@Cod3r
Quote:

Regarding Aros, I am glad you said it and not me, because I didn't want to offend anyone's work, but from a coding standpoint... ouch. A hodgepodge of code. It isn't pretty. I'm surprised that it works as well as it does. I spent four months just improving them enough to be more compatible!

I absolutely do not understand the last sentence.

Why would a seemingly sane person spend 4 months editing AROS code "to be more compatible"? To be more compatible to what? Why?

Sincerely,
Baffled

_________________
Wanna try a wonderfull magical Amiga strategy game?
Total Chaos AGA

@ChaosLord

obviously aros 68k tries to be compatible to 1.x-3.x range of amiga kickstarts/workbenches or amiga os, whatever you name it, and existing 68k apps. and it still has issues here and there.

@Cool_amigaN

Thank you for the kind words. I am not discouraged, since I already spent nearly 9 months of my own time on doing it, I won't stop just because someone else doesn't want it.

I have read about the Minimig, but it really isn't a full replacement/upgrade because it lacks many of the features for it to connect to the outside world i.e. printers, ethernet, etc. My idea is to add stuff like that on to the project, since the target hardware already has USB, serial and ethernet ports.

If I could convince my company that this would be worth something, then I would have the talents of guys of co-workers who worked with Motorola and IBM to draw something up. I really don't understand why some people are skeptical that this can be done, because it can, but the guys who can do it are usually very busy and well compensated to "more important" work. Trust me, those companies didn't hire incompetent engineers.

I myself, am just a coder. I can come up with some great code, but i'm not a hardware designer, that's why I build upon what I potentially can use as a platform. I'm not a ColdFire, PPC or any other fan. To me, a chip is a chip. It's capabilities are what make or break my interest to code for it. I like MIPS code, but I wouldn't use an ancient 80's MIPS processor to emulate an Amiga's CPU... would work, but be unacceptably slow.

Thanks for the input, I really appreciate it. When I ask the boss tomorrow about my project and if he'll allow me to show what it can do in pics or a video, i'll go from there. An official endorsement from my employer would really enable me to do more with it.

Quote:

Cool_amigaN wrote: @Cod3r First of all, congratulations for your achievement! Don't be fooled by the responses from this website since most ppl here are pro NG users, the classic amiga scene (68k) is far wider than the PPC one. There have been successful attempts to recreate a classic amiga environment; Minimig is a FPGA re-implementation of A500. Clone - A is an similar concept (FPGA - A500) that is still being worked on. A new FPGA project was recently completed that re-implemented an A1200. Natami is a project that not only "mimics" a A1200 but also opens new hardware paths by expanding the capabilities (e.g. faster type of memory). Note that by "re-implementing" it is meant not only being 68k compatble but also re-creating the custom chipsets (sound and video) of the classic amiga line. The coldfire platform on the other hand is a dead end since it is common knowledge that it isn't compatible enough (68k wise) and Workbench (classic AOS) will crush. Still your code can be helpful to various projects (e.g. Natami) mentioned above.

@trans

Great ideas. On the programming side of the coin, i'm confident in my abilities, but I have no HW design skills to shrink anything. I know a bit about CPU hardware, but I only know enough to help me better understand how to code for low-level hardware solutions.

I am really just a coder and would need help to design HW.

Quote:

trans wrote: @Cod3r You know, if you could get a system ~99.99% compatible with original Amigas and cost wasn't prohibitive then certainly some people will be interested. But will it be enough to make it worth it? I don't know. I suppose if you can produce them on demand at very low overhead then its possible. Now if you could take that same design and shrink it all down to a single SoC... Now that would be AWESOME, and I think just about every Amiga enthusiast would be interested. But I have no idea how possible such a thing is. Is the chip market evolved enough that you can order custom chip batch productions more-or-less on demand?

@Samurai_Crow

Thanks for your input, I will keep you posted on my work and I am willing to collaborate on Natami or other projects. I will be in touch!

@phoenixkonsole

Why of course! I'll send you the invoice in your inbox. Pay a 50% deposit up front (only $500,000.00 USD) and i'll have it shipped in three months. Deal?

Quote:

phoenixkonsole wrote: HA!! Wait.. your company creates 68k based stuff? Ok here we go: I would like an Ares CL AmigaClone running AROS. Where do I need to subscribe the contract ; )

@billt

I'm sorry that I can't answer that question, as I don't know ANYTHING about FPGA-Arcade. I just looked at the webpage, but it sounds like they have already done some similar work on what i'm trying to do. But without knowing more about the project, I sadly cannot make a comparison.

