mSIGNA full node?

[luke-jr]

According to bitcoin#390 (comment), mSIGNA can use any remote bitcoin node, so I'm not sure why it's classified as a full node itself...?

[gurnec]

There is a distinction, however it's a bit subtle (see #645).

I think there's a general "grade wallets based on their default and most common configuration" rule, which is why e.g. GreenAddress desktop gets a checkFailTransparencyRemote score. (It has a Google Chrome App available for install which would earn it a checkPassTransparencyOpenSource score, however because it's also available via a website, it's likely desktop users will just use the website therefore it's graded based on that.)

Perhaps the distinction should be made more clear? Unbold and/or change the "Full Validation" text?

[harding]

I think the concerning part of the comment @luke-jr linked to is, "mSigna a highly-secure lightweight wallet that is not dependant on a complete local copy of the blockchain (like Armory)"

Their getting started guide says, "By default, mSIGNA will attempt to connect to a bitcoin node running on localhost (i.e. a local running instance of BitcoinQt). [...] If you would prefer to connect to a remote bitcoin node, you can change the IP address and port under Network -­‐>Settings [...] mSigma only connects via the peer-to-peer protocol and does not require any special RPC access to other nodes."

That means our text---"This wallet requires you to install full node software that validates [...] transactions"---certainly needs to be changed. Maybe we should create a new category of "Full Validation Compatible"?

@CodeShark, @saivann, @crwatkins --- comments appreciated.

[crwatkins]

We are running into more wallets with many different modes (e.g. some wallets can function in both plain HD and multisig with a server) such that a single scoring cannot apply to all modes of a single wallet. The same issue occurs when wallets can work both with and without a dedicated hardware component. I predict that in the near future we will be establishing multiple scoring for the same wallet working in different modes. Until then, I think we will have some scoring that is not totally accurate in all modes. As long as the wallet will actually operate with the scoring we list, I'm OK with that for now.
In general, I'm hesitant to create new scoring or we are likely to approach as many scores as wallets.

[saivann]

Personally, I'd be fine if the text was adjusted to "This wallet uses a full node software that validates [...] transactions by default". I think this would also reflect Armory correctly.

But sure, the increasing diversity of wallets are making them harder to categorize and compare. I agree with @crwatkins that it's acceptable if scores do not cover everything in the short run. Like @gurnec mentioned, thus far I thought the most appropriate and simplest solution was to reflect what the user gets by default.