Hacker News Reader

Comment by j2kun | Hacker News Reader

I'm not sure why the size of a program has any serious bearing on it's greatness. Sure it's an incredible feat of engineering, but shouldn't the "greatest program ever written" be primarily judged by what it allows us to do, not need to do, to build, or what insights it provides?

For example, certainly the first widely used assembler was a "greater" program than any chess game, not only because it stopped us from having to track memory locations by hand, but because it allowed us to dedicate our brain cycles to actual problem solving and writing more useful programs.

View on HN

The 4 minute mile was amazing because of the four minutes, not because someone managed to run a mile. Amazing feats in the face of incredible constraints are inherently worthy, I'd say.

A better analogy would be 'greatest movie ever made' was so because it was done on a budget of $1000.

Better analogy in context of this thread that is

I think the analogy is wrong...

Maybe making an oscar winning movie with great CGI on $1000 is more appropriate

Was the godfather 1 done in a budget of 1k?

Why? Seems like a much poorer analogy to me, since it is trivial to make a movie for much less than that. It falls into the same trap as the commenter above, the 'greatness' is due to the quality of the program, rather than just making it exist at all.

The greatness of a run mile is measured by how fast it is run.

The greatness of a movie is not measured by its budget, nor is the greatness of a program measured by its number of lines or memory footprint.

Do you see now why your analogy is not as good?

Nope sorry. Do you see why your explanation is poor?

The amount was arbitrary. The point is that there is more to a greatest movie than resources spent

Which is exactly why it isn't a good analogy. The amount wasn't arbitrary, it was the memory of the computer. There wasn't more to it being the 'greatest program' than the fact that it was created. There's no aesthetic judgement here. It is chess, with AI, in < 700 bytes.

Well, make it the greatest movie of all time because it was done with 1 $.

He meant the "movie budget" could be arbitrary chosen to satisfy the "feat under huge constraints".

Impressive but not necessarily deserving of the monicker: "best ever".

He meant the "movie budget" could be arbitrary chosen to satisfy the "feat under huge constraints".

Right, which is exactly why that isn't a good analogy to something where the budget was fixed, and the 'greatness' isn't an aesthetic judgement, but whether it can be done.

The greatness of an achievement may be in and of itself.

The Mona Lisa is great purely based on its appearance, not that anything useful comes of it. The "7 Wonders of the World" are noted for their scale vs the ability of then-modern construction, not their ability to move water or house dead bodies. Great exploration often leads to laudable yet largely unproductive dead ends (Everest, Antarctica). In this case, cramming a viable AI playing the quintessential expression of human intelligence (playing chess) into less than 700 bytes is astounding, and at least a candidate for greatness.

The first widely used assembler may vie for "greatest tool", but it did little of interest in and of itself.

And yet, when you ask what is the greatest thing or idea ever created by humans, the reply is not usually any of the Wonders of the World, or the Mona Lisa, or a chess program. It is usually language or writing or something which was similarly a sea change in the development of humanity as we know it today.

My argument is that these are the better criteria for "greatness" than the relative ease of performing a specific task, though that task may implicitly represent the great thing. E.g., one may call the printing press the greatest invention ever, not because it was great but because it ushered in a new age.

And this ushered in the age of chess on your home computer :) So it was practical while simultaneously being an engineering challenge, which is what (I guess?) most of us find so appealing.

I agree. I couldn't care less what he had to do in order to fit his not really usable (from a chess enthusiasts perspective) chess implementation in N kb. Apparently N kb is not enough for a good chess program.

However, 1024 bytes was, at the time, all the RAM available, in a computer that was the only one many people could afford. For some, like me, a computer was in fact a very good way to find a chess opponent.

Therefore, for a chess enthusiast of 1981-1982, this program was groundbreaking. You could actually get a computer to play against you.

(Personally, I did not have ZX81 but I got ZX Spectrum. I had to mow a lot of lawn and dig some graves at the local cemetery to get the money for a 16K Speccie.)

> Therefore, for a chess enthusiast of 1981-1982, this program was groundbreaking. You could actually get a computer to play against you.

That's a really good point I guess. Maybe if I was a bit older I would be much more amazed by the feat.

Yes, there were some feats done already then. I guess people can't imagine what strange tweaks were used in the machines.

Take for instance the 48K ZX Spectrum: it had 16K ROM from 0000H to 3FFFH, then 16K RAM (some of which was display buffer) from 4000H to 7FFFH made of eight 16 kbit chips, and the remaining 32K RAM from 8000H to FFFFH was actually made of eight faulty 64kbit chips which Sinclair could buy cheaply. The chips that had failed in manufacturing testing were just tested again to determine whether the faulty part in the chip was in upper or lower half of the 64kbit address range. The high address bit of memory chip was connected permanently high or low to keep the address pointing to a working half of chips, thus enabling the extremely cheap broken 64kbit chips to be used as valid 32kbit chips.

(All eight chips of memory of course had to have be broken in the same upper/lower configuration).

The Psion chess program for Spectrum needed the 48K RAM model, although I think someone also wrote a full chess program for the 16K model (including castling, en passant and all).

Psion, by the way, was publisher for much of the early software for ZX Spectrum, including chess, spreadsheet and 3D design programs. Later on it was known for handheld devices, as well as the EPOC system which later on was to become Symbian, used in Nokia smartphones.

Reply on news.ycombinator.com