[uf-new] hGive: Microformats and Microphilanthropy

[Benjamin West]

Greetings Mike,
First, this is the right place for your proposal.  Nice.

>
> These are the kinds of things that microformats could make possible. We
> at DonorsChoose have been talking about applying microformats to our
> proposals (each has its own page) to make them semantic, but none of the
> existing microformats seem to fit what we're doing. I also recently
> talked with Tom Williams at GiveMeaning who is also interested. I'm
> posting this here as what I hope will be part of a larger conversation
> about microformatting microphilanthropy.
>
> Ideally we would have a microformat "hGive". This would allow
> organizations that are seeking contributions / in-kind donations /
> volunteers to use it, as well as organizations/people who are looking to
> volunteer, donate, etc. (I'm thinking of online volunteer clearinghouses
> such as New York Cares which exist in most US cities I believe.)
>

Sometimes people approach microformat technology with a great use
case, notice that none of the formats have their particular use case
listed, and proceed to suggest a new format.  However, formats are not
designed with a one-to-one relationship to use-case in mind.  Existing
formats can be used in new techniques to satisfy new use cases never
intended by the authors of the original format.  Documenting a
technique using existing formats is a successful goal in the
microformats community, and we find that most people can accomplish
very creative things by re-using formats.  In addition, it's usually
much faster to get things done when we re-use previous work.

> Here are some of the potential parameters:
> [snip]
> I'm sure there are other possibilities / desiderata, especially around
> volunteer projects (one time vs ongoing, group vs individual, etc) but
> this is what comes to mind.
>

Perhaps.  Be careful of defining properties too early.  Both
properties and structure for new formats exhibit emergent behaviour
following analysis of pre-existing examples.

> If there were an implementable standard, I'm pretty sure I could get
> DonorsChoose to start using it in the nearish future. And then, of
> course, Utopia Ensues.
>

I'm not convinced a new format is necessary, without first trying to
re-use the work already available to us.  The process outlined at
<http://microformats.org/wiki/process> is designed to make it
difficult to create a new format, and is designed to encourage new and
innovative techniques for re-using available formats.

Here are some suggestions:

Create a wiki page outlining your use case, but don't call it hgive.
Try calling it microphilanthropy, or just philanthropy.  Start
discussing the various datatypes necessary.  Break things apart from
one large microformat into smaller chunks of data.  For example, a lot
of your description involves the concept of a person (a people data
type).  Hcard is perfect there.  Treat the current microformats and
related technologies (including semantic html) as building blocks for
a technique that satisfies the use cases you outlined.  Another data
type is listing of skills: the resume microformat might be perfect for
that.

You mentioned specifying directionality as a requirement, but I think
it'd be a lot simpler if you designed a technique/protocol that always
allows consumers to search for offerings.  For example, instead of:
>     * Jennifer has an old computer monitor she wants to donate to a
> local school. She calls a few schools but no one needs a monitor. She is
> able to search for local organizations that are seeking computer
> monitors, and finds a nearby women's shelter that is in need, so gives
> it there instead.
>
Try:

Jennifer has an old computer monitor she wants to donate to a local
school.  She finds a website that allows people to list things they
own as donations.  When she enters her items on this website, it marks
it up using hlisting <http://microformats.org/wiki/hlisting>, and
embeds it in an hatom entry.  It's also rel-tag'ged with
microphilanthropy, and the women's shelter is able to find it through
the syndicated feed they signed up for, or the next time they log into
the site.

However, I just noticied that hlisting includes the concept of
directionality as well, so the women's shelter could also list their
need using hlisting, tag it as microphilanthropy, and the website
could ping both parties when it finds that match.

Anyway, that's the advice.  Try using existing microformats as
building blocks: rel-tag to glue things together, hlisting for
products, hresume for skills, hreview for opinions, hcalendar for
times and events, hcard for people, hatom for syndication.  At the end
of the day, I bet this will get you 80% there very quickly, and the
rest can be filled in by describing some small additional semantic
html techniques.


Thanks,
Ben West