The prototype has USB, ethernet, ide, serial, parallel. The ide is tied to a compact flash connector, though the traces are on the board for a 40-pin ide connector. It's about the size of a CD, but in the shape of a square.

Quote:

billt wrote: @Cod3r i think anything is interesting to hear about. how does your project compare to fpga-arcade in terms of features and performance, as that seems to be what any other 68k amiga-alike will be competing with. compatibility with system friendly stuff should be ok with a p96 driver etc. like draco was. if want more compatibility then put in an fpga for minimig stuff. what features does your board have? pci slots? ide or scsi? ps2 or usb? etc

@Arko

Could be. I haven't seen the other implementation at all, so I can't say. But don't you agree that the performance depends on who implemented the design?

For example, I fixed up the Aros code and optimized parts for my target hardware. My version runs 5-10 times faster in some routines than the stock source code I downloaded.

Did the source code I downloaded work? Yes. Was it optimized by a software engineer who is a specialist in low-level system code? No way. Not even close. If I tried harder, I could squeeze out even better performance.

My point is that ten engineers of varying expertise can all work on a solution separately, and you'd likely have have ten different levels of performance in what was produced.

Quote:

Arko wrote: Amiga fans where fascinated by the Coldfire because this CPU was marketed by Freescale as 68k compatible CPU. If the Coldfire has to use emulation technics (Interpreter or JIT) it will not give them a 68k CPU with much higher speeds than a 68060 could do. A Coldfire accellerator for the Amiga was already demonstrated and it was disapointing.

@pavlor

Quote:

You read it wrong. It was Amiga Inc. who opposed this trademark (and terminated its opposition after settlement agreement with Cloanto).

Nope. That's exactly what I meant -- That Amiga, Inc. doesn't posses the trademark. But now you worry me a bit. I assumed Cloanto held the trademark to protect it for the Amiga community, not to be a trademark troll like Amiga. Please tell me I am not wrong about this!

@olegil

I am totally starting to see that. Though most responses have been quite supportive and positive, some individuals here seem quite cynical. Like if I am going to waste my time (and theirs) posting about a non-existent idea for a market that is itself practically non-existent.

I never claimed to be a savior for Amiga or a hardware wizard. I'm just a good programmer who wasn't old enough to experience the Amiga the first time, and I like the concept.

But I also have no prejudice or preferences in what chip does what, PPC or 68k or anything else. I just make stuff work. Period. I could care less if my implementation was running on ARM, x86, Blackfin, etc. I'm CPU agnostic.

Regarding my employer, you are absolutely right in what you say. The industrial boards we make are sold for quite a bit more than the cost of the individual pieces of hardware, but the clients pay for it happily because they know the reputation of the company is solid. We don't ship half-working, project-level junk. Our systems are designed to work for 10-20 years without maintenance.

Quote:

olegil wrote: @Cod3r Let's just say it once and for all: If you make a profit, you'll be demonized. If you don't you'll tank and people will think you made a profit and ran for it. Companies are never gonna win in this "community". Now, if you have such a mature product and you KNOW it's gonna work, then maybe. I bet your boss is used to industrial markups (cost x 4 before special high-volume or return-sale deals). He won't be seeing those in this market, as people here will quite happily count the cents on the components on the board and figure it out. Not even figuring out that they're also paying for assembly, testing and R&D.

@Nameless

I have no idea what the hardware would be sold for. The boards themselves have been developed as replacements for aging hardware that is used in manufacturing. They had certain requirements, and my company built around them, while maintaining compatibility with their custom software.

Not sure of what it all would be sold for, i'm not a businessman, and have never sold anything but my talent doing freelance work. Sorry, I have no real answer.

Regarding Atari, maybe so. I have research their ColdFire project, and honestly, they seem pretty on the ball, but the Atari platform is pretty easy to implement. As a programmer, it isn't really interesting to me. I do like a challenge...

Quote:

Nameless wrote: @Cod3r Interesting project. Whether it is something worth pursuing as a commercial project depends on a couple of factors. What sort of pricing are you talking about for a final product? For a niche, side-product, maybe sell 100-300 boards type of thing, you can maybe get by if it's a couple of hundred dollars using AROS. People can bring their own OS/kickstart, if needed, as most here will own real Amigas. If you want a big commercial project, 10s of thousands of sales, then you'd need to get costs down... a lot. Into OUYA range or less. And also get AROS working perfectly, or figure out some licensing deal to get WB/kickstart. Also keep in mind you can get Atari people interested too. Although I think they already have their own coldfire project going on/completed? I lost track of that.

@Cod3r

What you write sounds very interesting.

