Maybe I already replied somewhere (this is from 2007 folks) but wutevs, I'm too tired to search... The console itself could be easy, you can get anything from an Atom with ION and 2GB of RAM or even the good ol' SH4, but the costs of mass distribution and marketing would be unbereable. IMO making a good portable (as in from one platform to another) enviroment for game development that isnt a pain in the ass to use (fucking flash for example) could be a nice start. Too bad I dont have the capital nor the team of software engineers to make it...
Yeah nice to see you too chum. I would like to know how you plan to released a console with little or no money. Care to clarify?
Hey, back when this dream started. There I had less money, than I have now. But a dream is also good. If the dream is creative, and getting PC parts are the least of the worries. But you on the other hand with your negativity, really give this thread a bad smell.
Shadowlayer: You are a fucking cunt, that is not the problem. That is a statement some of us agree on. But I think the whole thread was a philosophical view upon. How to make a modern 2D console. But the real world is always knocking on the door. The cheapest way, of making a modern 2D console would be, "container/dumpster diving" when it comes to PC parts, the next cheapest thing would be finding old (but cheap) PC parts. The next step , which probably would be the most expensive part is going commercial on it, and to be perfectly honest. Then yes we would not have the money to get it on rails at all. Since I calculate we need at least 100.000 US $ as a start, but it could be cheaper, and it could be more expensive than that. But your "negative reality check" all the time , and your bitterness with your console past made this thread become a bit dead smelling. That is the reason I have been after you in this thread. Because you began babbling about the SH2 processor and so forth.
I dont know about the rest of you but I'm getting tired of taking shit from crazy boy over here... WTF is your problem dk? you go around UCD all high and mighty because you're fucking mod? news flash assclown: I'm a mod too! and its not like being a member of the fucking congress, get it? your signature says "I delete, remove. And clean up. I am the Resident moderator." get a fucking grip man! why are you so frustrated in life that you resort to that kind of moronic attitude? You're completely unreasonable 99% of the time, you have a poor (at best) use of english so you fall in all kinds of misunderstandings with pretty much everybody, and instead of saying "sorry, I didn't get it" you start ranting and cursing until everybody leaves the room. With that kind of behavior you're nothing but a walking stereotype of the foreign internet troll. Seriously shithead, take a fucking chillpill, nobody is going to embark into any kind of project if you act like a crazy bitch all the time. How do you plan to deal with detractors if you cant tolerate even the most simple constructive criticism?
Still planning your dual-SH4 uber-system, eh? I pop in to see how the forum is going on and you're still afflicted by the most ridiculous delusions of grandeur. Some things never change. This is still fucking hilarious, by the way. Oh, you were so close! Your "units" were practically on store shelves! If by some heinous error in judgement you have been promoted to moderator status (it isn't reflected in your profile, but with forum software this outdated there's no saying), I daresay this post will be swiftly removed and I'll be banned posthaste. I guess we'll see.
Hi all, I thought I'd resurrect this thread because I'm working on my own hardware platform instead of doing real work at work. Right now it consists of: -100 MHz, ~30 MIPs custom unpipelined/uncached 16-bit RISC CPU with multiplier & divider -Low resolution (320x240)/high bandwidth VDP. At the moment it's got 320 up to *512x512* pixel sprites (over 1000x the screen space) but I'm trying to implement alpha blending which will reduce sprite performance from literally unlimited to just practically unlimited. Sprites are *bitmaps* in the 128 megabyte CPU space so are extremely easy to use. -Simple digital controls input -PCM audio system (haven't worked on it yet) -Very cheap hardware, it will all be implemented in 3 inexpensive parts: a FPGA, a single SDRAM and a SD slot. I can see this being a very good platform to develop for due to the lack of constraints; A SD card allows practically unlimited content; 12-bit video and no palettes mean you can draw whatever on your graphics tablet and place it wherever on screen; audio will be primarily software defined, I will try to make audio playback/mixing seamless over the SD card so the only thing you have to really worry about is the game logic; I hope to have over 20 MIPs left over after streaming data from SD and handling a good number of audio channels, more than allowing for very complex 2D/even 3D game logic. The CPU will have very few instructions, and just a handful of addressing modes so it should be very easy to program. I don't know much about encryption, but that could be implemented later on; game data could be tied to a key in the FPGA.
Would you like me to sell it for you Woz..eh... Calpis?=) Also, how would one go about coding it? With APIs being the norm for most embedded devices I am curious as to how you harness the power of your design in practical terms.
I dunno what to call this thing as yet. It seems most suited for the arcade. First comes completed specs, then emulator implementation, then I'll go back to the hardware is out of sync with the specs. I don't think anyone would jump on board until I have something to show. Well, it's a wicked simple CPU, just running very fast. I would plan to use assembly since I am not a compiler writer-- I'm sure it'll be a feat to write a reliable assembler... It is probably more practical to find a FPGA with a hardware PPC core, or use another soft core with a compiler, but I wanted to make something fully portable, and unique. If a simple design can get the job done, I don't mind reinventing the wheel. As far as API, it will have easy interfaces for loading data from SD and the audio player (if desired). Graphics are just loaded anywhere into RAM and manipulated with a sprite table. The goal is to get the platform so that the transition from creating art to using it in game is as seamless as possible. Thanks. I would like to show off what is possible with very clean, simple hardware, it doesn't have to be nearly as complex as what is described in this thread! (Well, by using low resolution and very fast RAM XD)
Calpis: it just shows, where there is a will. There is a way, and you showed one way , and punched through all the negative BS that was raining down on some of us dreamers. So again you have shown it is possible And again congratulations cause I check this thread today again, and it looks like I have mentioned you a couple of times. But anyway good luck with it, and when it is finalised with sound and all that, I think I would love to hear the specs again. Sorry if this is a stupid question, but what about ram ?
The plan is to have 64/128mb of RAM, but it's difficult to find a hand-solderable DDR chip, maybe it doesn't exist. I have no idea how I'm going to make the hardware if I have to send away for it! XD The CPU will address maybe 128mb (not decided) in order to maximize bandwidth for long addressing. All the RAM fetches are a huge bottleneck like in a normal computer so the CPU will have short (relative) addresses too in order to speed things up. Because audio will need practically no memory (can load directly off card), and the game program probably only needs a few megabytes, there will be TONS of space left over for uncompressed graphics and data tables. Just having 40mb of free space would mean having 4500x4500 pixels of graphics at a time to work with, or nearly *200* screens worth.
I m getting all sorts of funky ideas now, in terms of where your market might be at. On a technical note, would the design have a "nose boot" or is the microcode going to be fixed?
