They said that it would take 1 trillion years at current rates, but because the rate keeps changing it would take forever.
Bitcoin Mining on Mac Classic
- Thread starter DrewyNucci
- Start date
It speaks highly of this community that you got as warm and friendly a response as you did.
That would be quite funny, if you had actually made it. That you had gone to such effort for something that nominally works but is essentially useless would be quite entertaining.
As a Bitcoin enthusiast, I expect you're familiar with proof of work. That's what gives you cred (as in actual credits) on the blockchain, and it also works (as credibility) in real life.
Unfortunately, it appears your project has been upstaged by your initial misrepresentation of it.
I did have a look at the "Hand-tuned SHA-256 implementation", though (leading me to wonder whose hand was involved, if not yours), and I'd like to call attention to this gem:
Code:
; Optimized 32-bit rotate right
; Input: D0.L = value, D1.W = shift count
; Output: D0.L = rotated value
; Uses: D2, D3, D4 (caller saved)
ROTR_ASM:
MOVEM.L D2-D4,-(SP) ; Save registers we use
MOVE.L D0,D2 ; Copy value to D2
MOVE.W D1,D3 ; Get shift count
ANDI.W #31,D3 ; Mask to 0-31
LSR.L D3,D0 ; Shift right by n
MOVEQ #32,D4 ; Load 32
SUB.W D3,D4 ; Calculate 32-n
LSL.L D4,D2 ; Shift left by (32-n)
OR.L D2,D0 ; Combine both parts
MOVEM.L (SP)+,D2-D4 ; Restore registers
RTSWhether this can be optimized further is left as an exercise for the reader.
Maybe instead of stubbornly insisting other people are wrong about something, you could simply accept their viewpoint as their own, and be quiet about it. From here, it looks as though you posted a thing you told an AI to generate and then invited everyone to Test “your” software with no detail details about what it does or how it does it.
As a software QA tester, the idea of someone vibe coding a thing and sticking it out on the Internet for everybody else to Test is sort of… yeah, that’s not how you do that in a day and age when every binary could potentially hijack your network among other things.
Your “nothing is real, everything is vibes” attitude is a bit antithetical to most of what this community seems interested in. I think it’s an attitude you might tone down.
I get where you’re going with this, but your approach could use some work.
I have to admit though, I'm enjoying reading everyone's responses. One problem with AI is that it's able to generate new content faster than humans can consume it. So people taking the time to comment on all the different angles of this "project" results in a good time capsule of thoughts on the current state of things, as well as a way of recording "why it's a bad idea to do things this way."
I found @jjuran's response especially entertaining.
I found @jjuran's response especially entertaining.
I think you actually tested that your code works and does something, though. The purpose of this "project" and the method of its genesis aren't the big problem here, I think. Doing things for personal enjoyment is the purpose of a hobby after all despite how objectively silly it might be.
What I think gets folks' metaphorical goat I think is the utter lack of care.... seems that the OP just submitted a prompt to ChatGPT and dumped the output in a Github repository. Faked up screenshots, glaring errors, completely untested, and certainly nonfunctional. It's a waste of time; the digital equivalent of dumping assorted parts into a box, shaking it, and calling it a working machine. It's a meme with extra steps.
Plus, olePigeon was honest and forthright about his usage of LLMs for creating his app, and he explained why he used it.
There's nothing wrong with using LLMs to do some things, as long as you clearly disclose the fact and thoroughly test that whatever you used it for is sensible.
Exactly.
When someone uses LLMs exclusively to white a complete program and then completely ignore the fact by saying that they did themselves, it feels dishonest and deceptive. Especially when there's no care taken to make sure that said program actually makes sense and works.
For example, there was that so-called "clean room" re-implementation of the classic Mac OS that someone posted about a few months back. It probably would've been much more popular if not for the fact that "clean room" was, at best, a big stretch, because the creator didn't fully acknowledge that they used LLMs, and that said LLMs blatantly stole code from various places, including extensively from the leaked System 7 source code. Definitely not "clean room" reverse engineering in the generally accepted sense, at least as I understand it.
LLMs are a powerful tool. Used wisely, they can do all sorts of wondrous things. But, as the bromide goes, "with great power comes great responsibility."
c
..which is still next-to-unuseable, yet it is surprising it actually boots in qemu (I tested it)!
That's incredible!
However, I bet it's so crash-prone that it makes Copland seem usable and reliable by comparison!
c
I've never tried it, because it seems like a waste of time to try, but from what I remember, there were only one or two Copland builds that actually boot, and there's not a massive amount more they can do. Apple's demo of Copland on the Computer Chronicles wasn't from a real OS, it was fake.However, I bet it's so crash-prone that it makes Copland seem usable and reliable by comparison!
