pcplantdb

[Chad Knepp <pyg@galatea.org>]

Hello folks,

Responding to various and sundry:

John Schinnerer writes (6/6):
 > <snip>
 > 
 >  From a functionality standpoint I don't find much that's been done so 
 > far worthwhile, with the possible exception of where Bear has started 
 > exploring.  I can get more plant information faster using Google than 
 > using current PIW.  As I posted some time ago based on an ad hoc case 
 > study drawn from a question on a permaculture list, PIW is currently 
 > pretty useless for even basic plant searches.

John Schinnerer writes (6/7):
 > <snip>
 > 
 > You don't need to defer at all.  My whole point is that everydays users 
 > aren't going to get much from PIW at present.
 > You can go to the current PIW and try some searching yourself and see 
 > what you get.
 > Unless I enter a known latin name or something very simple like 'plum', 
 > all I mostly get is hundreds of hits that seem to have very little to no 
 > relevance to what I searched for.

Wow, I'm surprised and a little hurt by this evaluation of the current
version (0.3.1) of PIW on <http://dev.permaculture.info/>.  While your
evaluation is highly accurate of the current live version I think very
significant progress has been made on dev.  Specifically the advanced
search addresses most of your concerns
<http://dev.permaculture.info/advanced.html>.  Many of the features
that Ad Vlems listed are available as well. 

Ad Vlems:

 > I want to integrate three ways of searching:
 > - simple (search for part of a plantname)
 > - complex (make a selection combining all possible fields)

Both of these are done with the addition of the simple search being
rather smart in being able to interpret (google like) the text of what
is entered such as common name, botanical name, habitat, and other
characteristics and make reasonable guesses as to what is wanted.  For
specifics the advanced search is necessary.

 > - question directed, in which the next questions can be selected:
 > * I want to grow a plant that can be harvested in August.
 > Here you can change the month and add the kind of soil/sunlight in 
 > your garden.
 > * I want to create my own plantlist (that I planted in my garden), 
 > make an inventory of them and get a list of gardentasks per month 
 > (when to sow, harvest, etc)

A garden log/notebook for individual customization is also in my long
term feature goals.

 > * I want an edible plant that will bloom yellow in May.
 > Here you can change color and month.

Our current dataset is missing some of the information that would be
important here as well as locale support to universalize bloom times
for other climates/hemispheres.  

 > * I want to attract butterflies, which plants should I plant?

Exists in the dataset partially (polinated by butterflies), but
ability to query this is currently unimplemented.

 > * Which plants will grow in a pond?

<http://dev.permaculture.info/advanced.html> check the box pond/lake
under Suitable Locations.

 > * Which plants are good companionplants?

One of the major goals of this project and incomplete at this point.

 > When one of the questions is selected, you get a list of plants. When 
 > you click the plantname, you get the detailed info.

Exactly including a relavance score as to how close the result matches
your matches your query.

 > Could you give me a list of what your website would have for visitors 
 > to profit? Maybe we can help each other.

Done.

 > Your's truly, Ad Vlems

Additionally I would like to add that the dataset is completely
editable.  This is fairly significant IMO.


John Schinnerer writes (6/7):
 > <snip>
 >
 > What needs to be done is to organize this like a 'real' project (whether 
 > commercial or open source).
 > That means data scoping and design and normalization and functional (and 
 > other) specifications and all kinds of stuff that was barely or mostly 
 > never done.
 >
 > Otherwise it will remain a handful of well-meaning folks hacking 
 > together bits of often very clever stuff with no coherence and little 
 > real-world usefulness.

I disagree.  The current PIW dataset is very close to fully normalized
in the Codd sense of the term.  We chose to work with a pre-existing
design which was still a good choice I think because of the wealth of
the PFAF dataset.  The server API is fully documented on
<http://dev.permaculture.info/api/>.

More fundamentally, although I think that there would be significant
benefits from running this project as a 'real' project I'm not sure
that this is the best approach for PIW in that it is largely a
volunteer project.  The funding received from Threshold would have
been used up before anything was implemented, and prior to Threshold
we spent about a year and a half 'designing' with pretty much no
results.  I think that our hacking together bits does have and will
have more real-world usefulness as we continue.

Sean Maley writes (6/7):
 > --- Josh Shupack <ynaught@gmail.com> wrote:
 > > Hi, my name is Josh Shupack.  Bear invited me to join the group.  I've
 > > actually been on this list for many months and read through a lot of
 > > the archives.  I like to see what I'm getting myself into before I
 > > jump in.
 > > 
 > > I am a freelance web programmer working part time and am interested in
 > > getting involved.  I was kinda sitting back and seeing what was going
 > > on, but now it looks like I could be helpful.  I'm traveling around
 > > the country this summer visiting intentional communities and friends
 > > with my wife, so I won't have much time until we get back in Berekely
 > > in August, but at that point I will have a lot of time to put toward
 > > this project if I can help.
 > > 
 > > I have never programmed in Python, but am always happy to pick up
 > > another language.  My idea would be to get a functional skeleton up as
 > > fast as we can and get people using it and slowly add features.  I
 > > think there are some great minds and great ideas on this list and the
 > > main thing is to keep rolling and get people using it.
 > > 
 > > Let me know what you all think.

