Happy New Year and happy hacking in 2025!

[BigDumbDinosaur]

BigEd wrote:

Your smiley does not excuse your bad behaviour.

And just who assigned you to be the arbiter of bad behavior?

[BigEd]

If you keep posting sexist or racist comments, I will keep reporting them. I believe the forum is harmed by those comments, and would be better without them. They are always off-topic.

[BigDumbDinosaur]

BigEd wrote:

If you keep posting sexist or racist comments...

Quit making accusations you can’t support with unimpeachable evidence.

[jgharston]

Yuri wrote:

BigDumbDinosaur wrote:

barnacle wrote:

I have similar information, but I need to update it.
I'm 41 next birthday. Er, 0x41.

You mean $41, doncha?  Or, should it be %01000001?    Heck, I’m not quite $50.

I will be a spry $30 this coming March..... >_>

I've just turned 0x38

On the semi-main theme, I made a will 15 years ago and in it specified various computer/electronics related stuff. It's been 15 years, and I'm planning on updating it this year, and one of the updates will be the inevitable update of who and where to inform of my continuing existence failure.

[barnacle]

I've updated my 'in the case of my death' paper, but I need to rewrite my will (UK law) to fit with DE law. Or rather, pay an expert to do it.

Neil

[BigDumbDinosaur]

barnacle wrote:

I've updated my 'in the case of my death' paper, but I need to rewrite my will (UK law) to fit with DE law. Or rather, pay an expert to do it.

My buddy Paul, who is an attorney (and was a COBOL programmer in a previous life), once told me that writing an “unimpeachable” will is definitely not something for an amateur to undertake.  He has been doing wills and trusts for some 30 years, and has seen many a do-it-yourself will get voided in probate court over trivial errors.  He described a probate hearing in which family members almost came to blows when one of them realized he wasn’t getting his “cut” from the deceased’s estate, thanks to an invalid will.  Apparently, the will had been prepared by another family member. 

Hmm...maybe I should take up probate law so I can witness some of these family altercations. 

[Paganini]

As mentioned elsewhen, I was less active here in 2024 due to the fact that we moved to Texas just about a year ago, and all my electronics stuff was packed up! Our new home had a small unfinished outbuilding in the back yard. Over the summer I finished the building and turned it into a mini-wood-shop for luthiery work. This week I finally got around to unpacking the electronics workshop and installing an "electronics corner" with my old work table. Here are some pics!

Once it warms up a bit (the shop doesn't have climate control yet!) I plan to get started building a new project!

[BigDumbDinosaur]

BigDumbDinosaur wrote:

Getting back on topic, despite a fair amount of software debugging taking place as I work on the building blocks of my multitasking kernel, POC V1.3 reached 76 days of uptime last night, which means the unit survived the New Year.

The saga of up time has come to an end—she made it a hair past 80 days.  I’ve brought in the New Year with new firmware for POC V1.3, which, of course, meant shutting down and swapping out the ROM.  Up time right now is 1 hour, 11 minutes and 48.61 seconds.    Now I can resume on my kernel work and the unit testing that is part of it.  We’ll see how far things go before I do something dumb and cause a computing train wreck. 

[J64C]

My plan (which I started on tonight), is to expand upon my “J64C Concept” computer, to make it something more usable. Including VGA graphics, better I/O, etc.

My first build was surprisingly good, for a first attempt. Stable at 25 MHz. But the downside of having a single VIA chip as it’s only avenue to the outside world, making it somewhat limited.

I’ll make slight alterations, where I could have done better on the original, and go from there.

With what I have in mind, 12 MHz would be awesome, but 6 MHz would be more likely, due to “reasons” hehe, but you never know. I might surprise myself.

[drogon]

I'll add to this as it's still the new year and I've been a little busy...... I decided to re-visit my minimalistic 6507 project (Codename Project 28 as the 3 main chips, CPU, RAM, ROM are all 28 pin ICs) It's essentially a 6502 with no interrupts and just 4KB or RAM and 4KB of ROM. It runs my own variant of TinyBasic with bit-banged serial for the main IO, also with 4 LEDs and a button on board, as well as a 65C22 VIA for more IO.

The ROM is actually a 32KB EEPROM arranged as 8 banks of 4KB. One bank is the code, the other 7 can contain a saved BASIC program. (Literally as a RAM image - there is no attempt to be clever and implement a "filing system")

But there was a small nagging problem; If a program accidentally wrote to it, it could corrupt the contents and subsequent boots might crash. The EEPROM I'm using has software protection, but is shipped with protection turned off.

The ROM was full - almost. I had one byte free before I embarked on an exercise of "code golfing" but I managed to squeeze enough space to disable the write-protect when saving a program and re-enable it afterwards. This involved a lot of reading the manual and testing, as well as trying to eek out the space needed - compounded by the fact that the actual code to save/load programs needs to run from RAM for obvious reasons. One feature I temporarily disabled to give me space to write the code initially was the turnkey/autostart but I really wanted to keep that so I could power it on and have to "just work" without needing the serial port. I did eventually managed to get that back in...

