Re: [OAUTH-WG] OAuth Signature Draft Pre 00

[Mike Jones <[e-mail]>]

Nat had reviewed Yaron's and my proposal and encouraged us to proceed with it.  Yes, it's intentionally general and not OAuth specific (just like SWTs were).  The JWT token type will have uses outside OAuth.  Dirk and we agree that we need to come up with a unified JSON token spec.

Later today I'll write up a set of comments on the differences between Dirk's proposal and the JWT one as they stand today to kick-start the discussions of specifics.

                                                            Cheers,
                                                            -- Mike

From: [e-mail] [mailto:[e-mail]] On Behalf Of David Recordon
Sent: Monday, September 27, 2010 9:24 AM
To: Anthony Nadalin
Cc: oauth
Subject: Re: [OAUTH-WG] OAuth Signature Draft Pre 00

Mike and Yaron's proposal is different from Nat's though. Nat's is based directly around OAuth versus explicitly defining a separate signing mechanism and then a second spec to map it into OAuth. It also supports fewer options (no unsigned tokens for example) which makes it easier to understand within this context. Dirk's now seems to be four specs which then reference Magic Signatures for the underlying signing.

On Mon, Sep 27, 2010 at 9:17 AM, Anthony Nadalin <[e-mail]<mailto:[e-mail]>> wrote:
So we have been working with Nat on the signature proposal and talking to Nat he agrees that the JWT proposal is well under way, what I would like to make sure is that we merged in with your proposal

From: [e-mail]<mailto:[e-mail]> [mailto:[e-mail]<mailto:[e-mail]>] On Behalf Of Dirk Balfanz
Sent: Monday, September 27, 2010 9:13 AM
To: David Recordon
Cc: oauth
Subject: Re: [OAUTH-WG] OAuth Signature Draft Pre 00

I'm just as confused :-) I think what happened is that I posted a signature draft and then didn't follow up. Nat then very kindly agreed to help and put out a draft, but that also didn't get much momentum. So I went back and re-did my drafts. Also, somewhere along the way, Yoran wrote a draft. At least that's what it looks like from where I'm sitting. I might be getting it wrong (maybe Yoran's draft represents a merge of his and Nat's thinking? - I'm not sure).

At any rate, of course we need to end up with one proposal in the end. I'm fairly agnostic about the details, but I believe the following should be true about any merged proposal:

- very limited number of options for signature algorithms, key representations (should not require more than 10..20 lines of code in your given platform, without any additional library, to implement signature and key parsing).
- must support both public and symmetric keys.
- should not have security flaws

Dirk.
On Mon, Sep 27, 2010 at 6:59 AM, David Recordon <[e-mail]<mailto:[e-mail]>> wrote:
I'm a bit confused between the relationship of Nat's I-D and the documents you and Mike recently posted. Is the goal to have one I-D? Nat's seems to have fewer options and different modes which makes it easier to read and understand.

On Mon, Aug 30, 2010 at 11:47 AM, Yaron Goland <[e-mail]<mailto:[e-mail]>> wrote:
BTW, Nat and I, as mentioned below, are talking. Here is my current draft. Please keep in mind that it's really just a set of notes trying to capture all the issues involved in creating a secure token format so it's a bit dense. My hope is that once all the issues are captured it can be completely re-written to be in something that looks more like English and is easier for actual implementers to follow. But for now I think it gives a good sense of the some of the security challenges in creating a secure token format.
                Yaron

From: [e-mail]<mailto:[e-mail]> [mailto:[e-mail]<mailto:[e-mail]>] On Behalf Of Nat Sakimura
Sent: Tuesday, August 24, 2010 6:50 AM
To: oauth
Subject: [OAUTH-WG] OAuth Signature Draft Pre 00

Hi.

It has been a few weeks since then I volunteered to do this work.
I have written up to this pre 00 draft then have been doing some reality checks on some script languages etc.

No. This pre-00 draft is far from being feature complete.
I still need to copy and paste the Magic Signatures text etc.
Also, I should add how this spec is being used in some of the major flows.

However, since I will not be able to work on it this week, I thought it would be worthwhile to share this early draft so that you have some clarity into the progress.

Apparently, Yaron has been working on it as well. We will compare the notes and try to merge, I hope.

So, here it is!

#For those of you who have seen the private draft, it has not been changed since July 31.

Best,

=nat



_______________________________________________
OAuth mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/oauth


_______________________________________________
OAuth mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/oauth