Commander X16

21 hours ago, x16tial said:

I'll start my answer with a question:  Why is the Phase 3 X16 even needed?  Or Even Phase 2?



...The X16 is your Model 3, the workaday, every man's model.  The more comfortable, and more attainable version, with a lot more practicality.



Keep it open (as possible), and hackable, and available.  Let people figure out how to case it, customize it, do whatever with it.

Oh, and ship it with 2megs RAM, do your best to make sure everyone has the same platform.  RAM "Upgradeability" for this system isn't a good idea, in my strong opinion.



What's your Model S?  Do you need a Model S?  I don't really think you do.

The X16p is the model S. All of that open hackability with kit buildable through pin chips, paying a premium price for either a kit or even more premium price for a professionally built kit computer is not the Model 3. The X16c is the Model 3, the workaday, every man's model.

The CX16e and the X8 are alternatives. One or the other, not both. And since the X8 can be available when the release X16p and X16c are available, with all three having distinct support tiers and minimum support levels to launch those tiers, then the crowdfunding can decide which ones get released, and there is no Osborne effect.

_______________

I'm kinda bummed people are voting to have no phase 2.  It's honestly what I would probably get, since it's still customizable (you can add RAM, etc.) but it will be smaller and cheaper than phase 1 (if I'm understanding correctly).  I'm not all that interested in soldering a phase 1 myself, or spending a lot of money to have it soldered, and the phase 3 seems kind of boring to me, like others said, you might as well just use an emulator with a RPi.

I don't think the no votes ought to be taken too seriously ... leaving it without a "I don't care either way" option means it's strongly biased to give mostly option 2 responses from those who don't want to buy the CX16c, even if people were instructed to leave if blank if they didn't care. The only real important information there is what share are in the CX16c market, and whether the other responses say "no, DON'T give those people what THEY want!!!" or "I don't care" is not the basis for a sensible market decision. You design products and introduce them into the market because they appeal to the prospective buyers, and if that is enough to support the product, that's all that matters, no matter how much they fail to appeal to those who are not in the market for them.

13 hours ago, VincentF said:

As of reading your comment, @Scott Robison, I'm thinking a bit more objectively on the problem of X8 vs X16 ?



So, we have two platforms with some differences in hardware. It would be effectively ideal to have those two having the same interface so the work to convert software is not that complicated.

Everybody's talking about having the X8 but with the X16's VERA interface, but why not doing the inverse of that, making the X16 actually using the X8's VERA interface ? Or a mix of the two, to lift a bit of constraints on the final hardware.

First and foremost, because the Vera FPGA literally does not have enough pins to handle that kind of interface from an external system bus. When they went from the 8 register to 32 register bus interface, that requires two more address pins, and they literally had to take out the Vera serial interface because there were not enough pins to support both, so they opted to make the video/audio access more functional and the built-in serial port option is bit banging on a VIA.

Secondly, because it complicates the logic of the chip select circuitry on the board, and they might not have enough TIME to complete more complicated chip select logic and also run at 8MHz.

With the all-FPGA design, neither is an issue, since for the first, it's already inside the FPGA so doesn't use up any pins, and for the second, there really isn't any chip select circuitry when there aren't any distinct chips to select. It's just a matter of whether there is one cycle in each four in the 50MHz VGA generation process that the Vera is not accessing the embedded 128K RAM so that the circuitry simulating the 6502 side can access it ... if so, it can run the 6502 core side at 12.5MHz.

5 minutes ago, The 8-Bit Guy said:

  It's still the same features, the same registers, and same behaviors. The sprites, the layers, the PSG, it's all the same.  The primary difference is how you copy data to VRAM

Maybe you're right that it would only take a couple/few hours to retool those programs to the X8.   

I don't know,  and WE (this community) CAN'T know since the X8's detailed specs are simply not available AT ALL.   

(Compare that with all the ample tech details we've had on the X16).   

As  a result, I find its very difficult to have a meaningful discussion about it.     But just from what you said describing the X8 so far, it looks to me like probably none of the top downloads on the X16 site would work without retooling them, and how difficult that would be certainly depends on those non-public technical details.   

Because its not necessarily JUST the VERA, is it?    A lot of people are using banked RAM in their programs, for example.   

So why not put those details out.   Just the technical description of how the X8 works.   What's the memory map.    What are the specific details of reading writing VERA?   Etc.       

Reading between the lines, and from your last few posts in particular, it seems obvious that you personally are strongly leaning toward the decision to push the X8 and its what you want to do.   But since you do seem genuinely interested in securing community sentiment in that direction, I would think it might help to get the tech details for the modified VERA addressing and maybe, perhaps, get the emulator out there!   It would let people get some empirical experience with the process of converting something that relies on X16 VERA etc. into working with X8 VERA etc.     I'd love to tackle all the 'offical' BASIC programs in the DEMOs directory of the official repo and see how they go.  

One more question which I hope I haven't missed in prior posts (this thread has moved fast):    Current docs on the official X16 repo indicate that $A000 to $BFFF (the 8K banked ram area) is initialized to BANK1 as the default for the User, with the X16 reserving BANK0 for Kernal Variables and Buffers. 

If the X8 has no banking, what happens with range $A000 to $BFFF?   

 

43 minutes ago, x16tial said:

You're asking us to choose or not choose a product we know very little about.  We've had years now to familiarize with the X16, and what, a little over 24 hours, and 2 or 3 posts to find out about the X8, pretty lopsided...

BINGO!

Where do I send the money?

44 minutes ago, BruceMcF said:

Yes, @The 8-Bit Guy this was the most important missing option in Question 1:



1. Option 4: Yes, release the X16 kit in beta to a limited number of selected developers willing to build or pay for a built board, then once the the X16 is ready for full release, release the X16p, X16c and X8 at the same time. No wait for the X16e, so no Osborne effects either way.



If that option had been available, that would have been my vote: use the X8 to avoid the x16e development phase altogether.



As a side benefit, this would also give time to contact Stefanie to see if she can help @Frank van den Hoef integrate a soft YM2151 core into the X8 design on an FPGA that can handle both. Then you have the same audio feature set, the X8 is a subset of the same video feature set, and you have much less "feature fracturing" between X16 and X8. Given the raft of FM chips and FPGA soft cores of FM chips on the Feonix256, having the YM2151 in the X8 would also narrow the porting gap between X16 software and the Feonix256.

Indeed.  I'm willing to lay down money right now on a built board, with things not quite right yet.  

I have put a LOT of hours of assembly language programming into the X-16, and the game I'm writing now absolutely requires accessing VERA exactly as it is in the documentation.  I don't care much about the keyboard being flaky, as long as I can type LOAD"ACOM.PRG",8,1  RUN

Seriously consider populating the Store; at least put a PayPal Donate button on there.  Keep the 8 bit guy Patreon strictly for 8 bit guy, so that accounting and taxes don't become a nightmare.

And I think you're far enough along in the design and testing that the issues with the keyboard and SD card can be resolved.  That means that this project is ideal for a Kickstarter.  It isn't vaporware.  You've got the license to use the commodore Kernal, you've got tons and tons of documentation out there (much of it on youtube).  You're on the third iteration of the board and almost got it.  All you need is cash to offset your own costs and get the project rolling.  You mentioned $100k to guarantee the project gets off the ground.  Put up a kickstarter: if you get your 100k within a month of the launch, then good.  If you get significantly less, you'll know it's time to reevaluate.  But at least people wouldn't be voting in a poll, they'd be voting with their wallets.

(yeah, I'm posting a lot on this thread, but this subject has me very jazzed)

Something occurred to me, by mentioning the incompatibility between the VIC20 and the C64:

1 hour ago, The 8-Bit Guy said:

It's nowhere nearly as difficult as porting between something like the VIC-20 and C64 which have very different video/audio systems.

 you touched on something very interesting.

The C64 was a success *because* 1) it was better in most ways and 2) incompatible with the VIC20, an already successful computer (which was the first computer to sell a million units).

This is proven out by what happened with the C128.  A nice computer, but a roaring success?  Not so much.  And also to a lesser extent the C16 and Plus/4 are another example.

The X8 actually seems better in many ways.  Faster clock speed, USB, more efficient video access, as you said, etc.  Yes, less RAM, but with fast access to external storage, is this a big handicap?  Is it a handicap at all?

The X8 seems like a cool product, and at the price you mentioned, very accessible.  But should it be compatible with the X16?  I'm thinking now, probably not.

Why did you make the choices you did with X16?  Discrete chips, and everything else.  I seem to remember you being pretty dead set against anything FPGA, but now you're arguing fairly strongly for the first product you release to be one that seems to be at cross purposes with all of that.  But it is cheap, and it's available now.

Cheap and fast does tend to be a bad strategy most of the time, I have to say.  These are tough decisions.

