[Cbor] Re: to chairs: reject draft-edn-25 WGLC

[Christian Amsüss <[e-mail]>]

Hello Vadim,

(now with chair hat on:)

> [...] while time for -26 is expiring [...] / and time on wallclock
> says it's too late to work on this section before Monday (Carsten
> could use this time constructively but stolen from himself).

The Monday draft is not the last chance for things to go in: it is the
basis on which we can have a better structured way of addressing WGLC
comments.

> Instead of discussing valid technical arguments from mine (and not
> only mine) messages, he skips them and replies with same
> unconstructive "obvious" or "we don't need"

So far it seems to me to be a working technical discussion in the latest
threads. There are examples of stopped exchange: at least some of which
he indicated are for lack of clarity of the proposal (which not always
means writing more text to elaborate, but to focus on single concrete
issues), in others, that appeard to be implied. If I've missed instances
where that was not the case, please collect the relevant links to
[e-mail].

> * remove ellipsis "..." and tag 888
> * remove tag 999
> * remove string concatenation syntax with +
> * remove string encoding indicators (EIs) like "s"_2 or ''_ as 8949/8610 are
>   silent on them anyway - these are not normative yet
> * introduce optional EIs for app-extensions as orthogonal dimension,
>   like extname_1's'
> * introduce tstr'' and bstr'' extensions to handle both concatenation and
>   indefinite-length strings (names are from CDDL and could be changed, e.g.
>   text''/bytes'' or even t''/b'' to save user typing in t_2'short' or even
>   both as aliases, for readable and short variants)
> * remove ambiguous syntax (_ 'foo' "bar") for indefinite-length string as well
>   as ""_ and ''_ exceptions for them
> * introduce pragma/version mechanism, details not matter but /^#pragma.*$/
>   regexp preferred, because it must be a syntax which is syntax error in
>   7049/8610 EDN (and somewhat distinguishibly visible to human so e.g.
>   #pragma<<...>> is better than pragma<<...>>)

Thanks, that is a useful list. There is currently ongoing discussion
between cbor-chairs on how precisely we'll structure next steps, but it
will probably involve a per-suggestion structure, and having a list of
current ones is helpful (as are the per-proposal threads you have
started).

Best regards
Christian

-- 
Beware paths which narrow future possibilities. Such paths divert you
from infinity into lethal traps.
  -- Leto Atreides II