UCDs and arrays

[Markus Demleitner]

Hi Semantics,

On Tue, May 08, 2018 at 11:34:44PM +0200, Baptiste Cecconi wrote:
> I tend to say that stat.distribution would be a better option than
> stat.probability or stat.likelihood. 

Well, I've made it stat.histogram, and I've put in some rationale on 
http://wiki.ivoa.net/twiki/bin/view/IVOA/RFMforUCD

> Markus, please send a request to the UCD working group, as
> described here:
> http://wiki.ivoa.net/twiki/bin/view/IVOA/IvoaSemantics#UCD_list_update
> <http://wiki.ivoa.net/twiki/bin/view/IVOA/IvoaSemantics#UCD_list_update>
> (it will be good opportunity to check the process... :-)

First feedback on process issues:

(a) there should be a clear template for how such a proposal should
look like; the existing proposals are a bit nonuniform.  I've make it
"Use Case", "Proposed Solution" and a section "Discussion" now.

I have a hunch that we should plan for discussions to take place on
the mailing list, though, so I wonder if that last bit shouldn't just
be links into the list archives.

(b) I was severely confused as to which page to put my proposal on.
I've eventually decided the idea is to put it on RFMForUCD directly
and then archive away things now and then.  I like that.  But please
don't use headlines as link anchors to archive pages; they should be
in enumerations.  If I had to do it, I'd write

   * [[UCDList1dot42017June2018FebRFM|Archive: UCD proposals 2017-05 through 2018-04]] (leading up to list v1.5)

and go on like this.

(c) I had a strange mixture of total confusion and reassurance I was
right putting my proposal directly to RFMforUCD when I discovered in
the source of RFMforUCD a div with the old material hidden away with
CSS.  Let's not do that -- twiki keeps your history, so in general
there's no reason to comment things out.

           -- Markus