WOW.... seriously, guys,
WOW. I never expected this topic to explode.
Apparently you guys are serious about creating a new calc.
Looks like it's time for me to draft up a nice little infographic about the building process for electronics (which applies to both companies and communities)!
Note: Just because I started this topic does not mean I have the ideas, nor does it mean that I'm building a device. It's a community effort! 
As soon as I finish up an essay, I'll work on my infographic, and then upload it for everyone to understand the process.
First off, before I go and start replying to everybody,
DO NOT TAKE OUT YOUR MONEY, NOR COMMIT YET!!!!!!This is a great project, but we're leaping before we're thinking.
We don't even know how it looks like, nor what the final features are, nor the priorities - how in the world do we design a device that we have no idea about?Now, I'm not trying to dampen the enthusiasm here. It's very possible to create an awesome product from a community, but just jumping to build something without any serious planning first is kinda suicidal. Let's all think it out first, agree on something, follow the standardized steps to build this kind of thing, and then get building. (That's where my infographic comes in - it basically outlines the steps toward product development and such)
Money is kind of an important aspect in product development, and we will need it in the future, but NOT yet.
ALSO - there's 2 sides of this project - a Z80 side, and an ARM side.
Z80, obviously, is going to be a competing calculator against the TI-8x series.
ARM will compete against TI's Nspires. Make sure you're clear when suggesting priorities and features which one you're referring to.
(I might split this topic into 2 if needed)
Whew. Now all that is done, time for some personal replying:
i am all for this project, but i am ignorant for about everything this includes.
I am willing to invest in this project though. And will help as much as i can.
That's great!
But no money out yet until the project is ready to move forward.
If you do make a calc, be sure to include a sensor pad, just like a laptop
Also, use 48-bit deep color for the screen with 1024x768 resolution (j/k)
http://en.wikipedia.org/wiki/Truecolor#Beyond_truecolor:_Deep_Color
Oh noes! How will the Z80 or ARM chips ever handle this? lol
(I know you're jk)
Most likely for Z80 is a grayscale screen (I don't think Z80 is powerful enough to handle color)
For ARM, definitely a color screen.
Oh, I would be perfectly fine running KOS on this when it's done.
That'll be awesome!
Of course, it might need some modding, since all Z80 hardware are different in some ways.
If this project does go on (the calc building one, that is), you really need to create a serious compatibility layer for TI basic programs
and ASM programs. Don't do it now - focus on the core, guts, and such (kernel, libraries), and then start creating this. (if possible)
I'm game.
But, we will need a knowledge base.
I dunno how to hard mod at all.
There's some people in the TI community who love to hack around with electronics, and know how to design them as well.
A wiki could be helpful in collaborating efforts. (That was aimed more to DJ
)
Yeah, same here. Ben Ryves knows how to, and KermMartian, and probably a few others as well.
Yeah, there's quite a few. I know for sure that KermMartian can access some equipment for prototyping (I think).
how much are we going to spend on each calc to make?
max price for:
screen-
buttons-
shell of calculator-
battery-
back up battery-
hardware and chips-
advertisement-
shipping-
location of selling-
total cost?-
selling price?-
profit?-
The overall goal (for the Z80 calc, that is) is to be < TI's price. I'd prefer < $80, for added cheapness.
It's all part of the building process, so it's great that you recognize that. But, we haven't even figured out the design
or the hardware yet, so there's no estimate on pricing until that's done.
For profit, it's a tricky question. We don't want to overprice it like TI - remember, this is a community project, not
some kind of big greedy company project - but we don't want to be penny broke at the end either. I suggest
maybe $10-30 profit per calc? As I've said before, we need a solid plan before deciding.
Advertising is tricky too. Social advertising can be very cheap and successful if done correctly, but that takes a while.
As said before, we need something solid before deciding on these.
One question:
How the hell are you going to collaborate the build process?
That's a good question. I'm not collaborating (look at the top of my reply), the community is.
For hardware specs, etc., I suggest using a wiki for that, as well as a source revision system (SVN, Mercurial, GIT, etc.) for schematics and code.
if anyone will gie me something to read about the chips and stiff for making it that would be great.
No plans yet, but you can google up Zilog and how their chips work, or ARM if you're going for that side of the project.
One question:
How the hell are you going to collaborate the build process?
Probably a SVN with all the physical circuit diagrams, and the software.
Pretty much.
Very intresting project, altough it is all just specualting now. Reminds me of the Pandora . Well anyway altough I am really just starting with all this kind of stuff I do know something about processors and circuits. If anybody is going to take this seriously then you ahve to start to decide what kind of processor 'we' will use. I think a 32 bit processor should be more then enough, 16-bit might do the job as well. Some good 16 bit processors are:
INTEL 8088 @~7.2 MHZ
INTEL 80188 @~18.72 MHZ
Some good 32-bit processors (and not to expensive) are:
INTEL 80386 ranging from 16 to 33 MHZ
INTEL 80486 ranging from 25 to 100 MHZ.
Altough clockspeed isn't everything I suggest at least 50 or more MHZ. Another focus point would be a (G)LCD with a high refresh rate so we can use the processor at full speed when writing to or reading from the screen.
