You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
And then I was wondering about passwords. Are there any requirements or limitations on valid passwords? Considering these need to be interoperable this might make sense. For example, should spaces be allowed? Invisible characters?
I don't want to overly limit ourselves here (emoji are for example a fun alternative to passwords), but some sort of specification would be good to improve things like regex detection of codes.
If we come to the conclusion that passwords can be anything, it would at least make sense to add this to the spec.
The text was updated successfully, but these errors were encountered:
I absolutely would love to see experiments with emoji or other things, so wouldn't want to limit too much via spec
There are some concrete use-cases to consider as well:
different dictionary/word-lists (including "weird" things like emoji) should remain possible
humans are allowed to simply make up / pre-agree on a password (and nameplate) before any interacting with any software (so weird arbitrary limits that make this hard should be avoided)
file-transfer is an important use-case, but not the only one
We should move towards specifying something about how we expect nameplates and passwords to look like.
We have this open, which should be specified here: magic-wormhole/magic-wormhole-mailbox-server#45
And then I was wondering about passwords. Are there any requirements or limitations on valid passwords? Considering these need to be interoperable this might make sense. For example, should spaces be allowed? Invisible characters?
I don't want to overly limit ourselves here (emoji are for example a fun alternative to passwords), but some sort of specification would be good to improve things like regex detection of codes.
If we come to the conclusion that passwords can be anything, it would at least make sense to add this to the spec.
The text was updated successfully, but these errors were encountered: