I wonder if this also apply to transaction costs?
Imagine a 1 nano/transaction fee, it’s like hearing the sound of silence.
Directly after the dubious big bang an atom already had a function in the universe. Does a nano have a function in the SAFE universe directly after the big b 
?
You could keep the smaller CNR if you needed to save disk space! On the other hand, CNRs require you to download some information from the Network before being spend-able. With CN you can create everything offline until you go upload their SignedSpends. This saves network ops, meaning it’s significantly faster to do a transfer and we don’t overwork Network nodes.
If you already got the CNR it means you uploaded the SignedSpends of your CNs to the Network. At that point it’s too late, you can never get your CashNote back (you can’t spend it again).
It is in the CLI! 
If I lose all the CN or CNR information on my computer, can I recover it from the network?
Small enough that if it was uploaded to the network it would be visible to the person storing it, all in a single chunk?
What I am getting at is, do I need to carry my cashnotes around with me as they are not on the network?
I get that they are worthless without the PK but could I encrypt that and store it on the network too?
If we could make dealing with value on safe not a pita then that would change the world of digital currency.
I guess storing them off-network may be a good security measure in that if someone hacked your account credentials, they wouldn’t have access to your cashnotes. If you stored them on the network without some kind of extra encryption, if someone somehow gets your credentials, you could lose it all.
But, storing the cashnotes on your account would be convenient. Managing off-network secure storage as well as your credentials may be asking a lot, especially for small value sums.
Great opportunity for someone who wants to write a wallet app, now who was it that had one running on an earlier API…? 
Great update. Interesting to hear some details of how payments can work. Also promising to hear of some of the changes that should increase speed of uploads and reduce the number of transactions required per upload & reduce resource requirements in general.
As ever, looking forward to the next testnets to see how things are progressing 
Will some info on this be released as part of an update? I’d be interested to hear Jim’s thoughts on what the MVP will look like, what’s needed for it, and Josh’s list of what’s needed technically to get to beta.
Certainly sounds like we’re getting closer to rubber hitting the road time!
There could be a combination of online/offline depending on your personal risk profile.
My thoughts exactly 
It’s a pitty that AutoNAT v2 work on LibP2P seems to be a bit on a pause at the moment. I hope that doesn’t block us from progressing.
Is that holding things up in terms of nodes from home, or a bigger issue than that?
If it is just holding up nodes from home, maybe we could try to provide more info on helping people to get port forwarding working, even recommended routers etc?
NAT hole punch adds lots of complexity and lots of bugs with it.
It may be better to stabilize already implemented features first.
I’m still not understanding this. I think I’d need to see the byte layouts of a CashNote, CashNoteRedemption, and SignedSpend.
In particular, I don’t understand how a node can receive a CashNoteRedemption and verify that it is legit.
You say the size of a CN is variable and can be very large.
What is the practical upper limit and what determines the size of the CN?
A payment to many recipients?
The actual amount of the payment?
A payment from many to one - providing a very popular chunk perhaps?
Sorry I am lost here…
Could be very dangerous. If, in the end, you can’t recover your CN or CNR from the network, you would lose all your funds if you don’t have a backup if, for example, your hard drive crashes.
In this case a standard wallet should look like this.
.-Create an offline cashnote
.-You save the information of that transaction (for example the CNR). Of course in encrypted form
.-You transfer the cashnote
This way you will never lose your funds.
The risk here is losing/revealing your credentials and losing access to your network stored CNs isn’t it. So long as you keep your credentials secure, as normal what’s the risk?
The biggest risk seems to be leaving CNRs or CNs unencrypted at rest, locally.
As ever there will be a security/convenience trade-off.
My own thoughts on a wallet follow pretty much what @digipl said
but with a facility for a small amount of “petty cash” for auto payments that would require little/no user interaction.
I am not claiming these are terribly well developed thoughts yet.
Yes and that’s a privacy risk only, not so much a losing money risk. As you say
Like any system the base credentials are vital. Here with those you can decrypt and create the secret keys you need to transact SNT on the network.
I feel we need to simplify much more here and put the details in an appendix. So get the API solid and explain in a very simple way, those functions folk will use most (Spend, Retrieve etc.). The advanced stuff needs to perhaps be relegated to appendix or advanced section of the docs.
Yes, obvious that CN/CNR are useless without the private key but I forgot that 
, so it really is just credentials (apart from the primacy issue) that need to be secured to keep your cash secure.
Useless to whomever got them without your private key but still a loss for you if they go missing.
Either I have that wrong or I am not understanding your and Davids points.
I am hearing only secure your PK and your value is safe.
I would feel safer with my notes on the network than on a hard drive, obviously provided the network does what it says on the box.
Yeeeeaaahh… there’s talk about MVP, launch etc… but how about the box… has anyone thought about it at all?