KyPeN
100+ Head-Fier
- Joined
- Oct 5, 2003
- Posts
- 357
- Likes
- 29
Software engineer here.
You're gonna have to detail, and I mean detail, what "dedicated file copier software means." If I use Windows standard copy within Explorer, a Python script that copies the file over directly, or a C++ program that copies the bits over directly via its own (let's assume custom) buffering mechanism, they're all going to be calling into the same exact operating system API that interfaces with motherboard controller and harddisk firmware to copy blocks. The higher level code mechanisms don't really matter here, because the underlying code in the hardware is doing the heavy lifting and everything is using the same APIs the OS provides.
A proper analogy would be that if you have a multilingual person flipping light switches. If you tell him in English to flip a switch over, I tell him in Spanish, and a third person tells him in Japanese, he flips the switch exactly the same, regardless of how we told him, from where, or using what language.
Either the bits come out the same or they don't. Provide a SHA-1 hash of all files. Are they the same? Same data. Move on with life. Period. If you're going to discredit that, you better have SERIOUS data to back it up, because it would be invalidating not just audiophile "man, this sounds better when I copy over with your special software" scientific research, but also biological, industrial, and literally EVERY OTHER field that uses computers to copy data and assume it's lossless. Lossless is lossless, period. If you can provide data, and I mean HARD DAMNED DATA, that discounts that, I'm all ears. Otherwise, nope.
EDIT: Just a couple updates:
1) It looks like SHA-1 has been deprecated and SHA-2 or SHA-3 is more recommended for cryptography purposes. Any of these should be more than sufficient.
2) Can we take a moment to appreciate how impossible it is to actually validate the claims here? If any of us download the files, do a comparison, and prove they're identical, OP can simply attempt to invalidate that whole debate by arguing that the download over the internet was not a valid copy under the test requirements, because of the assertions in his initial post. You've just got to be there. Trust me. Yeah, sure. Like I said, I demand HARD data or this is so debunked it was never bunked in the first place.
You're gonna have to detail, and I mean detail, what "dedicated file copier software means." If I use Windows standard copy within Explorer, a Python script that copies the file over directly, or a C++ program that copies the bits over directly via its own (let's assume custom) buffering mechanism, they're all going to be calling into the same exact operating system API that interfaces with motherboard controller and harddisk firmware to copy blocks. The higher level code mechanisms don't really matter here, because the underlying code in the hardware is doing the heavy lifting and everything is using the same APIs the OS provides.
A proper analogy would be that if you have a multilingual person flipping light switches. If you tell him in English to flip a switch over, I tell him in Spanish, and a third person tells him in Japanese, he flips the switch exactly the same, regardless of how we told him, from where, or using what language.
Either the bits come out the same or they don't. Provide a SHA-1 hash of all files. Are they the same? Same data. Move on with life. Period. If you're going to discredit that, you better have SERIOUS data to back it up, because it would be invalidating not just audiophile "man, this sounds better when I copy over with your special software" scientific research, but also biological, industrial, and literally EVERY OTHER field that uses computers to copy data and assume it's lossless. Lossless is lossless, period. If you can provide data, and I mean HARD DAMNED DATA, that discounts that, I'm all ears. Otherwise, nope.
EDIT: Just a couple updates:
1) It looks like SHA-1 has been deprecated and SHA-2 or SHA-3 is more recommended for cryptography purposes. Any of these should be more than sufficient.
2) Can we take a moment to appreciate how impossible it is to actually validate the claims here? If any of us download the files, do a comparison, and prove they're identical, OP can simply attempt to invalidate that whole debate by arguing that the download over the internet was not a valid copy under the test requirements, because of the assertions in his initial post. You've just got to be there. Trust me. Yeah, sure. Like I said, I demand HARD data or this is so debunked it was never bunked in the first place.