Re: YA proposal to bring /TR/ into the 21st century

On 21/10/2013 19:25 , Arthur Barstow wrote:
> Speaking of making /TR/ more useful, here's YA proposal ...
>
> Instead of TR/<shortname>/ resulting in a specific snapshot, it is just
> a short document with a set of links to the document's various
> incarnations and the user can then go to the version of choice. F.ex.:
>
> <ul>
>   <li><a href="github/.../">Latest Editor's Draft</></li>
>   <li><a href="TR/YYYY/FPWD-shortname-YYYYMMDD">First Public Working
> Draft</a</li>
>   <li><a href="TR/YYYY/WD-shortname-YYYYMMDD">Last Published Working
> Draft</a</li>
>   <li><a href="TR/YYYY/LCWD-shortname-YYYYMMDD">Last Call Working
> Draft</a</li>
>   <li><a href="TR/YYYY/CR-shortname-YYYYMMDD">Candidate
> Recommendation</a</li>
> ...
>    <li><a href="TR/YYYY/REC-shortname-YYYYMMDD">Recommendation</a</li>
> </ul>
>
> (and use "Not Published" when applicable)

Allow me to phrase this in the dulcet tones that so typically 
characterise the suave diplomat that I am.

No no no no no no no no no no no no no no no no no no no no no.
No no no no. No. Just no.

The ACLs on this list mark it as appropriate for all audience; were that 
not the case every "no" above would be three syllables long.

This is a terribly, terribly bad idea. It is so bad, it makes things 
really, truly, horribly worse without bringing in any measurable 
advantage. In fact, I am compelled to say:

 
http://w3cmemes.tumblr.com/post/34763367921/standards-negotiation-is-the-art-of-sending-people

If /TR/shortname/ returns something that isn't immediately useful, 
people will pick another one, some the last in the list, some the first, 
others whatever seems like the best option. And then they'll copy and 
paste that, and reference it everywhere. Essentially, they'll be forever 
stuck on a dated version without ever going back to /TR/shortname/ to 
see the updated version. It successfully binds together all that is bad 
about dated snapshots while removing all that's useful about /TR/shortname/.

To put things differently: I think we shouldn't do that.

-- 
Robin Berjon - http://berjon.com/ - @robinberjon

Received on Monday, 21 October 2013 19:54:43 UTC