[TESTING] Curve25519 encryption support (update-2)

[Daniel Kahn Gillmor]

On Thu 2015-08-06 10:07:33 -0400, Werner Koch wrote:
> On Thu,  6 Aug 2015 10:29, gniibe at fsij.org said:
>
>> I pushed the patches by two commits.   "make check" works fine
>> with libgcrypt 1.6.3 and patched libgcrypt.
>
> Works well, thanks.  There is one aesthetic thing I would like to
> change:
>
>   sec   ed25519/C68CE6D1ED0319C8 2015-08-06
>   uid               [ultimate] Curve25519 Test 150806.1
>   ssb   crv25519/49238B9F0712C9BF 2015-08-06 crv25519
>   ssb   rsa2048/8AEAF74014699D2C 2015-08-06
>   ssb   crv25519/8EC3776830B08736 2015-08-06 crv25519
>   
> the columns are still not aligned nicely.  Given that the most common
> keys are RSA and DSA keys we would be abale to get a prettier output if
> we change the alias for Curve25519 to just "cv25519":
>
>   sec   ed25519/C68CE6D1ED0319C8 2015-08-06
>   uid               [ultimate] Curve25519 Test 150806.1
>   ssb   cv25519/49238B9F0712C9BF 2015-08-06 crv25519
>   ssb   rsa2048/8AEAF74014699D2C 2015-08-06
>   ssb   cv25519/8EC3776830B08736 2015-08-06 crv25519
>
>
> What do you think?  Shall I change it?

I like the alignment, but now i would wonder about why cv25519 shows up
before the / and crv25519 shows up at the tail end of the line, and yet
neither of them are the official "Curve25519" name.  Remind me why we
need both, and both need to be shortened (differently from one another)?

    --dkg