Re: GSoC 2018 - Modern cryptographic algorithms to netpgp, netpgpverify

When you read the answers below, I can understand if you think I'm being difficult.

But this is how we can work out whether you've understood the task, and have thought about how you'd go about it.

On 20 March 2018 at 18:03, Harsh Khatore <khatore.harsh.github%gmail.com@localhost> wrote:

On Tue 20 Mar, 2018, 9:34 AM Alistair Crooks, <agc%pkgsrc.org@localhost> wrote:
Yeah, netpgpverify is the new, all-in-one, no pre-reqs codebase solely for the verification part of signatures.

ed25519 also needs to be added to netpgp, which is the older and more crufty code base which covers signing and verification.

So, as netpgpverify is the new code base for verification part, netpgp will be used for only the signing part or for both as it used to do?

See the project specification - both are needed.

But before any code is touched, we'd need to know what gpg constants uses for these algorithms, since they're not in RFC 4880, and so we can interoperate with gpg in verifying and signing.

We can get the ed25519 specifications from RFC8023 and see for the constants but I have a doubt as to what are these constants that you referred?

Look further.

Think of the job that netpgpverify and netpgp do, and how they interoperate with other software.

We need to know what extra parts are needed (from different sources, along with their licences), and any other prereqs we might need for both netpgpverify and netpgp.

I am not able to get what do you mean by extra parts and prereqs, can you explain, please?

Think of other things that will be necessary to accomplish the task.

In summary, I'd like to see what you think will be necessary. I'd like to see the thoughts that go into how to accomplish a design and programming task, what tests are necessary, and what to be wary of in the implementation. Not in email, but in the proposal that you write - how you will accomplish (and surpass) the goals.

You (and anyone else) should now have enough information -- from the project specification, and from previous email -- to generate a proposal for the project.

Regards,

Alistair

PS. One copy of email is enough, thanks :)

On Tue, Mar 20, 2018 at 12:13 AM Alistair Crooks <agc%pkgsrc.org@localhost> wrote:
Hi Harsh,

I've been talking to others about it, but yours is the first mail I've received.

C proficiency is necessary. C++ not needed.

I can help you out with any specific questions you have - please mail them here (i.e. to tech-crypto, CC me).

Thanks,
Alistair

On 18 March 2018 at 10:57, Harsh Khatore <khatore.harsh.github@gmail.com> wrote:
Hi Alistair,

Sorry for contacting you soo late for the above-mentioned project. Could you provide me with help regarding the project so that I can work on it for GSoC? Also, do you have anyone else preparing for it or can I continue with this?

My knowledge of C and C++ languages is intermediate.

Thanks,
Harsh Khatore