incremental evolution as a process (was Re: [uf-discuss] Microformats.org usability review)

[Tantek Ç elik]

On 1/29/06 9:36 PM, "Mark Rickerby" <maetl at mcs.vuw.ac.nz> wrote:

> I have been working away at this over the past week, and I'm starting
> to realise that simplicity and minimalism is sometimes more tricky to
> get right than complexity and overabundance.

<snip>

Mark,

First of all, thank you for writing up your review.

There are many good points and good suggestions for improvement in there,
and I look forward to seeing a bunch of them implemented.

Your review made me think about some more aspects of what we are doing as a
(still) fairly new community which have largely been implicit, but deserve
some explicit mentioning.

As far as how did we get here, literally what you see with both the
microformats site and wiki is are the expanding minimum steps to enable more
and more folks to understand and implement and/or use microformats, within
the time available to contribute that everyone working on the site has had.

When presented with an opportunity to spend five minutes improving something
or writing something to enable a few key developers and publishers, vs.
spending days writing something to enable many more, we have chosen the five
minutes tasks.  This is actually quite different than classical publishing,
or standards work in many ways (I and a bunch of folks here have some
experience with both).  But these differences are not accidental, there are
actually very good reasons for them.  A few:

 1. Breadth. There are so many such five minute tasks, that if you spent
days on any one of them you would get very few things done overall.

 2. Adaptive Evolution. It is much better to spend five minutes with an
incremental improvement, and get potentially get some feedback quickly and
make corrections quickly, than to spend days on something only to find out
it wasn't what people were looking for.

 3. Distribution of work.  Many more people can work on many more small five
minute tasks, than can on any number of multi-day tasks.

 4. Lower barrier to entry/participation/contribution.  When anyone can
visit a page and spend only a few minutes improving things, rather than
having to spend days, orders of magnitude more people will help out.

A lot of what we've written up in the microformats process is specifically
designed to provide those interested in pursuing microformats with, for lack
of a better word, a series of bite-size "microtasks" to complete when they
can.

For these reasons and more, I suggest that any improvements try to take as
much of the incremental approach as possible.  Any improvements that you
would think might improve all of a certain type of document could perhaps be
done on one first to see how it goes, and adapt and adjust from there.

Now, not everything can be done as a 5 minute task, but we do try to
minimize such tasks to only those that are absolutely essential.  Drafting
new specs is among them.  Designing the site at the beginning was as well.

However, at this point, I'm very much convinced that such tasks are (and
very much should be) the exception rather than the rule.

One thing I have found that helps a lot with tasks that seem to require more
time than a five minute chunk, is to add them to the to-do page, and iterate
on them, breaking them down into smaller tasks.

As far as where to go from here, I'd definitely recommend that you add a
section for yourself on the to-do page and move a bunch of the documentation
you have for what you want to improve there, to collect the thoughts and
allow some iteration.

 http://microformats.org/wiki/to-do

Some of the suggestions you made will be really easy to do quickly, and
others will require additional thought and discussion (I didn't think it
would be effective to use email to discuss/debate all the assumptions etc.
that went into various suggestions - I think the wiki will work better for
that).  Depending on how much depth you go into, it may require additional
pages.

I hope some of this background and philosophy is helps provide understanding
and context for what we is microformats today, and I hope to see your
outline of improvement suggestions.

Thanks again very much for your review and suggestions.  I'm eagerly looking
forward to the improvements.

Tantek