And then just as I was feeling smug, a comment on Facebook made me go back and find more space to implement a new command "BTN" so I could read the button from a BASIC program without resorting to using the peek command (Which would look like: IF ?&F8 = 128 THEN .. pressed - the syntax will be familiar to Acorn/BBC Micro people - ? is peek and & is a hex number, and yes, despite not being a fan of it, I use Zero page for IO - I need every spare byte of that 4K ROM space including the 2 bytes of NMI vector!)

It already had an LED command that you can use to write/read the 4 LEDs - takes the bottom 4 bits of the operand: e.g. LED = 5

And after that, I feel I'm past hole 19 in this round of code golfing for the system.... But this sort of becomes almost an obsession - I'd like to implement some "wiring" like commands for the ports of the VIA but each command has a 6 byte overhead + the length of the command string, so a DW p,v (Digital Write Pin, Value) will need to find 8 bytes plus the actual code, then there's DR (Digital Read) and PM (Pin Mode) to 'complete' the 'wiring' like commands.

Anyway, I've just done a re-spin of the PCB and have a few new ones and my plan is to build one up to test it, then offer it as a kit to anyone who wants, so we'll see how that goes. I'm only aiming for 10 or so in the first instance for nothing other than a bit of fun.

Cheers,

-Gordon

[barnacle]

Best of luck, Gordon: to be honest, I'll be happy to get my tiny basic up and running on a 65c02; then I can get on with the rest of the hardware.

(Unless I get distracted by thoughts of an IDE with a 65c02 emulator and proper debugging (step into/step over being my main annoyance at present ) but writing graphical interfaces these days involves so many layers of libraries...)

Neil

[BigDumbDinosaur]

J64C wrote:

With what I have in mind, 12 MHz would be awesome, but 6 MHz would be more likely, due to “reasons” hehe, but you never know. I might surprise myself.

Think of your computer in terms of a drag racing car that you have designed and built from scratch.  Your first concern is to get it to go, your second concern is to get it to go straight, i.e., not crash, and your third concern is to get it to go fast.    Proper construction methods will make it all possible—consider how inexpensive four-layer PCBs have become.

Something worth looking at is using a PLD in place of discrete logic.  A single 22V10 GAL can take the place of a handful of gates, and in the case of Atmel parts, offer single-digit prop times.

As for I/O, you should see about giving it some mass storage capability.  It doesn’t have to be anything fancy...just capable of reading and writing blocks of data at reasonable speed.  floobydust has his “C02 Pocket” unit rigged up with a PATA microdrive (can’t recall the manufacturer)¹—basically a miniaturized version of the venerable ATA hard disks that used to be in PCs.  It achieves in excess of 300KB/second on single block (512 bytes) accesses, more than good enough to support running DOS/65, which is a disk-intensive OS.

————————————————————
¹EDIT: I belatedly recalled the microdrive was made by Hitachi.  Kevin had mentioned to me he was getting about 380KB/sec on reads, which is much less than the drive’s raw capability.  The chasm is due to having to use PIO to talk to the drive, which involves a lot of monkey-motion in the driver code.

[J64C]

BigDumbDinosaur wrote:

J64C wrote:

Something worth looking at is using a PLD in place of discrete logic.  A single 22V10 GAL can take the place of a handful of gates, and in the case of Atmel parts, offer single-digit prop times.

As for I/O, you should see about giving it some mass storage capability.  It doesn’t have to be anything fancy...just capable of reading and writing blocks of data at reasonable speed.  floobydust has his “C02 Pocket” unit rigged up with a PATA microdrive (can’t recall the manufacturer)—basically a miniaturized version of the venerable ATA hard disks that used to be in PCs.  It achieves in excess of 300KB/second on single block (512 bytes) accesses, more than good enough to support running DOS/65, which is a disk-intensive OS.

Hi BDD! Way ahead of ya! In the first point anyway. I’m using a Xilinx CPLD to take care of a ton of the logic. The whole board presently consists of six IC’s.

Your second point is high on my list and is greatly needed, I have no forum of external storage access yet. It is something that I absolutely need to address.

Your suggestion of a micro drive sounds very interesting. I’ll have a look at that.

[barnacle]

My next system - designed but only part built - uses a Pi pico to generate vga video, provide serial and keyboard ports, and an initial program load to the 65c02, which has therefore an all-ram memory space apart from a tiny bit of IO at 0xff0-ff7f. The storage is a 2G Compact Flash which has an parallel interface which is dead simple.

There's a long way to go still...

Neil

[drogon]

barnacle wrote:

Best of luck, Gordon: to be honest, I'll be happy to get my tiny basic up and running on a 65c02; then I can get on with the rest of the hardware.

A 65C02 is a luxury I don't have. I could save a few 10s of bytes with the C02 instructions (Mostly STZ and PHX/Y ...) and right now 10 bytes would be great!

Quote:

(Unless I get distracted by thoughts of an IDE with a 65c02 emulator and proper debugging (step into/step over being my main annoyance at present ) but writing graphical interfaces these days involves so many layers of libraries...)

Neil

On my "bigger" Ruby boards, I left the video for something else - in my case a "smart" video terminal that run under Linux using the SDL library. It was a spin-off from my Linux BASIC I wrote some time back - the terminal version takes standard Acorn MOS VDU codes for text and graphics which means I can run other things on the Ruby board like BBC Basic and all (almost all) built in graphics, etc. "just work".

Keen to see your Basic though - keep going!

Cheers,

-Gordon