Wrathchild

Here's how far I got elite_tunes.zip

Please, please - if you're experimenting with sounds then can you use the C64's Elite's 'Blue Danube' tune as a base . That has a good mix of instruments .

Had in on watch for a few days now... hope of winning it low though Also saw their 'maniac mansion' for the C64 - but the price of that shot up quickly - would be fun to make such an 'engine' for the A8. I started reverse engineering it - but its very intense code upon first look - it really needs a C64 expert's input. I think I'll make it a New Year's resolution to try and collaborate on some C64->A8 ports - then I might get further with some of the started projects, e.g. Bard's Tales, World Class Leaderboard. All the best, Mark

Certainly I need a push - but for now still overloaded at work - usual scenario of 'get more clients - employ no extra staff to look after them as the existing employees should be able to manage it' . In not being able to focus directly on it, I actually got distracted onto a simpler project - but that will let me experiment with some 'save/load satte to cart' routines - more details soon . Only progress made was a bit more studying of the swap-out code used when entering a building (or combat for that matter) along with a look at what a 64K machine does different to a 48K machine, which is basically some of 'walking around the city' modules are loaded into the high memory so that when you 'leave' a building they can be loaded back into lower memory instead of using the disk drive. So not much new there that's not known already. My approach is going to be to take 3 dumps of the games memory, 1 for before going into a building, 1 when in the building and 1 when having left the building. I will do this for 1) running the disks under emulation and 2) running the cart image. Then I'll compare them all against each other to fathom whats going wrong in the cart version with respect to being thrown out immediately. (If someone wants to do the dumping for me that could help - just hit F8 to get the monitor and then "s 0000 FFFF filename.mem") In August I may be able to focus a bit better as wife and kids will be away in the Czech Republic again, but that won't stop normal work still eating up my time (plus I've got a tax return to do and other 'jobs' to do around the house in their absence) Another thing I aim to do is to get the Dungeon to a similar state. I have this running the intro from a cart and so the next stage is to replace the loader routines with cart access code. I've found these but they differ (slightly) to the City's implementation and so I need to 'label' the source code in order to help me slot the replacement loader code already written for the City. Keep the faith - I'll get there one day! Best wishes, Mark

Based on that, has anyone done a closest Atari palette color match for each color in the C64 palette?

I've stripped out a PRO image and this has slightly different data in those sectors. But the loader does the same thing in an emulator, that is to stick on sector 36. I'll post those tonight when back at home. As for checking out the real thing, I'll have to get an A8 down from the loft to do that (where's the blush/embarrassed smily?)

In the words of Dangermouse... "Good grief Penfold!" - Nice work!

The disk is protected every 18th sector with phantom sectors from #36 to 180. On side one every other sector (181 onwards) is bad/missing/not formatted. Here's the ATRs anyway - I could have a go at the protection - I guess that's already been done though?! Regards, Mark spy_games.zip

I've have the ProSystem so can give it a go. One day I'll get around to buying a scanner too.

