New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Issue32 ssh setup.3 #57
Conversation
6803c34
to
edef282
Compare
whoa, did you really get the full test suite to work on pypy? I thought the C code in pynacl was a blocker for that.. |
I think it's just building the bundled libsodium and using CFFI to call it. But yes everything passes on PyPy with the one change for Deferred |
I like it. I want to nit-pick about the implementation, though :). Could you force-push a new branch with the following changes?
I haven't looked closely at the actual thanks! |
Okay, I can mess about with Obviously, it would be good to have more eyes on the |
Let's see:
|
As per #32, a name like |
|
I agree about root stuff. How about: if you appear to be root then:
(Basically I guess that is just "if you're root, you have to specify --user= and it can't be root). Or, just bail out right away with "don't run this as root; switch to the target user first". |
For
|
Re
And if you're logging into your own account and want to add a new key (maybe you used a password the first time, or maybe you logged in from one host and want to add the pubkey of a second host), you do the same:
If we think the add-key-to-my-own-account is more likely, then we'd not require the posarg, default to modifying your own Re pubkeys: yeah, that plan sounds fine. The only thing I'd add would be to print out the list of discovered pubkeys, if there's more than one. And I think I'd be ok with leaving out a special check to avoid sending root's pubkey (doesn't seem dangerous to me). I might call it For extra credit we could make it interactive (print a list, assign a number to each, ask the user for the number), but maybe that's too much work. Unlike text/file sending, the sender doesn't need to confirm the transfer. But the receiver does (at least if |
edef282
to
df163df
Compare
Okay, I force-pushed after rebasing and doing the things in #57 (comment) |
df163df
to
dfb956e
Compare
okay, i made some of the cleanups in the last couple messages from @warner in the last commit on this branch. I moved to "wormhole ssh accept/invite" sub-sub commands, --user, changed method-names, more permissions checks, ask user which key (if there's multiple keys), ... |
bde8009
to
38d636e
Compare
- move to 'wormhole ssh' group with accept/invite subcommands - change names of methods - check for permissions - use --user option (instead of --auth-file) - move implementation to cmd_ssh.py - if multiple public-keys, ask user
38d636e
to
fe29b31
Compare
Current coverage is 86.57% (diff: 46.15%)@@ master #57 diff @@
==========================================
Files 18 18
Lines 2694 2718 +24
Methods 0 0
Messages 0 0
Branches 379 380 +1
==========================================
+ Hits 2342 2353 +11
- Misses 257 270 +13
Partials 95 95
|
Looks good: I made a few tweaks, and I have a list of additional fixes to make later, but I'm going to land this now. Thanks! |
Here's a bit more sketched-out version of the idea of a top-level
send()
andreceive()
API...This PR also takes out the "just invoke 'wormhole send'" part of the other PR.