pcplantdb

["John Schinnerer" <john@eco-living.net>]

Aloha,

> Can I make a suggestion (and give us a bit of homework)?  Can we all
> answer these 4 questions:

Initial thoughts here...

> 1.  What would be want the ultimate plant/guild database to be able to
> acheive?

I would assume it'll be an organic and evolving creation.
If we want to design 'the ultimate' right off we'll either get stuck in
analysis paralysis (or just plain disagreement :-) or be disappointed when
it's not as ultimate as we though or when it takes 'forever' to implement or
other similar design pitfalls (not specific to software design but happens a
lot there too...).

> (i.e. I want to be able to insert x,y,z criteria and design a medicinal
> guild for my local region...)

I don't think of it as something that will do my thinking for me, but as a
resource to draw on for various sorts of information that supports my
designing.  I don't want to be handed a 'solution' - I want support as a
designer.

So for example I might want to search for something like (in English) "humid
tropical climate plants that produce food and do well in high rainfall,
minimal soil situations," or "temperate dryland drought-tolerant root barrier
plants" and so on.  I would know something of the functions I'm seeking and
the context I'd use them in and would be looking for plants that suit.

Likewise if there was a particular plant I wanted for some reason but didn't
know enough about, I'd want to get the vitals to see if I could use it in my
context (soil, moisture, shade/sun, size, growth rate, climate zone, etc.)

Sometimes my searches would be very broad; sometimes they might be quite 
precise.

>Also specify fee (free or paid),

I see it as an open source project - community created, source freely
available under one of the typical open source licenses (GPL or similar).
I have no problem with some money being raised/found to fund development as
long as it remains open source.


> format (web or cd rom)-- how this tool is accessed.

Web - CD's cost money to buy, burn, mail, etc. and they're not very
compostable. Main thing is, if this is a dynamic database they will be out of
date potentially as soon as they're burned.

If a turnkey cross-platform CD can be produced with open source
tools/platforms I wouldn't oppose that - it would be more decisions and more
work is all.  If someone wanted a CD for some reason (bandwidth) they could
download an image once and burn their own and for others.

> 2.  What characteristics of current plant/guild databases can serve as a
> model- features we admire, and would we want to include in a 'next level'
> project?

Haven't seen too many in detail besides PFAF, which has lots of good info and,
importantly, references for that info.

The PFAF searches such as habitat, native range, uses, etc. are very nice, and
of course a full text search is always nice.

Pictures can be good, if they're optional in search results, well taken and
representative, and sized well for low bandwidth users (can always have a
high-bandwidth option too).

> 3.  What are current plant/guild databases lacking, that we would aspire
> to include in a 'next level' project?

They mostly represent plants as a list of information/statistics.  What's the
context?

At least as important, what are the relatings of a give plant with other
plants (and other living systems)?

For example, a guild-enabled DB would include info on what roles the plant
might play in guilds, and PC design in general (root barrier, ground cover,
mulch source, forage, mineral accumulator, alilopathic tendencies and
tolerances, etc.) - including details (how deep/dense/fast growing a root
barrier, how thick/tall/shade tolerant/quick moving a ground cover, how much
cutting for mulch is tolerated, forage for what in what season(s), what
minerals are accumulated, what is exuded and what is tolerated, etc.)

> 4.  only AFTER answering the first 3, brainstorm tech formats/platforms
> that could acheive this end and how.

To me it seems natural to model plants as objects, with attributes and
methods.  I suspect this would enable easier modeling of the actual diversity
of attributes, especially as arbitrary attributes (and methods too if that
proved useful) could be added (along with some that could of course be
predetermined).

Some methods would simply expose attributes (data/information), such as my
examples above.  Some methods could interact with other objects' methods, if
that turned out to assist with guild/design applications.

Example: Looking for understory plants for a given dominant tree - the tree
object would query other plant objects (in this case those with an
'understory' attribute or method) to determine those that cooperated best with
it's own needs (root depth/spread, nutrients, water, etc.).

In short, I'm favoring an object DB because I think it better suits the nature
of what we're trying to look at.

That's the short of it for now...

> DID YOU USE THEIR CLASSIC VERSION OR THEIR NEWER VERSION?  PERSONALLY I
> FIND THEIR CLASSIC VERSION MUCH EASIER TO USE.

I didn't see those options - just a link to the online version, whichever that
is.

> AND REMEMBER- THINKMAP IS
> A SOFTWARE DEVELOPMENT APP (BACKEND), AND YOU ARE DISCUSSING THE *VISUAL*
> INTERFACE (FRONT END).  IF YOU HAVE PROBS WITH THE INTERFACE, IT CAN
> DEFINITELY BE MODIFIED!

Isn't that (the visual interface in general) what struck you as so useful a
representation for a PC plant DB though?  I don't mean the specific shapes and
colors and so on in the thesaurus; I mean the whole floating quivering 3-D
linked way of re-presenting in general.  Graph Based Interfaces appears to be
the generic term...?

> This is rather like asking the waiter if the food at his restaurant is
> good when he comes to take your order.
>
> ACTUALLY, I WAS TALKING TO HIM MORE AS A FRIEND THAN AS A CLIENT WHEN I
> FIRST THOUGHT OF THE IDEA- BECAUSE WE WENT TO SCHOOL TOGETHER.  HE IS
> YOUNG LIKE ME AND I DON'T THINK HE IS AN EVIL ENTREPRENEUR JUST TRYING TO
> GET BUSINESS.  DON'T BE SO CYNICAL :)

I don't doubt my comment sounds cynical.  It's based on my observations and
experience in tech industry.  And, I certainly don't think your friend is
evil, nor that there's anything wrong with him trying to get business, whether
from professional drive or personal passion for the product.  However well it
would or wouldn't work, that it's a proprietary platform is the main
consideration IMO.

cheers,
John S.


John Schinnerer - MA, Whole Systems Design
------------------------------------------
- Eco-Living -
Cultural & Ecological Designing
People - Place - Learning - Integration
john@eco-living.net
http://eco-living.net