10 hours ago, The 8-Bit Guy said:

... The X8 could be available immediately and be well under $50.  I'm not sure how far under $50.  I'd say as low as $25 and as high as $50.  



The Phase 1 system sold as a DYI kit could be well under $300. Maybe under $250.  Add another $100 to $150 for a pre-assembled kit.



Again, you can't hold me to these numbers because so many things are unknown right now with the cost of chips.  But that hopefully puts things in a ballpark for people trying to figure this out.  



For those trying to figure out what the advantage would be of an X8 versus what is envisioned for the X16 Phase-3 (known as the X16e).  Well, the X8 would still be half the price.   For example, the X8 might be $35 and the X16e would be like $70.   There is simply no way to ever produce an FPGA based X16 as cheaply as the X8 can be produced.   And the X8 brings with it most of the functionality and personality of the X16.  And it's not an emulator.  So, there's that. ...

A separate reply because this is more as a professional Economist who brought my exposure to these markets in the 80s into my doctoral studies in the early 90s than as a not all that competent 6502 assembly language and Forth programmer ...

... that price break between the X8 and the CX16e ... whether [25/50], [35/70] or [45/90] pretty much nails it down for the X8, leaving the only question as release timing.  There is no question that in that double digit price range, half the price is well over twice the sales.

Even more, for the built versions including the assembly charge and limited order quantities per production batch on the CX16pm, that is a classic vertical price ladder. To have two mass market options and two deluxe options, you really want the higher priced mass market option around 50% higher than the budget option. The X8 (evidently) dominates the market for those who are not worried about how close it is to the system reference design, but at twice the price, the X8 will cannibalize much of the CX16e market even among those in the two digit price market even if at the same price they would have a strong preference for system compatibility. The CX16e in turn would cannibalize some of the CX16c market.

With three product price ladder, the market segregates much more cleanly. For the target market for the CX16c, there isn't really any cannibalization by the X8 ... for the target market for the CX16p, whether kit built or prebuilt, the X8 only enters into consideration as an optional additional purchase.

As far as timing, the launch of the crowdfunding for all of the pre-built options should be available in parallel at the same time. You want the crowd-funders choosing system options based on system preference, not based on timing. Hitting the individual crowdfunding targets literally decides which ones launch. Set the crowdfunded built CS16p option with a size cap, don't cap the DIY CX16p, the CX16c, or (obviously) X8.

 

3 minutes ago, Ed Minchau said:

Indeed.  I'm willing to lay down money right now on a built board, with things not quite right yet.  



I have put a LOT of hours of assembly language programming into the X-16, and the game I'm writing now absolutely requires accessing VERA exactly as it is in the documentation.  I don't care much about the keyboard being flaky, as long as I can type LOAD"ACOM.PRG",8,1  RUN



Seriously consider populating the Store; at least put a PayPal Donate button on there.  Keep the 8 bit guy Patreon strictly for 8 bit guy, so that accounting and taxes don't become a nightmare.



And I think you're far enough along in the design and testing that the issues with the keyboard and SD card can be resolved.  That means that this project is ideal for a Kickstarter.  It isn't vaporware.  You've got the license to use the commodore Kernal, you've got tons and tons of documentation out there (much of it on youtube).  You're on the third iteration of the board and almost got it.  All you need is cash to offset your own costs and get the project rolling.  You mentioned $100k to guarantee the project gets off the ground.  Put up a kickstarter: if you get your 100k within a month of the launch, then good.  If you get significantly less, you'll know it's time to reevaluate.  But at least people wouldn't be voting in a poll, they'd be voting with their wallets.

I've been watching all your dev videos on the asteroid commander game and yours is one of the projects that I though of when this thread landed.       

45 minutes ago, Snickers11001001 said:

... If the X8 has no banking, what happens with range $A000 to $BFFF?  

Though it hasn't been said there is no banking, it's just been said there's only 64K system RAM (because that FPGA has only 128K RAM internally). If they wanted to do the CX16 memory map in discrete parts but skip the 512K SRAM's, they could set the banking to 3 8K banks in the 64K Low RAM. That would be even easier to do in an FPGA, just allocating the lowest two bits at the $0000 address.

If the Kernel and Basic use Bank 0, that would leave two 8K banks in the "High RAM" window for program use.

All speculation, of course, due to only having a vague description of the X8 previously and not much more now, but that's the approach that would make porting between the two systems the simplest.