Re: [Gen-art] Gen-ART Last Call review of draft-ietf-json-i-json-05

[Tim Bray <[e-mail]>]

On Fri, Jan 23, 2015 at 10:16 AM, Meral Shirazipour <
[e-mail]> wrote:

> Nits/editorial comments:
>>
>> -Please spell out acronyms at first use. E.g. JSON, I-JSON
>>
>
>  ​I disagree.  People know what JSON stands for, and for those who don’t
> spelling it out wouldn’t help.  De facto, the name of the format is JSON.
> Also, the mention of JavaScript is a red herring.​
>
> I agree for JSON, but I-JSON I had to look it up.
>

​In http://www.tbray.org/tmp/draft-ietf-json-i-json-05.html#rfc.section.1
it says: This document specifies I-JSON, short for "Internet JSON".   So I
think we’re probably OK here.​

>
>  ​I think this has been discussed elsewhere on teh thread.  It’s
> irritating that 64-bit integers would be string-encoded even though most
> computers can handle them in hardware, but it is a real plus for
> interoperability.  Anyhow, the place where this advice really applies is
> for​ huge crypto integers.
>
> yes, that part of the text was not clear to me. For a standards track RFC
> it may be worth taking another look at it.
>

​Actually, following on this discussion, I agree. The text originally
referred to huge crypto integers and was revised to “64-bit integer” at the
request of the WG.  But now I see that that is sort of counter-intuitive
because it assumes the subtext of ​“yes we know that most computers can do
64-bit ints just fine but don’t do that because stupid JavaScript”.  So I
think the draft needs to either fill in that backdrop, or go back to using
huge crypto ints as an example.   I’m inclined to put in a few words of
backdrop.