Re: More clocktable breakage

[Achim Gratz <Stromeko@nexgo.de>]

Nicolas Goaziou writes:
>> I'd say anything org-element-* should exclusively return syntactical
>> things.  Context dependence needs to be dealt with elsewhere.
>
> I'm not sure to understand this. Syntactical things are all about
> context dependence in Org. Do you mean /context independance/ needs to
> be dealt with elsewhere?

No, I meant context of application, rather than context in the
syntactical sense.  Org-element-* deals with syntax, nothing else.
Whether you need strict syntactical interpretation or something else
gets decided someplace else.

> Could you provide examples about that? In particular "providing
> timestamps as arguments to any processing functions" sounds odd, in
> particular from someone who cringes when I suggest to add a second
> optional argument to a single function.

Whatever it sounds like, what you want in the clocktables example and
the properties example (and elsewhere) is something that looks, walks
and talks like a timestamp if you'd put it into the proper context.  In
each of these places the text that looks like timestamp but isn't
(org-element says so) will be fed into some machinery that emerges with
a result that is indistinguishable from what you'd get if that text
would have been a proper timestamp and then uses that result to do
whatever it wants to do with it (i.e. most certainly not build up an
agenda, although it could do that as well).  It uses a bit of Org syntax
in the improper context to achieve this (and this requires precisely to
ignore that context or at least check with something more loose than
org-element).

>> If I'd follow that road, I could edit what looks like a timestamp
>> everywhere, regardless of context.  I don't know if that's the right
>> thing to do and I don't even expect consensus among the Org user base.
>> I personally see no need to do that.
>
> I do see it, tho. This is analogous to links in comments. If you see
> something looking like a timestamp (or a link), you can expect some
> commands to operate on it. This is exactly what is biting us at the
> moment.

In a comment that timestamp-looking text doesn't have any function, so
it's in a different category, I must insist.  As I said, I can see
somebody wanting to have this text be editable like any other timestamp
also, but it's really the other uses where it's used meta-syntactically
that I'd like to focus on.  ONe of the differences to text in comments
(or generally quoted material) is that there is an expectation that this
sort of timestamp is correct, since they are intended to be input to
further processing.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs