Quote from: tvbcof on July 01, 2015, 02:53:01 AM

Quote from: thezerg on July 01, 2015, 02:19:27 AM

Quote from: tvbcof on June 30, 2015, 07:02:56 PM

OK, let's try again. Ever heard of 'packet filtering?'

If you are the engineer you claim to be you know damn well how easy it would be to hide Bitcoin txns from a DPI DSP engine inside a HTTP image request, an audio or video stream, or of course via an encrypted connection.

Please don't post arguments that you know are illogical but you think will convince others. It does not further the discourse.

In my experiments with steganography many years ago, I find it cuts down available bandwidth by orders of magnitude. At that time it was unclear if it was even theoretically possible to hide things fully, although it was pretty clear that one could cause the waste of huge amounts of an attacker's CPU if you made him look.

The utility I used most was 'steghide'. A while ago I looked it up again. The source code was still available but it had not been changed in years. In looking just now I notice that there is a French version of wikipedia which has an entry but I can find no English one and Google translate doesn't work on it. Weird. https://fr.wikipedia.org/wiki/Steghide

Actually, there is a point to be made that individual transactions could be hidden steganographically for spends no matter what the network transaction rate and it is kind of questionable whether the backbone itself could run fully steganographically hidden even with 1MB blocks under dedicated attack by funded and motivated attackers (who own the network.)

Ideally I'd like to see it practical for the backbone network to be operation fully on side-channels which make no use of the global internet at all. In any real attack situation it is very unlikely that any infrastructure operator would enjoy the streaming class of network traffic that we know today. That is another reason I would like to avoid developing a reliance on real-time behavior (or even ~10 min/conf expectations) or structured block formations and so on. The Blockstream folks making mention of expectations in the days range gives me hope that they are designing for worst-case scenarios (which is exactly what I need to feel comfortable about storing significant value in Bitcoin.)

Edit: BTW, I've got very little expectation that encrypted connections without government backdoors will survive indefinitely. No design that I trust for longer duration work will make that expectation.

Its steganography to hide a bitcoin txn from a human inside a hidden channel. But its so simple to hide from a packet inspection engine I wouldn't even call it steganography... for example the "image" downloaded could look like random bits to a person (obviously not a meaningful image) but the packet inspection engine would not be able to determine that.