[r-t] CCCBR meeting

[Don Morrison]

On Tue, May 31, 2016 at 3:42 AM, Mark Davies <mark at snowtiger.net> wrote:
> However, I think the way forward will be easier once the methods
> committee have direct ownership of the method libraries.
>
> So this is my first task. I want a new methods website with shiny
> new features and webservices, hosted centrally, i.e. not under the
> control of one person. I want to design it so that different
> protocols for method classification and extension can be plugged in;
> it should be possible to say "what did methods look like with the
> classification rules in force in 1967?" and "what do methods look
> like with this new classification protocol I've just invented"?
>
> Of course there ought to be the standard lookup features, including
> some convenient RESTful webservices for apps to use. I'd like to
> populate the database automatically from the incoming Bellboard
> feed. One of the other methods committee guys was saying, wouldn't
> it be cool if we could do some visualisations, like live mapping of
> the geographical distribution of where new methods are rung?

I strongly advise you not to let the fun, technical stuff take precedence
over what is far more important: complete, accurate content, and clawing
things back from a nearly proprietary web site to something more under
community control, where a well-functioning, openly communicating methods
committee is a reasonable proxy for "community". I think last year's
taking-my-bat-and-ball-and-going-home experience highlights how important
it is to get the first things done first.

Really the first order of business ought to be to get something, publicly
available, on a non-proprietary web site. And then, as quickly as possible,
get the content right by making it complete, filling in the lacuna
deliberately created by years of foolishness.

Such a first pass does not need to have the snazziest wiz bang interface.
Even something as basic as one or more static text pages should be fine, so
long as the format is regular enough that it is easily parsed. The current
text format is not really satisfactory, as it is so irregular and complex
that it's quite tricky to parse accurately, especially over all the
possible classifications of methods. Better to have something more regular.
The current HTML version is much better in that regard, but doesn't even
include everything the Central Council found acceptable even before
yesterday's meeting; and, of course, the HTML version isn't really human
readable, at least not by most humans.

Making it complete, and doing so sooner rather than later, is far more
important than making pretty.

Only once it is on a non-proprietary web site, and is as complete as
practical (to first approximation; there will always be improvements to be
made to the content going forward, of course) should effort be put into the
fancy technology stuff. If instead you start down the technology path, it's
going to impede the more important part and we won't have anything
satisfactory for years, if ever. Far, far better to get something complete
and public as quickly as possible, even if ugly, and then incrementally
make it wizzier. Spending year's designing something does nothing for the
ringing community. Building something, even if initially inferior to what
you might design, is a real service.

It may be worth noting that there has been talk about pulling the method
collections out of proprietary-land for over a year, with the claim that
this is a top priority, and, as near as I can tell, absolutely nothing has
happened to make that a reality. Do not perpetuate this same mistake. Be
more concerned with making *something* happen than making the best
possible, or even just a pretty good, thing happen. Never forget that
better is the enemy of good.



-- 
Don Morrison <dfm at ringing.org>
"The historian is miserably subject to the brevity of time and human
patience, and must dishonor with a line men who were immortal for a
day, but now lie hidden between the peaks of history."
              -- Will Durant, _The Age of Faith_
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://bellringers.net/pipermail/ringing-theory/attachments/20160531/68c83937/attachment-0002.html>