for their productivity cars. >Is there any hope that you could convert this thing into something that could be used with SIO2PC? Similar to the 16k Plain ROM. Of course loading from disk always comes with the risk that a bug also trashes ACTION! itself. On idea that I have is to have a version that uses the RAM under the XL OS optionally as "backup". >wanted to take advantage of the Action! 80 column functionality If there anything special in ACTION! itself, apart from the fact it uses the E: handler? Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 8, 2015 Share Posted February 8, 2015 Now that's what I call fun with code. It took me quite a while to understand this until I finally encountered the "inc sargs,x". ; ; ; Analysis: RAM Protection ; ; The the RAM version of Action! contains a global equate "ramzap" that ; activates conditional assembly of eleven self-destruction code fragments. ; These fragments change the content of parts of the Action! ROM area ; in many different ways. They are intended to prevent the creation of ; a file version that can be loaded to the RAM of the machine. ; When the Action! code is in the ROM area, these code fragment have no effect. ; When the Action! code is in the RAM area, thee code fragments destroy the code. ; When the "ramzap" equate is set to zero, the code fragments are replaced ; by a NOP sequence of equal size. This way the overall program structure ; and label addresses remained constant, but Action! or its parts could be ; run from the machine RAM or a special RAM cartridge. 25 = 0001 ramzap equ 1 ; to zap or not to zap, that is the question? -- 60 ; fake out zero check in XIOstr 61 .if ramzap 62 B02A 91 A5 sta (arg5),y 63 .else 64 nop 65 nop 66 .endif -- 283 .if ramzap 284 B6F6 91 A1 sta (arg1),y ; zap RAM if any 285 .else 286 nop 287 nop 288 .endif -- 448 .if ramzap 449 BD4F 91 A8 sta (arg8),y 450 .else 451 nop 452 nop 453 .endif -- 739 .if ramzap 740 BEED F6 00 inc sargs,x 741 .else 742 nop 743 nop 744 nop 745 .endif -- 851 ;.IF RAMzap 852 ; LDY LTab+1,X 853 ; LDA LTab,X 854 ;.ELSE 855 BF5A BC E5 04 ldy lsh+1,x 856 BF5D BD E4 04 lda lsh,x 857 ;.ENDIF -- 71 .if ramzap 72 A036 9D 4A A0 sta mulb,x 73 .else 74 nop 75 nop 76 nop 77 .endif -- 143 .if ramzap 144 A1FD FE 2B A1 inc stm,x 145 .else 146 nop 147 nop 148 nop 149 .endif -- 48 .if ramzap 49 A3BD DE 95 A3 dec chkerr-$10,x 50 .else 51 nop 52 nop 53 nop 54 .endif 55 A3C0 60 rts -- 1449 ; -------- 1450 getexp .if ramzap 1451 A8B4 EE 92 AF inc cgopscd 1452 .else 1453 nop 1454 nop 1455 nop 1456 .endif -- 2263 .if ramzap 2264 ADCA 91 64 sta (adress),y 2265 .else 2266 nop 2267 nop 2268 .endif -- 93 ;SPLInit .proc ; init compiler RAM 94 .if ramzap 95 A334 20 B6 A3 jsr zap4 96 .else 97 nop 98 nop 99 nop 100 .endif -- 72 A3B0 1E zap2 .byte $1e 73 A3B1 6D A5 .word delch 74 A3B3 FE zap3 .byte $fe 75 A3B4 9A A5 .word backsp 76 A3B6 7E zap4 .byte $7e 77 A3B7 2C A4 .word insrtch Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 8, 2015 Author Share Posted February 8, 2015 (edited) JAC! You are a miracle! Please find attached the file version from: OSS Action! Version 3.6 with DOS XL 2.20 and REAL files.atr with some strange REAL files on it. The DOC file can only be opened with the cart... due to RAM problems... Did found it in a lost archive... Have fun further, we are with you all the time! :-) Wil try to get the Basic XE source, too. stand by Edited February 8, 2015 by luckybuck Quote Link to comment Share on other sites More sharing options...
+JAC! Posted February 8, 2015 Share Posted February 8, 2015 The one thing I wonder about is what the "AMPL" in the file and label names stands for. "Action Machine Programming Language"? Or even related to this.. http://www.ampl.com/contents.html Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 8, 2015 Author Share Posted February 8, 2015 Before Action! it was called: AMPL But Action! sounds more cool. :-) Quote Link to comment Share on other sites More sharing options...
gozar Posted February 8, 2015 Share Posted February 8, 2015 >Is there any chance that you can relieve us of this accursed OSS cart format, altogether? The binary will be available in multiple output formats, including the OSS formats so people can update their carts. 16k plain ROM is fairly straight, once the source is relocatable. Of course it comes at the expense of 8k less RAM for code & source. That's why OSS came with that special design for their productivity cars. >Is there any hope that you could convert this thing into something that could be used with SIO2PC? Similar to the 16k Plain ROM. Of course loading from disk always comes with the risk that a bug also trashes ACTION! itself. On idea that I have is to have a version that uses the RAM under the XL OS optionally as "backup". >wanted to take advantage of the Action! 80 column functionality If there anything special in ACTION! itself, apart from the fact it uses the E: handler? What about using extended memory? I would think most of us that want to use it have more than 64K in our Ataris. And what about extended memory for code space and compiling? Enough questions from me, what you've done is incredible! Quote Link to comment Share on other sites More sharing options...
luckybuck Posted February 8, 2015 Author Share Posted February 8, 2015 Already discussed above. One thing after another. You already suffered more than 3 decades. And now a few days shall matter? Come on... Quote Link to comment Share on other sites More sharing options...
+CharlieChaplin Posted February 8, 2015 Share Posted February 8, 2015 Well, I have a hacked/cracked file-version of Action! in my collection that works with 48k RAM (not much free RAM available) and another file-version that works with 64k RAM (utilizing RAM under the OS, more free RAM available). But as of yet, I do not have a file-version that works with extended RAM... Action_file.zip Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.