Re: [Extra] Some comments on draft-ietf-extra-imap-64bit

[Alexey Melnikov <[e-mail]>]

Hi Stephan,


On 15/09/2017 17:06, Stephan Bosch wrote:
> Hi,
>
> I gave draft-ietf-extra-imap-64bit a quick look. I have a few comments 
> on the syntax specification.
>
> First of all, the elements fetch-att, msg-att-static, and search-key 
> are just copy-pasted from RFC 3501. Those are meant to have number64 
> instead.
Yes, fixed.

> I wonder whether it would be a good idea to have intermediary elements 
> called e.g. `msg-size = number64' and `msg-lines = number64', so that 
> future changes/extensions to the message size and line count syntax 
> aren't such a pain anymore.
Might be a good idea. If you send me the specific ABNF fragment, this 
will speed up the process.

> The "capability" syntax element is not extended (arguably not all 
> specifications do this).
This is typically done in prose.

> Finally, the following IMAP capabilities have syntax that is affected 
> by this extension:
>
> QUOTA - https://www.ietf.org/rfc/rfc2087.txt
Good point, I will have a look.
> BINARY - https://www.ietf.org/rfc/rfc3516.txt (not only literal8)

Looking at the ABNF:

    append         =/  "APPEND" SP mailbox [SP flag-list]
                       [SP date-time] SP literal8

    fetch-att      =/  "BINARY" [".PEEK"] section-binary [partial]
                       / "BINARY.SIZE" section-binary

    literal8       =   "~{" number "}" CRLF *OCTET
                       ; <number> represents the number of OCTETs
                       ; in the response string.

    msg-att-static =/  "BINARY" section-binary SP (nstring / literal8)
                       / "BINARY.SIZE" section-binary SP number

    partial        =   "<" number "." nz-number ">"

    resp-text-code =/  "UNKNOWN-CTE"

    section-binary =   "[" [section-part] "]"



I don't think we need 64bit section numbers, but I missed updating 
"partial".

> URL-PARTIAL - https://www.ietf.org/rfc/rfc5550.txt -> 
> https://www.ietf.org/rfc/rfc5092.txt (partial-range)
> APPENDLIMIT - https://www.rfc-editor.org/rfc/rfc7889.txt
Yes, these need to be looked at as well.

Best Regards,
Alexey