Just got around to typing this up... The guys selling the Infocom Hitchhiker's I won recently also had a box of Atari gear that he was going to list, but I offered him (GBP) £100 for the lot: Along with an 800XL and 130XE + two not working 1050's were the following originally boxed disks - overall not a bad buy I think Title By Instructions Disks Memory The Dallas Quest US Gold/Datasoft N 1 32K The Goonies US Gold/Datasoft (with hints) Y 1 48K Moon Shuttle Datasoft Inc Y 1+C 16K Mercenary Novagen N 1 48K/64K Mercenary - 2nd City Novagen Y 1 48K/64K Boulderdash II Databyte/First Star Software N 1 48K B/dash Const. Kit Databyte/First Star Software N 1 ? Starquake Bubble Bus Software Y 1 48K Flight Sim II Sublogic (with scenary) Y 1+1 48K The Institute ScreenPlay Y 1 48K The Pawn Magnetic Scrolls/Rainbird Y 2 64K The Guild of Thieves Magnetic Scrolls/Rainbird Y 2 64K Jewels of Darkness Level 9/Rainbird (lenslok) Y 1 32K Snowball Level 9 (Tape) Y C 32K Adventure Quest Level 9 (Tape) Y C 32K The Payoff Atari (big box) Y 1 48K Gemstone Warrior US Gold/SSI Y 1 ? Questron US Gold/SSI Y 1 ? HHGTTG Infocom Y 1 48K Hollywood Hijinx Infocom Y 1 48K Infidel Infocom Y 1 48K Moonmist Infocom Y 1 48K Enchanter Infocom Y 1 48K Trivial Pursuit Domark Y 2 48K Fighter Pilot Digital Integration Y 1 48K Spitfire 40 Mirrorsoft N 1 48K Hacker Activision N 1 48K? Hardball US Gold/Accolade N 1 48K Spy Demise/Strikes Back Electric Dreams N 1 ? Movie Maker Electronic Arts Y 2 ? The Moonshuttle disk and tape combined release was interesting. Regards, Mark

I experimented with just POKEs whilst BASIC was loaded and it seems that Gr.0 behaves oddly. If I set bit 3 of GPRIOR then I'd expect PF0,PF1,PM0-3,PF2,PF3 As the color registers used by Gr.0 are 709 (text brightness), 710 (background) and 712 border - I'd thought that these corresponded to PF1,PF2 and background, and that the PMs would sit between - but this must not be the case as a player merges with the text foreground and background colors? Maybe I'm missing another poke somewhere to inhibit this, but I more than think that its a no-starter . Try the attached test basic program. Regards, Mark Test.zip

Thinking about it, you only therefore get control of just the 'text' colour and not the 'background' colour, though you can still have 'inverse' text, so its not a full solution by any means [EDIT] To add to this, the actual gr.0 text color (709) would have to be set to zero (black) so that a space, due to being in inverse mode, hides (i.e. blocks) any underlying p/m's... hope this is making sense

Has anyone ever tried this with player underlays? Looking at it in theory, you could use 4 sprite sets (low-rez 1K each = 4K). Each of these would be used for base colours (red,green,blue and white). Every vblank the next p/m set is selected in order to merge these colors. Hence with 'off' giving black (b) you get these mixes: bbbb - black Rbbb - dark red bGbb - dark green RGbb - dark yellow bbBb - dark blue RbBb - dark magenta bGBb - dark cyan RGBb - grey bbbw - another grey Rbbw - red bGbw - green RGbw - yellow bbBw - blue RbBw - magenta bGBw - cyan RGBw - white Using 4 players and a combined 5 player from the missles gives 40 pixels. These can underlay a gr.0 screen, and the inverse char mode flag is set (I think) so that underlying colour appears through any given character. By using a memory area to simulate a colormap - e.g. you put a value of 0-15 at the desired offset, this can be translated into also correctly setting/clearing the appropriate squares in the p/m memory areas. If we wanted to be a little more clever, some additional bit in the high nibble of a byte in the colormap could be used to add things like blinking. Then you'd effectively have a kind-of mode to display teletext type stuff. Doubts to its success are obviously, is the color persitance going to work at that kind of rate. Just wanted to know if its been done before before I even attempt such a thing? Would be nice if I get around one day in porting the Infocom V5 interpreter to the A8 (e.g. from the C64 Border Zone game). Regards, Mark

Indeed, $400->$47F being the cassette/single density disk I/O block buffer. I find myself retaining its use in the replacement I/O routines for carts, e.g. copy correct 128 bytes from cartmem (when correct bank selected) to $400 and then swap out the cart and copy from $400 to desired destination address. Though actually easier and quicker to shift 256 byte pages around instead, this often retains compatability with the existing I/O routines found in games. I also often find that using the end of the page 1 stack (i.e. $100 upwards) is a great little holder for code snippets. If you know the stack isn't going to go below say $180 then there's half a page free! I've haven't really looked at CC65's hw-stack usage to know what's safe when writing in 'C' though. Regards, Mark

Well I came totally unprepared to my parents this week - I hadn't put IDAPro on the laptop I brought down with me and so I haven't been able to inspect the disassemblies Still, I'm now minus two wisdom teeth and still ache a little - no bruising though which was a surprise I should be back at home on Sunday so I'll tay and take a look in the evening. Regards, Mark

Thanks, I should be able to produce a mapping of what loads from where.

