@Dandy

Quote:

Hmmmm - the A500s all had a "baseline" "Expansion Port" and a "Trapdoor Port". Also pretty useless from a users perspective without "adding a decidedly non-baseline piece of hardware to it", me thinks...

How about the clock port of A1200's? It was a very simple interface which allowed a lot of neat hardware to be developed without the added complexity and higher costs involved in developing full-blown Zorro cards. The possibilities with the Xena/Xorro system are huge in comparison, and where you previously needed complex and expensive electronics, you hardly need anything because the Xena can perform the same function. All you need is to hook up some Xorro lines to some I/O ports on the card, possibly with some DAC/ADC, program the Xena and off you go. Minimal electronics + minimal development costs = cheap hardware solutions.

@COBRA

I don't think a subway is anyless complicated as a highway was (except for the fact that the highway had a clockport itself plus the connector for the norway).

Clockport devices were created for only one reason, it was the only way of targeting the non-towerized-A1200-market with anything that was not an accel .....

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

The Xena/Xorro thingi needs an extra card to come to any use. And since you are allready designing that extra card, why not just throw that 5$ microcontroller onto a PCI-card and be compatible with all OS4-compatible (and non OS4-compatible) HW ?

Microcontrollers don't just get "thrown" onto PCI cards... You need PCI interface circuitry to interface the microcontroller with the bus. It also means that you would not have the JTAG connection to the CPU, so it wouldn't be in-circuit programmable. Unless you add some clever programming interface as well to the board, which can be controlled through the PCI bus (even more cost). Also, Xorro boards are not designed by A-EON, they will be designed by talented enthusiasts who want to harvest the capabilities of the Xena/Xorro system, possibly making some cool hardware addons such as Amiga joystick interface cards and the like. Such things with Xorro/Xena can be made in a fraction of the time of a PCI card with minimal development costs and thus much reduced cost per board when released. The potential benefits outweigh that $5 per board cost by far.

@Kronos

Quote:

I don't think a subway is anyless complicated as a highway was

You don't think? You say that based on what information? Have you designed any hardware for Zorro slots and/or Amiga clock ports? Or have you spoken to people who have? Do you know the differences between the clock port interface and the Zorro expansion bus?

@COBRA

Regarding Zorro-slots (and were are talking Z2 which is allready many times faster than the clockport), it's well documted, there are plenty of DIY-projects found on Aminet and elsewhere so it's clearly no voodoo involved.

Bout Xorro:

The only way one could do Xorro-card is either o have lot's of equiqument handy or to order it at a PCB-service (which can be quite costly for a single proto). PCI-proto board (including basic decoding locig) can be had for about 200$ in single quantities and the resulting HW would be directly accesable from the CPU as a memory-mapped device. If the XMOS chip can't be programmed without that JTAG connection, well add that or just use another better microcontroller ...

HW-developed that way can be used in allmost every computer, and computers having more than 1 PCI-slot can utilize more than 1 piece of HW developed that way.

Last edited by Kronos on 15-Mar-2010 at 10:55 AM.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

Quote:

Hmmmm - the A500s all had a "baseline" "Expansion Port" and a "Trapdoor Port". Also pretty useless from a users perspective without "adding a decidedly non-baseline piece of hardware to it", me thinks...

The difference is that XMOS is clearly advertised as a custom chip and we have yet to hear from a use as a custom chip. There is no need to argue until a proper use is presented...

_________________
http://www.warpdesign.fr/

it's a "customisable" chip!

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

Indeed. I don't know what source Leo thinks he's using for the "advertised as a custom chip", cause everything I've seen so far has been talking about the XENA as a "customisable" chip, specifically mentioning it in the context of Amiga chipset only being "custom" chips.

I think it's a great idea, we just need to utilise it.

I'm extremely eager to try out some ideas I have on this chip.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@Kronos

The big difference between Zorro bus and clock port is that the clock port is simply mapped into a fixed address which can be accessed so you basically can just wire pretty much any logic to it directly and get away with it. With a Zorro card you have to support the autoconfig protocol, your hardware has to be able to accept dynamically assigned address, store autoconfig info, init code, etc. There are also much stricter timing requirements for a zorro card, because there can be several cards on the bus and they share the bus, so although it's not voodoo magic, it is considerably more effort to develop zorro cards. I still remember our conversations with Klaus Burkert (designer of the Picasso IV card) who told me about all the problems he has come across with the zorro bus (hidden bugs of the buster chips, etc.).