Regarding Aros 68k... Aros was at the started creating to run on fast hardware (X86) so optimization was not a real issue, much more important was that it "works". So every help would be appreciated to make it more useful on slower hardware.

Regarding Kickstart/OS you have two choices, using the original stuff (legal issues and no source code) or Aros 68k with Kickstart Replacement.

A lot of stuff already works. I have created a distribution for it (Aros Vision).

Description is here:
http://www.natami-news.de/html/aros_vision.html

Compatibility page based on Aros 68k and WinUAE:
http://www.natami-news.de/html/compatibility.html

Some screenshots:
http://www.natami-news.de/html/screenshots.html

At the moment all big Aros-distributions are integrating Aros 68k for 68k support out of the box (f.e. Icaros and Broadway X) so it means thousands of potential users and I expect also adaptions to f.e. FPGA Arcade. So it will be best bet for 68k in future.

@terminills

I'm going to leave it at this: I don't have to release ANYTHING to you to prove my abilities. How dare you try to pressure me into giving away 9 months of my work because "you don't believe it"? If you read the thread, you'd see that I am going to ask my employer if it is okay for me to release pics or video of my work running on the board.

How would you like it if I said, "I don't believe you have a girlfriend/wife, post nude pics of her or she doesn't exist"? That would be quite juvenile, and insulting wouldn't it?

Obviously, I opened a dialog to see if I can help the community, not get into arguments about my abilities. Comments like yours will run away any decent talent that comes through here. Reading posts like yours on here for months is EXACTLY why I never joined before. I simply don't have the patience and time to deal with it.

People who spend months and months of their own time at no cost developing something shouldn't be subjected to inquisition. I never said I had the best solution available. You can't believe that someone could optimize botched-up Aros code? Really?!? If it was so solid, it would just work. Wouldn't need improvement.

I have a great resume. If you have a job offer paying 80k or more a year, i'll gladly email/fax you my resume for consideration. Have passport, will travel.

Quote:

terminills wrote: @Franko to quote. "Code patches or it didn't happen" ;)

@Cod3r


Difference is I'm not claming anything you are. ;) How does the old saying go "those who can't talk about it those who can do" ;)


PS: you do have to follow the APL btw. ;)

Last edited by terminills on 15-Jul-2012 at 03:16 PM.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@Cod3r

"botched-up Aros code"

others also work in their spare-time for it so please stay politely (as you expect of others)

@Pleng

I was my incorrect reference to call the Amiga-compatible OS the name Workbench. It is an error, nothing more. I don't want to use any proprietary trademarks or names, please understand me. I could care less if is called Wanderer, Workbench, BenchTop or anything else.

Again I state for the record-I am not interested in names, only functionality. You can name my product for all I care. That's how much I don't care about the name.

I just referred to the OS as Workbench because I thought that's what it was called. Like some people refer to "Ubuntu" as "Linux", instead of "a trademarked Debian-derived GNU/Linux distribution".

I'm 25. Never had an Amiga in the past. I bought a secondhand A500 for compatibility testing five months ago. I have not reverse-engineered anything. I read some Sams books I bought on eBay.

I have Aros running on a ColdFire prototype board through low-level emulation. Aros doesn't even know it is not running on a 68k processor. And I increased the compatibilty and speed by replacing some system level routines with my own code. Some programs that didn't run on 68k Aros now run on a ColdFire prototype board. Most games that I tried don't work or crash. Some non-game software written for the standard Amiga OS does work.

I'm not sure if any of this is interesting or not. Could be or not. I haven't read about anyone else doing it, but it is truly not impossible enough for anyone else not to have done it before. I admit, I could be re-inventing the wheel here.

Quote:

Pleng wrote: @Cod3r In your original reply to me you stated that, even thought you're not interested in the Amiga name, you don't see why it would be a problem and you cited the "IBM Compatible" case. I suggest you look up the history of that as it was based on reverse engineering of a BIOS which enabled people to make the "compatible" claim. Secondly, you have a lot of people here telling you to ignore the naysayers, claiming that their answers are fulled by a bias towards PPC, and telling you that there is a much bigger market potential for Classic hardware. Of course, these posts are coming from classic enthusiasts, so you might well need to understand that there's potentially just as much bias in these posts. I stand neutral - I don't have any classic or NG hardware, and haven't really had anything to do with the Amiga in years. I just pop in here out of interest when I have too much free time on my hand. I have no idea which propaganda is more accurate I'm afraid. All I know is that all Amiga markets are small markets. Thirdly, you started off by saying that you got "Workbench" running on your system, but then admitted that what you have running is Aros. Aros already runs on many platforms, so I'm not really sure how special what you've achieved is? Does it run any Amiga system friendly software natively (which is the point of Aros M68k)? Or does it only run software which has been compiled for Aros? If it's the latter then, really, I don't see anything special in what you've got - There are already cheap x86 and ARM architectures that run Aros quite nicely. A coldfire Aros board that can't run any classic software doesn't seem to be anything to get terribly excited about, unless I'm missing something?

