Hacker News Reader

Comment by fasteo | Hacker News Reader

Good times indeed !

Off topic. Little quiz:

  RANDOMIZE USR 0

What for ?

View on HN

OK, here's one (for the ZX81). Why did I choose ports 16509 and 16514 for the libvirt daemon (https://libvirt.org/remote.html)?

10 REM TAN

20 PRINT USR 16514

16514

Good one. TAN (the keyword, not the 3 letters) is hex C9 which is also RET in Z80 machine code. The address of the first byte after the keyword in the first statement in a ZX81 program is 16514. USR 16514 calls this address, which immediately returns. The PRINT statement then prints the result of the USR expression, which is the contents of the BC register on return, which I guess happens to contain the start address (since not otherwise set) as a side effect of the implementation.

16509 is the address of the start of the program in memory. Bytes 16509-16510 are the line number (2 byte LSB int). Bytes 16511-16512 are the length of the line. Byte 16513 is the command keyword (REM).

After thirty odd years I still have these numbers encoded into and handful of neurons somewhere.

Thank you for getting me to dig them out and great that they live on in libvirt!

ps The memory mapping reminded me that ZX81 managed to squeeze a 32 x 24 screen into a 1k total RAM. I think someone managed to write a chess program with those restrictions - but only using 8 x 8 or so of the screen.

Indeed, 1K chess by David Horne: https://en.wikipedia.org/wiki/1K_ZX_Chess

I had this chess game on tape, and I've often wondered if the two are related, but this one needed 16K: http://www.zx81stuff.org.uk/zx81/tape/ZXChessII(Black)

Not only you!

Shortest number of keypresses for a system reset?

bingo !

I remember trying to write machine code on my Speccy and it seemed that 99% of attempted runs resulted in a system reset. Always assumed that the professional games writers were using a CP/M machine or something similar with proper debugging support!

Indeed, as you can find out by magazines like RetroGaming, many were using 16 or 32 bit machines and then deploying into the Spectrum.

Ha, I thought that must be true at the time. It convinced me not to persevere. Good to hear I was right.

If I remember correctly most developers could only afford such systems after slowly establishing themselves, and they weren't cheap for 80's money.

So preservance using what the 8bit systems offered was a must anyway.

It would be interesting to recreate the full 1980s Spectrum development experience. Emulating a 1980s 16 bit machine on a modern PC, which runs a development system running, presumably, a Spectrum emulator!

You might find this interesting, https://retrocomputing.stackexchange.com/questions/3314/how-...

Thanks - looks fascinating!

Bonus points. The difference from NEW being?

RAMTOP!

Reply on news.ycombinator.com