Hacker News: jhallenworld

New comment by jhallenworld in "PicoZ80 – Drop-In Z80 Replacement"

jhallenworld — Thu, 09 Apr 2026 21:04:12 +0000

I want to make something like this as a classic CPU ICE, with trace memory, disassembly, etc. (note that you need a crystal oscillator circuit for many CPUs- 6802, 8085, etc.)

It would be useful for debugging classic computers like Altair 8800, etc. What you do is get a boot trace (record the first 100,000 instructions) of a working machine and diff it with the one from your broken machine. This finds the problem in like 5 seconds.

New comment by jhallenworld in "IBM Announces Strategic Collaboration with Arm"

jhallenworld — Thu, 02 Apr 2026 21:00:25 +0000

How well do commodity systems protect your financial transactions from cosmic ray-induced bit errors?

New comment by jhallenworld in "Afroman found not liable in defamation case"

jhallenworld — Thu, 19 Mar 2026 16:28:36 +0000

I thought it was big time defamation risk for Afroman to call him a pedofile.. but maybe the cop is afraid of discovery in this case..

New comment by jhallenworld in "Use GPT-5.4/CODEX to reverse engineer ancient machine code"

jhallenworld — Fri, 13 Mar 2026 21:39:30 +0000

I gave it the .S19 file containing the EDOS resident 6800 FORTRAN compiler machine code and asked it to reverse engineer it. It wrote a disassembler in Python along the way.

Use GPT-5.4/CODEX to reverse engineer ancient machine code

jhallenworld — Fri, 13 Mar 2026 21:39:30 +0000

Article URL: https://github.com/jhallen/exorsim/blob/master/stuff/edos_images/fort/fort_notes.md

Comments URL: https://news.ycombinator.com/item?id=47370319

Points: 1

# Comments: 1

Bill Gate's Comes Clean

jhallenworld — Thu, 26 Feb 2026 04:11:12 +0000

Article URL: https://wabcradio.com/2026/02/25/bill-gates-comes-clean/

Comments URL: https://news.ycombinator.com/item?id=47161739

Points: 2

# Comments: 2

New comment by jhallenworld in "AIs can't stop recommending nuclear strikes in war game simulations"

jhallenworld — Wed, 25 Feb 2026 20:38:25 +0000

>The biggest danger of a nuclear weapon is being hit by flying debris.

I thought it was being burned alive in the resulting firestorm because the intense light starts fires over a large area: way beyond the blast zone. This risk could be reduced if we painted everything white- a double win since it would also help reduce the city heat island effect.

New comment by jhallenworld in "AIs can't stop recommending nuclear strikes in war game simulations"

jhallenworld — Wed, 25 Feb 2026 20:31:14 +0000

Colossus/Guardian was the first AI to realize that the humans could be easily coerced by using their own nukes against them.

From the Colossus: The Forbin Project (1970) film.

New comment by jhallenworld in "Show HN: Formally verified FPGA watchdog for AM broadcast in unmanned tunnels"

jhallenworld — Wed, 18 Feb 2026 22:36:15 +0000

>NCO: 12 Numerically Controlled Oscillators generate carrier frequencies (505–1605 kHz)

This seems like a crude way to do it.. why not provide all 110 carrier frequencies by using a polyphase channelizer? The bandwidth of the entire AM broadcast band is pretty low..

New comment by jhallenworld in "Ask HN: What are you working on? (February 2026)"

jhallenworld — Mon, 09 Feb 2026 02:22:26 +0000

Adding EXORdisk-I support to my MC6800 simulator so that it can boot EDOS and EDOS-II disks.

EDOS was a direct 6800 port of FDOS. FDOS was the first DOS available for microcontrollers, using iCOM's FD360 8-inch floppy drives.

https://github.com/jhallen/exorsim

https://www.youtube.com/watch?v=TpHKygZ7OHY

New comment by jhallenworld in "We tasked Opus 4.6 using agent teams to build a C Compiler"

jhallenworld — Thu, 05 Feb 2026 22:28:25 +0000

Does it make a conforming preprocessor?

