When you think you're clarifying yourself, you're actually getting less coherent. I can't even parse what you're saying in your deeper replies (and I design mixed-signal audio systems for a living).
You're doing the same kind of "affect wisdom by alluding to the elephant without mentioning the elephant" hand-waving that conspiracy theorists do, but as yet you haven't shown a meaningfully concrete example (other than your misapprehension that the same sort of implementation-specific inconsistencies that are found in floating point calculations exist in the fixed-point math used in most audio systems).
Look. If I just made the .1+.2 example in passing by as how code doesn't always behave like we expect, does that mean I'm claiming we even use addition to work with sound? If everyone misunderstands a claim I haven't made, and asks me to explain that claim, how exactly do I respond?
It was an example I don't know how audio works and don't claim I do. But I do know moving bits doesn't always end up the way we expect, and different functions has different outcomes.
I shouldn't have to explain sort and sort-reverse has different outcomes, it's implied and not my point. All I'm saying is don't call people without domain knowledge idiots when they're on the right track, kind of. The rest are strawman arguments I'm supposed to explain. I'll give you an example:
> ME: 1s and 0s in the same order are the same, but some data may be lost in the transfer
> Poster: Please explain? If the 1s and 0s are still in the same order, what data has been lost?
What I say: 2 files 1010 and 1010 are the same at start, but might be 0101 at the destination. How is irrelevant, there are a million ways.
What I'm asked to explain: If 1010 is 1010 at destination, how are they different.
Then I mentioned transfer, oh-oh. Now we're going into transport protocols and not just moving bits no matter the means. Incredibly frustrating and ironic given the context. I see the misunderstandings, I just can't seem to explain it. It's not hand-waving I just don't get why I have to explain things I never claimed.
This seems to boil down to a claim that when transfering data from A to B then B may be different from A. No, it won't, and that's the point. Data transfer via a reliable protocol, e.g. TCP (and when using UDP you simply apply a framework with its own error correction, as e.g. OpenVPN UDP does). I mean, in my job we transfer petabytes of data.. if B wasn't equal to A.. no. What can, and does, happen is that the data at A or the data at B could experience bit flips typically due to cosmic radiation - this does happen, which is one reason ECC RAM is better for some situations, but if this happens during transport then the protocol will handle that and you still get B = A.
> Then I mentioned transfer, oh-oh. Now we're going into transport protocols and not just moving bits no matter the means. Incredibly frustrating and ironic given the context.
How is it still happening. Did you read the post? Do you see the irony?
Could you please stop posting in the flamewar style and otherwise breaking the site guidelines? you've been doing it repeatedly, unfortunately, and we have to ban that sort of account.
> ... I don't know how audio works and don't claim I do
OK
> All I'm saying is don't call people without domain knowledge idiots when they're on the right track, kind of
What makes you think you're on the right track?
> But I do know moving bits doesn't always end up the way we expect
> What I say: 2 files 1010 and 1010 are the same at start, but might be 0101 at the destination.
Your assumption is wrong. Moving bits around is very well understood, and uses error checking where needed. 1010 does not turn into 0101 at the destination.
> How is irrelevant, there are a million ways.
Your explanation is inconsistent with the knowledge and experience of millions of engineers and users, so it's up to you to explain how.
I'm not claiming I am. Read the first comment I made. I'm simply stating the quote "bits are bits" is misunderstood because the one who made it doesn't have domain knowledge. I'm discussing in general terms but people expect details and proof of concepts.
I'm not arguing anything, everyone is reading me as if I was. Then I'm supposed to explain how TCP will fail and so on. Never even mentioned TCP, I said transfer.
People are too literal. But it's fine I've given up trying to explain it, this discussion has turned into the special olympics of misunderstandings. And you're all winning, congratulations.
> Your explanation is inconsistent with the knowledge and experience of millions of engineers and users, so it's up to you to explain how.
Since you insist, explain what, exactly? Quote the claim you want an explanation of.
> Since you insist, explain what, exactly? Quote the claim you want an explanation of.
Things you said:
> But I do know moving bits doesn't always end up the way we expect
> What I say: 2 files 1010 and 1010 are the same at start, but might be 0101 at the destination.
I'd like to know what makes you say these things. The reasoning behind it, or maybe experimental data.
When audio data is digital, it works in exactly the same way as all other digital data, something computer scientists, electrical engineers, network engineers, etc. happen to have a lot of knowledge about and experience with. We know how it behaves, we know bits don't just change. Therefore I say it's up to you to explain how bits can move in unexpected ways, and how they can be different at the destination than at the start.
You can't just say you're discussing in general terms as a defense for using unfounded assumptions.
That's my entire point. I never claimed anything. I'm told I did and asked to explain all these strawmans. I see the title attracted all the "actually" people in the world.
Look at my first post. I'm trying explain how I read "bits are bits" and that it doesn't necessarily mean that he's stupid. Are bits always bits?
You're doing the same kind of "affect wisdom by alluding to the elephant without mentioning the elephant" hand-waving that conspiracy theorists do, but as yet you haven't shown a meaningfully concrete example (other than your misapprehension that the same sort of implementation-specific inconsistencies that are found in floating point calculations exist in the fixed-point math used in most audio systems).