computing pi with square roots / RPN calculator build

[BigEd]

Hee hee, well there will always be ideas for making a thing into a bigger more complex and more marvellous thing. They might not be worth doing.

[barrym95838]

Only two years for a contraption of that magnitude? I see that you either have a lot of spare time and energy, or you take your hobby quite seriously!

Or, maybe that's only comparing your situation to mine ... my full-time job is physically and mentally draining, I'm getting old, and I have poor time-management skills. I have an automotive project that is half-done, and it only took me 25 years to get to that point.

Mike B.

[calculi]

Hi Mike,
Yes, 2 years... but I cheated seriously, I suppose it's the same for all of us : when you're doing something, the other topics don't progress. All in all, I spent between 600 and 700 hours on that project. I'm lucky enough that, back home from work (which isn't tiring as it used to be, activity is low), I often have 1-2 hours for me. It hasn't always been the case, and I'm afraid it won't be anymore soon, priorities are about to change, and not for the best.
About energy : I'm more than 50, I know my energy will lessen someday... I've been collecting calculating devices* for 30 years. I began dreaming of my "own" calculator 20 years ago, so it's some kind of an culmination for a 30-year old passion. To me, passion is what makes a life. During last year when I was finishing the firmware, some days in the morning I woke up with ONE thought in my head : "I've got the solution for yesterday's bug !". Software coding syndrome, I suppose...

Speaking about mechanics : I've got tremendous respect for mechanical devices, some are genuine masterpieces, and can last almost forever if well crafted and cared of. You need more time to create them but they'll survive almost every electronic device... If only I had enough room (and time, and money, and so on...), I'd set up a workshop ! One life isn't enough

Marc

*of course there are mechanical calculators in my collection...

Edited - readability.

[calculi]

BigEd wrote:

..... They might not be worth doing.

You're right ! Sometimes it's better to say "I'm satisfied with that, I don't need to tweak it anymore". But I can't help thinking about such or such enhancement. Hacking is so fun !

[BigDumbDinosaur]

calculi wrote:

Eh eh... should I build and program a new, 65C816-based machine ? In such a case I need 2 more years

Hardware-wise, the design won't be too much different than your 65C02 unit, unless you plan to have "extended" memory (more than 64KB) in the new machine. Even then, a reasonable bank latching circuit could be implemented with basic gates. If 64KB is enough then you could almost copy your 65C02 design.

Software-wise, your code would have to undergo a fair amount of revision to fully exploit the 65C816's capabilities, especially in the realm of 16 bit operations. The raw math functions would greatly benefit from 16 bits at a time operations. Also, with better stack management features, you may find yourself looking at new ways to implement some operations. The payoff would be a substantial improvement in performance at any given clock speed, as much as 40 percent in some cases.

So of those two more years, I'd say most of them would be consumed in software development. As we have often said when it comes to race cars, horsepower costs money: how fast do you want to go?

[BigDumbDinosaur]

barrym95838 wrote:

I'm getting old, and I have poor time-management skills. I have an automotive project that is half-done, and it only took me 25 years to get to that point.

My large-scale locomotive project has been in the works for 11 years, although the unit has been runnable since 2008. I have pretty good time management but being self-employed, often find myself burning both ends of the candle trying to keep up with the workload. Faltering health isn't helping any...

EMD F7

[calculi]

Hi BDD, I can only agree with you on both sides, hardware and software : it took me 2 months (cumulated) to build the hardware, the rest of the time for the software. Currently I'm more inclined to build a slightly enhanced "release 2" of my 8-bit hardware, with a PCB-based SBC and a custom cabinet + keyboard, than to launch a 16-bit version. Of course I may change my mind someday... but I feel running fast enough, by now -especially compared to HP and TI calculators from the 1980s : more than 100 times faster

I recently browsed your locomotive-related posts : impressive ! What a nice work, I sincerely hope you'll keep going. One of my grandfathers worked for the French railroad company, he did maintenance on steam locomotives. I loved hearing him recalling his past experience.

BTW : I posted a suggestion on your "Patience needed" thread, which Tor completed. Will that be useful to you, of course I don't know....

Marc