Thanks for the informative discussion about CSV. My apologies for falling behind on my own question.
It’s not my spreadsheet, and I don’t know how it works—however, yes, I think that
every spreadsheet in the world can export to CSV. It’s the lingua franca of tabular data. SQLite’s shell tool can also import CSV, which I myself find more useful.
I don’t know about this particular person’s workflow—but merely in my own opinion, I think that CSV would be a much better interface than copy-pasting for SME business use with numerous bulk payments.
Unfortunately, my reason for looking at Sparrow in the first place has been obviated. I’m looking for a user-friendly Silent Payments sending and receiving wallet, to recommend to people who are
not me.
I’ve been learning more about SP and SP lightwallet protocols. Thereupon, I cannot in conscience recommend anything that uses the Sparrow/Frigate strategy.
Sparrow reveals the SP scan key to the server. (Not the spend key: The scan key.) Frigate only processes these keys in memory, and does not store them—but there’s nothing to stop Chainalysis from making a small patch to Frigate, and setting up a hundred GPU cloud honeypot servers that log absolutely everything.
The only defense against this is to train users
against using this type of protocol. Tweak servers like Blindbit don’t get any information that could be used to spy on your SP coins.
Revealing the scan key is much worse than revealing ordinary Bitcoin addresses within a wallet’s gap limit, which is bad enough. The scan key can be used to see all past, present, and future payments to an SP address, for all possible labels—
forever. It’s a static, long-term
encryption key; entrusting it to a server sounds to me like entrusting your PGP private key to a webmail server. If your webmail server unverifiably
promises to process your PGP private key only in memory
(pinky-swear!), is that secure email?
If you want to stop the accruing privacy damage of connecting to an untrusted Electrum server, there is no way to undo the past—but for P2WPKH (etc.) addresses, you can protect your future privacy simply by bumping your wallet past its current gap limit
(don’t forget change addresses!). Then, remember to skip the gap limit again if you ever need to restore from seed... Anyway, you don’t need to make a new wallet. Whereas with SP, if you have revealed the scan key, the only way to prevent future spying is to make a totally new SP address. This adds to the cost of SP scanning, as you’ll still want to scan the old address in case someone sends funds to it.
Nontechnical users (and even most technical users) have a poor understanding of privacy threats. I think that the Electrum protocol ecosystem runs entirely on hopium, blind faith in untrusted parties, and wishful thinking. Ignorance is bliss; if you never know which public server logged what, why should you care, right? This is worse with SP, as it will result in a
false sense of security with the equivalent of
infinite address reuse. People will behave differently, if they
believe that blockchain analysts can’t trivially link all of their coins under common ownership. Worse: Their counterparties will behave differently, with neither choice nor informed consent in the matter.
And it’s all so unnecessary. There are lightwallet SP scan strategies that reveal
nothing to the server, at only moderate load to the client. It’s not a situation where dangerous levels of trust are the least-worst of bad options for many practical use cases. SP can and should be always fully trustless.
Please don’t say, “Well, people can run their own personal Electrum/Frigate servers.” That’s not the reality of >99% of the lightwallet user base; security and privacy must be designed accordingly, lest one create an ideal operating environment for honeypots. Oh, and as for the people who say, “Trust my Frigate server not to log your scan key; I’m an honest community member who definitely doesn’t want to remember your key,” sorry but not: That makes me trust you less.