gingeropolous: Besides the privacy, there's no advantage to steganography unless we 'successfully' make it cheaper. It's computationally expensive to prove and should be more expensive to do.

It shouldn't be more expensive because TX extra should be cheaper than outputs. It should be more expensive because steganography isn't perfectly efficient like TX extra.

So while yes, computational bloat is possible, I believe the comment is TX extra has to be attractive for arbitrary data

As a side note, output steganography isn't even the most efficient possibility is. CLSAG `s` is AFAICT

You can do it immediately, without prep, in a single TX. It appears identical to any other TX on the network and lets you do just two outputs. Whoever receives the message can eliminate decoys, yet because we're increasing the ring size to 16, if we only have 5 decoys used for messages it's solely a reversion to current privacy for whoever receives the message.

That gives you ~150 bytes, which can be used for ~500 characters with a compressed minimal charset. It has no computational cost nor does it add storage.

You also don't lose out on decoy selection distribution because instead of saying the first 5 are messages, you can use the encrypted amount to specify which were used in steg.

I'm not sure how that transforms to Grootle though, and the discussions on that post were largely about post-Seraphis theory.

*you do have to encrypt it with a competent algorithm before hand to produce a uniform distribution from your non-uniform data. A 37 charset takes 5.2 bits. You naturally get 281.84 bits. I'm assuming a zstd dictionary created in advance from the english dictionary can achieve a 50% ratio.

" Whoever receives the message can eliminate decoys," ... i might need to draw this out

s/*/\*/, s/bits/chars accordingly/

Because the receiver is using a `s` value as a message, they know it wasn't a real `s` value which must meet a mathematical formula

Accordingly, it must be a decoy. Accordingly, they can eliminate a candidate from the ring as your real spend.

My comment is that with the new hard fork, solely reverting to current privacy (and solely to the recipient) gives you ~500 chars of a-z 0-9 space per input without computational overhead and no additional storage

right, yeah that part makes sense

there's also no way to identify such TXs because they'd appear completely standard. they'd be 1/2 or 2/2 (unless quantum computers enable flagging them, no promises there, blockchain messaging is stupid)

lulz

I wouldn't assume similar methods are possible post-Seraphis though with Grootle. That's why I commented on outputs, which will still offer it :p

roight roight roight

I believe so long as TX extra isn't more expensive, even if it's as expensive, TX extra will be used by everyone who doesn't need such privacy. Any who does need such privacy should not be using a blockchain.

So while I appreciate your intent, I do believe it's not supporting both, so long as we don't push people to steg insteadf

s/insteadf/instead/

" Any who does need such privacy should not be using a blockchain." ........ so why allow tx_extra. I feel like my ramblings are more suited for lounge at this point.

Never depend on a single income to get financial freedom. I can teach you how you can earn from $5,000 to $6,000 weekly,without sending money to me or anyone, but you will pay me 10% commission when you receive your profit. if you are interested, click on the link to join the chat room 👇

Just to work with their favorite privacy coin

<gingeropolous> "" Any who does need such privacy..." <- Because the usages discussed don't need privacy.