Re: [kitten] Small question about token in draft-schmaus-kitten-sasl-ht-09

[Florian Schmaus <[e-mail]>]

On 13/01/2023 18.29, Alexey Melnikov wrote:> I am reading the 
description of token in your draft:
> 
>     HMAC() is the function defined in [RFC2104] with H being the selected
>     HT hash algorithm, 'cb-data' represents the data provided by the
>     selected channel binding type, and 'token' are the UTF-8 encoded
>     octets of the SASL-HT token string which acts as a shared secret
>     between initiator and responder.
> 
> This made me wonder why is the token defined as UTF-8 string and not just as binary data?

Out of convenience because the accompanying XEP would return the token 
as String [1]. Of course, we could have the XEP return base64 encoded 
binary data, but this appears like an additional hoop in the XEP. But 
now we have the hoop in the I-D, so one could say the it just switched 
places.

What also contributed to this design decision is that I prefer tokens to 
be strings (potentially even restricted to [A-Za-z0-9-]). Simply because 
it makes handling and debugging easier compared to arbitrary binary 
blobs. But that may be just my personal preference and I like to hear 
other opinions too.

- Flow

1: XEP-0397 § 4, https://xmpp.org/extensions/xep-0397.html#obtain)