According to https://bitcoin.org/en/wallets, there are these 2 types of wallets (apart from full node):

Centralized means: "This wallet relies on a centralized service & Server by default. This means a third party must be trusted to not hide or simulate payments."

SPV means: "This wallet uses simplified validation and random servers from a list."

I want to understand: Galleon and Airgap are both in the "centralized" category, right?

This means I suppose if your server(s) is down, user's transactions get stuck, right?

This is a simple obvious single point of failure scenario. Would it not be in everyone interest, to prevent such situation technically? Can you , in addition to your server, also pick a couple of other trusted servers randomly, for validating transactions, like many SPV wallets in bitcoin? So that if one server is down, transaction would still get through?

Is something like that possible and does it make sense?


1 Answer 1


Andy from AirGap here.

I'm not sure which "validation process" you mean. Could you please provide more details about what exactly you mean?

As you mentioned, AirGap currently connects to nodes/servers hosted and secured by us. We plan to add a feature that lets users enter a custom node/server in case ours ever goes down. We might also add a small list that users can choose from.

If you have any other questions, please let us know!

  • thx.@Andreas-Gassmann, SPV stands contrary to that: the wallet provider implements an automatic mechanism by which every transactions is routed to MULTIPLE different (random) nodes in the network (in addition to your own nodes). Look how the bitcoin SPV wallets work. I would like to see that for tezos wallets. Why are you only routing to your own servers? Why not additionally to other trusted nodes of the network to avoid single point of failure? And why won't you implement sth automatic, instead of letting users manually define additional nodes? this is a feature every user would want Dec 17, 2019 at 20:33
  • You are right for the last step of a transaction, the actual broadcasting to the node. This can and be more distributed because even if you have a bad actor all, they can do is ignore it, but it will still be added to the network through other nodes. But the broadcasting is not the only part where we need the node. The tricky part is when you get your balance for example. A malicious node can just give you a wrong balance, so your transaction is based on false data. So for these things, you really have to trust the nodes. Dec 18, 2019 at 14:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.