See: http://www.avatardesign.net/alternatereali...04/00000619.htm Anyway... guys - please try not to get this thread closed a second time. 20th Anniversary release? It is still not known at this time if a cart-port is actually technically realizable, hopefully I'll know more by next weekend. One question about this 'opening up' of previously unavailable areas. I take it that this has already been tried and tested and can be reliably confirmed that it wouldn't not introduce a bug of any kind - has there been evidence that PP had this code tested/working up to the point of release where it was pulled? If not then there's significant risk in just enabling them. Regards, Mark

http://www.atariage.com/forums/index.php?showtopic=67762 Check out also the Gauntlet ROM implementation on this Forum too. Coding to cart is fun - but I haven't really tried doing 'C' work here yet. I'd like too try it to see, but there may be some considerations that need to be looked into with the way that CC65 is handling the heap/stack etc. Has anyone else attempted this yet? Regards, Mark

Thanks Atarigrub! I had some issues loading the CAS file into A800WinPlus, so I hacked the binary into an executable (XEX) instead, taking the opportunity to turn the JMP to the next stage loader routine into an indefinite loop instead. Regards, Mark Myst_adv.zip

Hi, This was a little experiment I did to check that it was impractical to write DLIs in C. VBIs could be OK, but really you're best of using assembler for these . However, it should give you an idea of how the configuration file can be used to provide your memory blocks, segments can be defined to reside in these memory blocks. Start with 1 per block to keep things simple, but later you can use the .align statement to provide addresses on boundries. It then shows how these can be made visible to a 'C' program via a simple assembler file and finally the C file gives examples of utilising the pointers. Hope that helps, no doubt some of it could be done slightly differently. Regards, Mark vbi_dli.zip

Hmm, the only trouble I had was remembering to tick off the 'read-only' setting in the 'Atari/Settings' menu. I haven't used SpartaDos too much so it could be that.

Hi Dan, Are you writing in 'C' or in assembly? I think the standard build model for the Atari assumes that DOS is loaded, therefore all that memory is already available to you. In 'C' you'll need to define a pointer to the start of your required memory area, and in assembler you can either use a define, or alternatively you can use a customer linker configuration file to declare the blocks for you. I can probably dig up some examples of this if required. Regards, Mark

Can you post up a small example program source set and the steps you are doing and I'll work out where you're going wrong? If your app requires DOS then the 'H:' drive approach is correct, otherwise for stand-alone EXEs the 'Load Executable' on the 'File' menu should suffice. CC65 produces the correct header, however if you are including multiple sections in your binary then you'll probably have to provide your own configuration script to the linker. Regards, Mark

I hope to have an updated version ready by the weekend 16/17 July. I won't have my A8 hardware with me during most of the development so the Flash saving won't be added for that one. The main ideal will be to hopefully sort out the Enter/Leaving problems and to eliminate crashing that occurs on real hardware during the replacement I/O routines. Slow progress, work busy and enjoying good weather with family. Regards, Mark

Hi, Does anyone have (or can make) a CAS dump of one of the Brian Howath title? I have a couple of the tapes in my parents loft - but thought I'd ask before hunting around and then trying a WAV2CAS myself. All of the downloadable images I've seen have are just EXEs of the game part. With the tape versions an intro was shown with the pieces of the '8' flapping about to join up in making the 'Channel 8' title. Pretty lame, but I'd like to put it on a menu-cart with all of this series of games. Cheers, Mark