New comment by jhallenworld in "Child prodigies rarely become elite performers"

jhallenworld — Thu, 05 Feb 2026 04:13:09 +0000

It's like those articles that say super high IQ people are not always successful.

So I think human brain development is like some kind of optimization algorithm, like simulated annealing or gradient descent. I think this because there is way more complexity in the brain than there is in human DNA, which has pretty low information by comparison. Anyway, child prodigies occur when the algorithm happens to find a good minimum early on.

New comment by jhallenworld in "New York’s budget bill would require “blocking technology” on all 3D printers"

jhallenworld — Wed, 04 Feb 2026 03:56:54 +0000

The receiver is like the asset tag on computer servers- it's the one thing that is definitely not replaceable since it has the serial number used for entitlement.

New comment by jhallenworld in "Banning lead in gas worked. The proof is in our hair"

jhallenworld — Tue, 03 Feb 2026 16:33:20 +0000

If the end product ends up marginally cheaper, the company will be able to sell more of it, and this will lead to more profit. And sure, when you ignore the cost of the pollution, this certainly benefits the consumer, by allowing them to afford more energy and energy-based products (i.e., just about everything).

But then we come back to ignoring the cost of the pollution. It certainly gets paid for eventually, but by who? Also, it's cheaper for everyone if the pollution is eliminated to begin with rather than being cleaned up later (which is certainly a more energy intensive endeavor).

New comment by jhallenworld in "Banning lead in gas worked. The proof is in our hair"

jhallenworld — Tue, 03 Feb 2026 15:26:33 +0000

Lead helped fuel company profits because it was cheaper than the other anti-knock additives, like ethanol.

New comment by jhallenworld in "Floating-Point Printing and Parsing Can Be Simple and Fast"

jhallenworld — Fri, 23 Jan 2026 22:49:24 +0000

I have been using Walter Bright's libc code from Zortech-C for microcontrollers, where I care about code size more than anything else:

https://github.com/nklabs/libnklabs/blob/main/src/nkprintf_f... https://github.com/nklabs/libnklabs/blob/main/src/nkstrtod.c https://github.com/nklabs/libnklabs/blob/main/src/nkdectab.c

nkprintf_fp.c+nkdectab.c: 2494 bytes

schubfach.cc: 10K bytes.. the code is small, but there is a giant table of numbers. Also this is just dtoa, not a full printf formatter.

OTOH, the old code is not round-trip accurate.

Russ Cox should make a C version of his code..

New comment by jhallenworld in "Microsoft gave FBI set of BitLocker encryption keys to unlock suspects' laptops"

jhallenworld — Fri, 23 Jan 2026 20:38:44 +0000

Agree, use Linux, use LUKS.

PGP WDE was a preferred corporate solution, but now you have to trust Broadcom.

New comment by jhallenworld in "TPM on embedded systems: Pitfalls and caveats to watch out for"

jhallenworld — Wed, 21 Jan 2026 17:34:27 +0000

I mean a separate chip.

New comment by jhallenworld in "TPM on embedded systems: Pitfalls and caveats to watch out for"

jhallenworld — Wed, 21 Jan 2026 17:13:00 +0000

Do you really need a TPM if you have something like ARM TrustZone?

New comment by jhallenworld in "A glimpse into V8 development for RISC-V"

jhallenworld — Thu, 08 Jan 2026 02:07:58 +0000

>V8 emits a near-jump for all forward jumps. If the target ends up too far away, the near-jump target is adjusted to a jump pool entry that contains a long-jump to the actual target

This sounds weird to me... Why not assume that all jumps need to be long to start with (so that the code is valid), then relax them to short jumps during an N-pass optimization stage- so that you get the smallest possible code?