@Cod3r
Quote:

I was my incorrect reference to call the Amiga-compatible OS the name Workbench. It is an error, nothing more.

Don't worry about it dewd I refer to my windoze xp desktop as "Workbench" all the time.

In fact, if everyone would refer to their desktop as Workbench all the time then it would be ruled a "generic term" and it would lose its Trademark protection.


Quote:

You can name my product for all I care. That's how much I don't care about the name.

Ok, I will vote for BorkWench

_________________
Wanna try a wonderfull magical Amiga strategy game?
Total Chaos AGA

@Cod3r

back to topic..
How do you support us and how should we support you?
If you have a design tell us the price and i tell you a price for the needed drivers and software.

Deal? You show what you have, we show what we have.

_________________
AROS Broadway - AEROS - Aminux - AmiCloud - indieGO! Appstore - AmiWallet - VAN lossless video codec - AMC Amiga media Center -KrypUnite - LibertyNet - MinX - amigaNX

@itix

Exact timing is easy. You need a programmable system clock. Sync everything to it, change it with a callable routine. Base the enhanced system on a multiple of the classic's hardware clock. Of course, video sync and port timing has to be considered, but as a programmer, you know what I mean, just make sure you keep the correct timing in place for those by choosing an even multiple, if possible. And the code checks to make sure that everything will scale to the proper values.

Sure, you may not squeeze that absolute fastest performance out of the hardware, but it will close enough to the limit and be much more compatible with the legacy emulation.

Of course, I don't and wouldn't use C (or any other high level language) for exact timing. Asm is my best friend, Forth if I absolutely need some portability, but I tweak the object code anyway. Screw portability when you need compatibility. That's the principal.

Quote:

itix wrote: @Cod3r Quote: Are you a programmer? How can you say something like that without knowing my level of expertise? Is it because you can't do it? With all due respect, you would have to be a bit misinformed to believe a professional, low-level systems programmer couldn't emulate a 1985 desktop computer in 2012 with a more powerful hardware platform. You can emulate custom chips, it is already done (WinUAE), but dont expect it works. Games rely on exact behaviour like timing, response times, undocumented features or HW bugs. You have to go through lot of games to verify custom chip emulation works and when it doesnt work you have to figure out why it doesnt work. I am a programmer and I havent ever done anything complex like that but I know how complex Amiga custom chips are. It takes lot of effort to emulate custom chips properly. Quote: In regards to the CPU clock, you know you could always drop down the speed for compatibility with troublesome software. Simple to do... with a timing routine. It is probably enough.

@amigadave

I won't even deny the fact that I myself would rather have a Natami than what I have at this point. It really does seem like the most interesting project out there.

In regards to a faster ColdFire chip, I assume it could be possible to switch to a higher speed CPU on the board, i'm not sure though, as i'm not the designer of the hardware. But I would assume it is possible.

Quote:

amigadave wrote: @Cod3r, I have to agree with jingof that the Natami is one of the most interesting Amiga projects ever to be proposed and there are hundreds, if not thousands of Amiga enthusiasts that are hoping and waiting for it to become reality and be produced and offered for sale to the public. If you can be of any assistance toward getting the Natami finished and into production by offering your coding skills, you will become an instant hero in the Amiga community (anyone who helps the Natami become available for me to purchase will be my personal hero, right next to Trevor Dickinson). Your ColdFire board might be interesting, but I doubt it would ever sell in large quantities to the Amiga community, unless it can surpass the capabilities of the existing MiniMig by a wide margin, while providing near perfect Amiga emulation and providing Ethernet, USB2, SD Card and/or Compact Flash storage & more RAM. The FPGA Arcade Replay board with it's proposed daughter card will soon provide 75MHz to 100MHz 68060 CPU speed with all of the just mentioned expansion ports and extra RAM, so if you can't provide your Cold Fire project at the same or less price, or better performance than the FPGA Arcade Replay board and it's daughter card with a real 68060 onboard at up to 100MHz, it might not be worth pursuing it further, unless you only want to do it for fun and are not doing the project for the possibility of a small financial gain with sales. Is it possible to replace the currently installed 200MHz Cold Fire CPU with the 500MHz Cold Fire previously mentioned in this thread? If not too expensive, this might distinguish your project from the proposed FPGA Arcade Replay w/daughter card & 68060.