Another Idea I had but which I am not sure about is practical is having several Z80's. The reason to choose Z80's in the first place is because this whole calculator community is familiar with it and we want to develop for this community right? Ease of use is a second.
My last suggestion would be the most updated version of the Z80, the eZ80 (http://en.wikipedia.org/wiki/Zilog_eZ80). Being 9as wikipedia says) 4x as fast as a regular Z80 it might be enough. Frankly what I need to know is how resource intensive the final OS (linux) will be, based on that decide the processor and based on the processor decide the other hardware.
So what is going to be this things purpose? Is it going to be a gaming device with some math capabilities or is it going to be a hybrid? Shortly siad, what do we want it to be capable of? Maybe everyone should make a wishlist?
Ahh, the Pandora. It's a real project though, so that's not misty at all.
No Intels - it's a bit hard to get chips from them, and licensing fees may be needed.
A good LCD is a must here, that's for sure.
There's 2 projects for calc building - Z80 and ARM. We do want to keep the Z80 (TI) community alive too!
eZ80 sounds good
This is strictly a calc project - we're attempting to knock down TI.
Games and other things will be included, but this is mainly math/calc.
if anyone will gie me something to read about the chips and stiff for making it that would be great.
this is a vast resource about digital electronics. It contains pretty much everything you need to knwo except how to solder. It teaches all the theory. (http://www.asic-world.com/digital/tutorial.html)
Definitely something worth looking at.
great info, thanks.
To your question. I think the calculator should have enough math for at least algebra 2, if not more.
It would be great if this actually became a selling calculator. one of the things that really sticks out with this is the color. We could have graphing capabilities with each function a different adjustable color. So many things could happen with color for math. We could do alot with it. If we would also be able to run asm programs through it, well that opens up alot for making games and everything.
things i would like to see:
1-10 for imporance for me.
color-7
mouse or keyboard port-5
math-8/9
speed-???
connectivity possibility with ti calcs.-6
Great ideas!
Math should be top priority though - 10/10 or higher.
for me it would be:
color - 10 (It's a must have)
mouse-7
keyboard port (from TI's calcs)-3
math-8
speed- 10 (It's a must have)
connectivity possibility with ti calcs - meh, this is software. I have no worries about this being implemented, since we will probably use the same usb port.
Good
Math should be 10, otherwise it's not a calc.
All right.
Let's make this official.
Are we doing this?
Do we have the money and the commitment to do so?
If so, say 'Aye'.
NO - we're not ready yet! NO INVESTING!
color - 10/10
mouse- 8/10
keyboard port or mouse port-2/10
math- 9/10
speed- 11/10
Resolution: 3:2 ratio? At least 144x216
You should have more ports in case some break.
Color: 6-bit?
Have at least 256 contrast settings and maybe built-in 4-bit grayscale.
Aye.
Kind of confusing - which calc is which?
And NO COMMITMENTS/MONEY!
Aye, altough I'd prefer to take a siderole. I am willing to invest a little, it really depends. Also I have alot of other projects running in parallel so well, yea I am not constantly available. We need people like BenRyves and KermMartian though, and we need their time :p
I think we should first maybe produce a dozen devices maybe more, maybe less.
As for importance:
Color: 8
Mouse: 7 (if it's going to be a linux driven system then it might be useful, adding a PS2 port isn't that hard (or we could go usb))
Keyboard: 9 (allows for greater productivity, if used wisely often faster then a mouse and if there is a PS2 port already...)
Math: 6 (important but not a must)
Speed: 10 (speed allows for great creativity AND, you know 1+1=2 faster then the other folks cheating the test)
LCD Resolution: QVGA or VGA (respectively: 320x240, 640x480) I don't really care that much, if it raises the price to much I deem QVGA enough. If we are going to have a (multiple) window based GUI then 640x480 is pretty much a must.
Color Depth: 8 bit (pretty high quality, if it affects speed to much the take 4 bits)
Math = 10, no exceptions. How else are you going to sell a calculator that doesn't do math?
No money yet, please!
Keep in mind that Linux is programmed in C, so z80 isn't the best option. (compilers are not very good for z80)
Linux and ARM is beautiful. Z80? Not sure about that.
Aye!
NO COMMITMENTS/MONEY!
Aye!
i will invest. (praying that it would be a good investment)
NO COMMITMENTS/MONEY!
Keep in mind that Linux is programmed in C, so z80 isn't the best option. (compilers are not very good for z80)
Aah true, forgot about that. Maybe a PIC is more flexible then? MicroChip has some pretty good ones. Maybe we could even have a dedicated GPU? (seeing that Micrcontrollers aren't that expensive). It adds to complexity though. Does anybody know any good 16/32b C compatible microprocessors? I honestly don't altough I bet the intels I referred would work...
Nah - Z80 is the best option for competing against TI-8x.
ARM can handle Linux for sure. Good example: Androids.
EDIT: So many things - so many ideas! We need a wiki setup, pronto. Set one up on Omnimaga or Cemetech, no wiki hosting sites!