Josh, look at the existing code and see what you think (esp. the
server which is quite elegant).
<http://dev.permaculture.info/eden.tgz eden_db.tgz and eden_api.tgz>
You will never regret learning Python and will probably never code in
anything else after (if you have a choice).

 > > Regards,
 > 
 > Why should we stick with Python?  Shouldn't we take an inventory of
 > skills and choose accordingly?  PHP can be insecure, but should be
 > sufficient with proper care taken by the programmers.
 > Alternatively, Perl with Templates wouldn't be difficult.  However,
 > Mediawiki has been implemented effectively which is PHP.  If we aim
 > to extend Mediawiki, much of the interface is complete with simply
 > adding markup features (programmer can choose Python, Perl, C or
 > Assembly, if they want).  This also makes it possible to divide and
 > conquer by assigning features/markups and taking measure on which
 > programmers are more reliable for providing completed features.
 > Where the markup language of a wiki isn't always accessible to a
 > computer novice, a separate data entry path can assist with data
 > entry (popup page?  Firefox extension?).

The current software running on dev is implemented in the classic
client/server split.  This decision was made in order to create
opportunity for multiple clients to access the same dataset.  Since
the server implements XML-RPC access of all of it's methods, anyone
anywhere can write a client in any language you mentioned and access
the dataset.

Scott Pittman writes (6/7):
 > <snip>
 > 
 > I would like to bring this conversation back to the beginning for a moment:
 > 
 > The idea for this data base came out of a conversation I had with 
 > Stephanie a couple of years ago when I was lamenting the lack of good 
 > "guild" information in permaculture circles.  I can only recount a half 
 > dozen examples for our arid climate in New Mexico.  
 > <description of guilds snipped>
 > 
 > Stephanie then decided to devote time to raising money to put together a 
 > guild data base and was successful.  We were also sucessful in 
 > recruiting several programmers who helped launch this whole process and 
 > brought it to its current state.  I think that we have accomplished a 
 > lot and would be loath to give up when it seems we are in sight of the 
 > goal.
 > 
 > I have met botanists and biologists around the world who were captivated 
 > by the guild concept and they have a lot of knowledge, from studying 
 > plants in their native habitat, about guilds in the natural 
 > environment.  Russian biologists come from the Vavilov tradition of 
 > looking for the origins of all comestibles - potatoes from Peru, corn 
 > from southern Mexiico, and apples from Kazakstan.  If one had the 
 > opportunity to look at apples in the forest of Kazakstan the evolved 
 > guild there would become apparant, and from that we could develop 
 > analagous structures for the human environment.  What I had hoped to do 
 > was enroll my international contacts to add their accumulated wisdom to 
 > the data base once it was formed, I still hold out that expectation.

I'll reiterate that I think the possibility of software being able to
both contain/understand all the possible relationships between species
and extrapolate a unique and workable guild of many species starting
with one or more species is an extremely difficult task... namely the
contain/understand part because a tremendous amount of data gathering
and entry will need to happen first in order to capture these
relationships.

The reason that I was not listed in the original proposal is that I
did not agree that a 'guild building' tool should be our primary
focus.  That said I think that much of the framework for gathering
this data is in place, and as it is today PIW is EXTREMELY useful for
figuring out what plants might work well with others, esp. if you use
the advanced search and a bit of thought.  4 years ago I thought that
the PFAF database was great for this...

 > I will have to defer to other programmers on the list to guide me as to 
 > whether this project is worthwhile or not, I certainly hope so!

I think it highly worthwhile, that it is useful now, and I also think
people are mostly just upset that nothing is happening and are
forgetting the significance of what we have done already (or haven't
actually looked at it).

 > I would appreciate more feedback from those of you who know how data 
 > bases work and what still needs to be done.  If we had a short list of 
 > what has been accomplished and what needs to be done I think it would 
 > help us all.

Hopefully this answers some of these questions.

Onward:

It seems like both my absence and my reluctance to participate in the
near future has really upset folks.  Ok, I can rethink this; esp. on
hearing the cry of 'start over' which pains me to think about.  I was
initially going to propose offering a day a week to correspond and do
some billable work but I'm not sure I can make it every week.  I can
offer 4 days a month.  Would this satisfy folks?  If not I'm willing
to give up being lead developer if someone else steps forward that the
group approves of.

Cheers sort of,
Chad

-- 
Chad Knepp
python -c 'import base64;print base64.decodestring("cHlnQGdhbGF0ZWEub3Jn")'