Regarding Xorro vs PCI, what you say doesn't make sense. You have to get a PCB manufactured in any case. PCB manufacturing cost wouldn't be any different. What would be different is:

1. Designing PCI bus interface requires additional development time and cost
2. Additional components required for the PCI interface lead to increased board cost
3. Additional board complexity with a PCI board means higher design risks (more chances of faults)

Also, designing a full-blown PCI interface may be beyond the skills of many potential customers who just want to implement simple control applications.

Quote:

well add that or just use another better microcontroller ...

JTAG is the industry standard interface for programming and debugging microcontrollers, so any "better" microcontroller you may want to use will also have the JTAG interface (and thus the same issue of having to implement some kind of JTAG driver logic on the PCI card that is controllable via the PCI bus).

So I think now you surely see that having that $5 XMOS chip on that X1000 which is expected to cost in excess of $1000 is a very elegant way of adding interesting and unique capabilities to the hardware without having any significant effect on the overall cost of the system.

I hesitate to join what looks more like a lynch mob than a discussion..

Whether we can call the X-Mos chip "custom" is an interesting discussion. Being software configurable makes it (arguably) more "custom" than any of the original Amiga chips.. because the programmer of the moment can make the chip whatever he wishes (within the limits of the chip).
While most of the readers here will be arguing about what it cannot do, some of us will be having a lot of fun figuring out what it might be capable of.

An earlier discussion concluded that the chip cannot be useful without an additional plug-in card. I disagree. There will be applications that use the chip completely internally, without any need for outside connections. One obvious possibility would be for commercial software to verify that it is running on a real, non-emulated Amiga system. The very presence of a X-Mos chip on the bus could be verified, and might be quite difficult to emulate in software, given the extreme flexibility of the XMos chip.

OK, so that's not the "people pleasing" killer app, but for a software publisher that is considering writing for the new Amiga, it could well enable better apps than we would otherwise get. and it would be useless unless it was included on EVERY new Amiga.

I'm going back to my quiet place now,
LyleHaze

_________________
question=(2b||!(2b))

The only thing we can almost guarantee a Xorro board will contain will be more XMOS chips.

I am certain there is some potential industrial use for a whole swag of them working together. Although I am at a loss as to what application exactly would benefit from that particular type of processing. Perhaps some sort of inexpensive signaling technology is called for in some market. Surely one of the prime reasons to have one of those XMOS chips on board is merely for the processor to directly talk to such a Xorro card, and relay instructions and data to one or more multi-chip expansion cards. You could argue XMOS chips could do all that without any desktop motherboard being involved, but for particular higher-end purposes that may seem a little mickey mouse (ie. amateurish).

Edit: Interesting videos discussing these possibilities. http://www.xmos.com/videos

Last edited by BigBentheAussie on 15-Mar-2010 at 12:19 PM.

_________________
Leo Nigro, CTO Commodore USA, LLC
Opinions expressed are my own and not those of C= USA.
Commodore/AMIGA "Beautiful, High-Performance, Home Computers for Creativity and Entertainment."

@COBRA

One can build a Zorro-card to fixed address (some of the early HD solution did that) wouldn't be anymore "evil" than trying to run more than 1 clockport-device at the same time (some allow to be mapped to another address, others not). Autoconfig is something you could add later on if and when you consider making it more than a proto just for your own personal use.

The Picasso4 is a (autosensing) Z2/Z3-card and I think we can all aggree that Z3 is a royal pain in the a....

You don't have to get a PCB made to start experimenting/developing with PCI, proto boards can be bought ready-made including the full bus-interface. Sure 200$ and upwards isn't fun, but try getting a PCB with edge-connector made in single quantities and your not that far off.

JTAG or no JTAG, I'm sure finding a proto-board either for PCI or USB for any given microcontroller ins't that big of a hurdle.
And since a PCI-card can be controlled directly by the CPU there won't even be a need for any extra microcontroller in most cases.

@lylehaze

Actually being fully programmable and an of the shelf part makes the Xena the direct opposite of what the words "custom chip" are usually meant to mean.

But I must applaude you !! You've managed to to find the most stupid and outright retarted use for Xena. Not would such a developer cut himself of from potential sales to "emulated Amigas", no he would also cut himself of from customers useing OS4 on other well supported HW.

And just to make something clear (as some people have a tendency to go fishing for insults), I didn't call lylehaze stupid or retarted, I just think he had a brainfart he should have better kept to himself.

