|
|
| First NatAmi screenshot | ||
 |
||
| Previous Image | Next Image | ||
|
Description: The NatAmi prototype is running! It is running Kickstart 40.70 and is booting to WB 3.1 And as you expect I am working on the Blitter :-) Picture Stats: Views: 3437 Filesize: 53.87kB Height: 768 Width: 1024 Posted by: bluea at June 17, 2006, 04:50:40 AM Image Linking Codes
|
||
| 0 Members and 1 Guest are viewing this picture. |
| CD32Freak Posts:646 | August 19, 2006, 07:48:01 PM bluea Quote Does somebody know the rulesYes, Dave Haynie :-P ;-) |
| bluea Posts:12 
| July 10, 2006, 01:58:08 AM Quote
Yes, its quite "easy". I implemented the A4000 and A1200-Gayle registers as a Chip-Select logic for an 2.5" harddisk. So I can boot from that drive. @MAtt_H: I think its too early to compare... but it may go in that direction... (hopefully) For now I am really happy that I found 3-4 serious bugs in the blitter. So meanwhile I can read text and view icons correctly. But there are still graphic errors which I have to eliminate. There are that many basic things to figure out... for example the bitplane modulos. OS2/OS3 sets it to -4 on a OCS chipset. To 76 on ECS. And with AGA sometimes to 72 and sometimes to 76... But with FMODE ist still 0. Does somebody know the rules? :idea: |
| Dennis Posts:128
| June 25, 2006, 07:34:00 PM Quote The Blitter block portion is done but in line mode only the start-point is drawn. I think Blitter pointers and modulos may be ok, but some misunderstanding in ohter registers... lets see. Line mode is indeed very weird. From what i've understood from the "historical documents", it was really an afterthought by the designers, hence the weird implementation. Took me a month to get it all working properly. Quote I am booting from Harddisk. Floppy unfortunately gives too much zeroes and 0xffs in MFM data How do you do that? have you implemented gayle-like ide port? Quote Is the blitter really used to decode MFM data? - Im just wondering because I can not detect any write to blitter registers while (or shortly after) a track is read from disk On kickstart 1.2, yes. Don't know about KS3.1 though. In Minimig I couldn't get past the "hand" screen without a proper working blitter. So before I started on floppy, I already had a working blitter. Dennis |
| bluea Posts:12
| June 25, 2006, 05:50:46 PM Hello Dennis! Im glad getting congrats from you! Yes, 6-8 planes should be possible, but only 3 are tested. Copper is running well, only 2 sprites because of FPGA capacity. I fixed a bug in the CIAs yesterday, so keyboard is working also. Now I can write true amiga testprograms with user input. The Blitter block portion is done but in line mode only the start-point is drawn. I think Blitter pointers and modulos may be ok, but some misunderstanding in ohter registers... lets see. No, the c-one FPGAs are not really big enough. I exchanged the "small" 1k30 to an 1k100 like the other one. But now both of them are now used up to 95-98%. I am booting from Harddisk. Floppy unfortunately gives too much zeroes and 0xffs in MFM data :-( Another todo... Is the blitter really used to decode MFM data? - Im just wondering because I can not detect any write to blitter registers while (or shortly after) a track is read from disk |
| Dennis Posts:128
| June 19, 2006, 06:42:15 PM Hi Thomas, Congratulations! and take a :pint: . I know what you've been through :-) . Looking at your screenshot I suspect you've got bitplanes, copper, sprites, cia's and parts of blitter working? Also, are the fpga's on the C-one big enough? Are you loading WB3.1 through floppy? If so, then you're blitter is at least decoding mfm data correctly! Anyway, very nice work. Dennis |
| cv643d Posts:1197 | June 19, 2006, 02:57:09 PM Wow.. cool. Time to buy a 060 chip again |
| icbrkr Posts:411 | June 18, 2006, 12:30:12 AM As the owner of a C-one, I'd love to know how you're doing this  |
| Matt_H Posts:6412 | June 17, 2006, 11:51:38 PM So more like the BoXeR would have been? Sorry to be comparing your project to others, I'm just trying to understand this. It looks very exciting. Quote 16 or 32 MB SRAM for chip.memory. I know a few people here at Amiga.org will be very pleased to hear this ;-) |
| bluea Posts:12
| June 17, 2006, 07:26:57 PM hmm... none of both. Now it is running on the c-one because it is a "nice" platform. I had to do some ;-) modifications on the c-one board but now its possible to run parts of an Amiga on it. It is my test-envoronment now. The next NatAmi version will run on an own board containing an MC68060 with SDRAM as fast.memory, 16 or 32 MB SRAM for chip.memory. Lets see, if I have more time I upload more information to the website. But at the time I am doing all on my own. (the new HW design, working on the chipset, the website, ...) |
| Matt_H Posts:6412 | June 17, 2006, 05:02:07 PM Impressive. Is this project to be a new attempt at something like the Inside Out? Or is it a new module exclusively for the C-One? |