Skip to Content.
Sympa Menu

pcplantdb - Re: [pcplantdb] Synth, comments, etc

pcplantdb@lists.ibiblio.org

Subject: pcplantdb

List archive

Chronological Thread  
  • From: Chad Knepp <pyg@galatea.org>
  • To: Permaculture Plant Database <pcplantdb@lists.ibiblio.org>
  • Subject: Re: [pcplantdb] Synth, comments, etc
  • Date: Sun, 12 Dec 2004 15:03:19 -0600

John Schinnerer writes:
> Aloha,
>
> > Bear K. writes,
> >
> > Can I put in a recommendation that there be an API to query this (using
> > SOAP?), and develop that for the back-end sooner then later, so that
> > different front-ends can be built. Discussing the functionality of an
> > API for the back-end could help clarify what the back-end needs to do.

Eden will eventually turn into a low[ish] level python package that
provides a [documented] universal API to any kind of client a
developer wishes to create. One of the distinctions you've noted
already is that the data display/rendering is a front-end client
issue. The distinction between the front and back end in Eden is
currently a little blurry, but I will be making development changes to
make the distinction clearer. Eden will eventually become two things,
the back-end and a default web based front end. When I get to a point
of being able to distribute them as separate packages I will begin
encouraging independent development/extention of the web client.
Adding support for WebDAV, SOAP, xml-rpc and such is goal, but
currently a ways off. This will probably be implemented with either a
Twisted or Medusa daemon running in userspace (hopefully we can do
that on ibiblio). In retrospect one of the nice things about Zope is
that it had this stuff already built in. I used WebDAV with cadaver
quite a bit and really liked it.

This API is partially documented in Phase I goals in the form of what
it will provide. One of it's major roles is an authentication wrapper
around the database itself. It will provide methods to query, add,
and update the database as well as some "helpful" methods like
creating an "English" summary of the plant based on it's attributes.
The later might more appropriately be included in the front-end,
though... In terms of the API itself I was imagining a period of
development where we tried a lot of things before eventually settling
on a stable API. I'm also very open for the API to be extended in
client specific ways or for evolving needs.

> This is already part of the plan...just hasn't been brought up lately.
> Thanks for pulling it upwards...

Something else you mentioned a while ago that I forgot to "amen" to
was the possibility of the data storage model being a peer-to-peer
distributed network. This is a ways off... I am also inspired by the
model of Zopes Zeo and think that client side caching or client side
asynchronous replication might be good methods for keeping the
population of datasets large enough to be stable ;-)


--
Chad Knepp
Perl - $Just @when->$you ${thought} s/yn/tax/ &couldn\'t %get $worse;




Archive powered by MHonArc 2.6.24.

Top of Page