On the other side, a developer following his adavice ....... well lets just say some people should never ever touch anything even remotly connected to SW-developing

Last edited by Kronos on 15-Mar-2010 at 12:54 PM.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@COBRA

Quote:

Also, Xorro boards are not designed by A-EON, they will be designed by talented enthusiasts who want to harvest the capabilities of the Xena/Xorro system, possibly making some cool hardware addons such as Amiga joystick interface cards and the like.

Too bad it is already done. Maybe you prefer using A1200 keyboard? No problem, Keyrah is for you. The geek port already exists -- it is called USB.

_________________
Amiga Developer
Amiga 500, Efika, Mac Mini and PowerBook

@itix
The sentence from Cobra you quoted begins with "also" which is the perfect word for Xena. It opens a lot of possibilities depending on the programming and configurations (wheather or not you are using the one on board or in concert with additional chips, even 2nd generation ones that might populate future cards). Trying to relegate it at a single use or possibility and then say there are other things that do it it's not the best way of analyzing the matter...

@Kronos
Your wording is highly offending nonetheless, "not you yourself but what you say is retarded", "you had a brainfart keep it to yourself", oh, but I'm not offending anyone of course!"

_________________
SamFlex Complete 800Mhz System + AmigaOS 4.1 Update 4
Amiga 2000 DKB 2MB ChipRam GVP G-Force040 Picasso 2 OS3.9 BB2
AmigaCD 32

@itix

Quote:

already done

Yup! I agree. But, what about a game controller with 32+ different output
sensors that could work/feel different, depending on what the game is doing,
because it has a number of input signals that can modify it's behavior.

And then, , , different games requiring that the same controls be used in
a different fashion only need a different Xena (software controlled silicon)
program.

(Just for grins that is.)

@itix
So it could access more than the 64kb of ram?. If this is the case, the xmos chip is interesting and could be used for many applications/amiga things.

_________________
Amiga 1200 desktop. Apollo 030/50 Mhz 8mb ram + ClassicWB + Wb 3.1
Amiga 500 + ACA500plus 8mb + 30gb CF
Raspberry Pi 3b+ and Amibian 1.4
Mac Mini G4 1GB Ram with the butterfly!!

@Kronos

Quote:

lylehaze said: One obvious possibility would be for commercial software to verify that it is running on a real, non-emulated Amiga system.

Then,
Quote:

Kronos said: he would also cut himself of from customers useing OS4 on other well supported HW.

Sounds like a usable idea to me. If A-eon, let's say, were to spend big in time
and money developing an application to be used on/with their HW would they really
want it to be used elsewhere.?

Oh! At least HIS brain isn't constipated.

-----------------------------
lylehaze! I was beginning to think you fell off the edge. Good your still here.

@Mechanic

Want to identify HW ? Use Mac-addresses, specific drivers (is sys: mounted on mystersocSATA.device).

Want a game controller with 150 buttons ? Force-feedback ? Maybe even an LCD-display ? Well go to local electronics retail shop and just buy one ....

Heck it will even come with a plug that lets you connect it to every computer produced in the last 10 years or so (wether you'll find a driver is offcourse a different story).

Input devices, costum or standard should be USB these days, everything else is just nonsense....

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

But I must applaude you !! You've managed to to find the most stupid and outright retarted use for Xena. Not would such a developer cut himself of from potential sales to "emulated Amigas", no he would also cut himself of from customers useing OS4 on other well supported HW. And just to make something clear (as some people have a tendency to go fishing for insults), I didn't call lylehaze stupid or retarted, I just think he had a brainfart he should have better kept to himself. On the other side, a developer following his adavice ....... well lets just say some people should never ever touch anything even remotly connected to SW-developing

Wow.. At first I wasn't sure it was worth replying, but if I can get a response like that from Kronos, that really does make the reply worthwhile!

Oh, and not to worry, I've never even been tarted the first time. :)

But I must decline your award for most stupid use of the Xena chip. While I am sure that you've had a lot of experience judging others, I do not recognize you as an authority on what is best for the Amiga.

Have a nice day,
LyleHaze

_________________
question=(2b||!(2b))

@ thread

While this thread is quite interesting to read, let's keep the personal digs to ourselves please. Nothing bad so far, but let's keep it that way.

Carry on...

_________________
Moxee
AmigaOne X1000
AmigaOne XE G4
I'd agree with you, but then we'd both be wrong.