pcplantdb

[Richard Morris <webmaster@pfaf.org>]

>  > Q2 Version control
>  > 
>  > I feel DB would need sort of version control system.
>  > At the very least to stop some script kiddy just deleting
>  > all the data. We get quite a few 'You Suck' messages added
>  > to our readers comments so chance of corruption is high.
>
> /. <http://slashdot.org/> is a great example of how to handle
> registered users and anonymous ones, self moderate, and allow the good
> information to filter to the top.  The "You Suck" posts could easily
> be moderated into deletion.  Moderation only happens by registered
> users.

There is a big difference between slashdot and what we are
trying to do. slashdot is a readers forum where comments
are added. It's essentially an 'add-only' style where
information can only be added. If we are looking at a more
wikii style general edit system, moderation becomes rather
tricky. I suppose we could allow unregistered users to
add data and trusted users to edit/delete. I still feel
if edit/delete is allowed then we need to keep a
history of changes.

In my wild dreams about the
project it would be very cool to have some sort of moderation
system, (voting for favorite plants etc),
but this would involve the same sort of structure
as for atomic attribution.

>  > Q3 Edit style
>  >   I've mentioned before that I see some problems with
>  >   allowing free edits. Problem is especially with
>  >   medicinal uses in that some one could add some false
>  >   info to DB with possible health risk. Attribution
>  >   and only allowing additions could help protect
>  >   the projects name.
>  > 
>  > 	In my ideal world I'd really like some review process
>  >   so that edits are checked by a review board to
>  >   preserve accuracy.
>  > 
>  > 	On the reflection this is a bit over the top.
>  >   And wikii style free edits is probably OK.
>  >   (as long as there is version control and some way
>  >   of tracking changes, & restoring backups)
>  > 
>  > What do people think of these issues.
>
> Sound like you're answering your own question here.  See also my
> response to Q2.

Thinking aloud.

> Richard Morris writes:
>  > > 
>  > > So Richard, I must confess I'm feeling a little pissy about this whole
>  > > license deal.  I don't feel that you've been taking me seriously.  I'm
>  > > glad you decided to respond to me email.  I'm disappointed that it
>  > > took my releasing PCPDB in order for you to do it.
>  > 
>  > I don't quite get what your wound up about.
>  > I've been trying to workout a way that pfaf would feel
>  > happy with contributing their data to the project.
>  > This involves getting the license right from the ground up.
>  > Averyone seems to think that our current copyright statement.
>  > is full of holes so it just seems a bit odd to use that.
>
> Well starting sometime in early Feb. I initiated an email dialog
> urging a change in the license.  Your responses came about every 2
> weeks falling off to not replying at all to my last message prior to
> releasing PCPDB.  I got wound up because I felt like you were ignoring
> me. 

Sorry about lack of replying. Problem is I've got
a lot of balls to juggle at the moment, and I have a lot
of problem keeping on top of my email (managed to get inbox
down to 350 1/Jan now risen to 484). I've also been immersed
in my own maths computer project which has sucked a lot of time 
recently. Email response time is inversely proportional
to amount of thinking I need to do and PCPDB take a lot of thinking.

>  > Why not just get it right from the start!
>
> Releasing PFAF under the original PFAF license does not seem like much
> of an attempt was made to get it right from the start.
>
>  > > Since PCPDB is based on the November 2000 release of PFAF, I believe I
>  > > am within legal rights by conforming to and displaying the license
>  > > contained within that distribution.  I have closely examined and in
>  > > the past I have encouraged the adoption of the CC by-nc-sa by PFAF.
>  > > One of the reasons I was encouraging the adoption of a *real* license
>  > > is because the PFAF one is legally way fucking abusable.  
>  > 
>  > SO why use it!
>  > 
>  >  > Read the
>  > > first sentence of this paragraph again.
>
> Well, one reason to use the old license, is because it has no clause
> restricting sub-licensing such as paragraph 2, 4 of the GPL.  I
> essentially can decide to release a derived work, under any license so
> long as the new license doesn't violate any clause of the PFAF one.
>
>  ><big snip>
>  >   
>  > We will be contributing an upto date snap shot of the pfaf db
>  > with the CC license. It will take me a bit of time before
>  > I can do this (doing a new release of the DB takes about a day).
>  > 
>  > If you wish you can get the Feb 02 release from
>  >   ftp://ftp.comp.leeds.ac.uk/pub/pfaf/ascii.zip
>  > it contains scripts for easy import into sql. Archive still
>  > contains old copyright notice.
>
> That sounds good, except for the part about still releasing under the
> old license.  I would stop that practice ASAP.

No, I mean the archive still has old liscence as I've
not got round to changing it.

> Also, when you say we, who are you talking about?  Is there really
> much interest in this project.  I've been watching the weblog pretty
> closely and have been completely underwhelmed by the response.  Not
> sure how many unique users but only about 100 page requests since I
> posted the release announcement.  I'm seriously questioning if there
> is a need for this kind of project.  It's been fun so far and I've
> learned a lot, but currently, playing with high availability clusters
> seems like more of a growth edge for me.  Essentially I need to hear
> that people want this.

As for interest in project, I can give a few examples from pfaf
web site stats.
        UK database get 15,000 page impressions per week.
        US mirror gets a similar amount.

        In the last year we have had about 150 people add
comments to webpages/database. About a third of those contain
some sort of relevant plant based info.

And thats with the rather rusty current design which does not allow 
people to add new plants.

OK so 50 relevant additions may not feel like much
but it does help justify the time I spent adding that
feature. I had the same feeling as you when we only got
one comment a week when the comment section was added.
But time passes and if now feels like a respectable
body of information has been added.

I do get quite a few emails from people who have been
compiling their own datasets. With the right interface
both programatically and legally some of those might
be willing to merge their data into the project.
If we got just one of those a year then that would be
a great benefit to spread of plant knowledge.

I know mikel was enthusiastic about a year ago and seemed
to have plans to contributing more info.

Really I see see pcpdb as a way of future proofing the pfaf
dataset. I'm really looking to ten years in the future
where input from pfaf will be minimised, I'd like to keep
the dataset alive for prosperity.

So far the db has not been widely publicised and still
very much in beta, so hits are limited. Heck 100 hits
is aprox 20 people who have sufficient interest in the
project to check out an early version.

There is such a vast array of different people who might
be interested, we have only started skimming the surface.


> From: BK <lildragon@saber.net>
> Reguarding attribution, the concept of tracking every single change and
> crediting would seem to bog the DB with more info then about actual plants.
> It would seem useful if there was a credits list, each unique user who added
> info would be appended to the list, and based on the number of
> changes/Kilobytes of additions their ranking on the list would rise towards
> the top. With some sort of limited explaination of the changes.

Sound OK.

> I'm definitely interested in the DB, and I would be cool if it could somehow
> have extensions into other PC realms (designs, guilds, PC sites of
> interest). I've been really wanting to sit down and play with PHP and SQL
> and try setting such things up, but haven't yet gotten around to it. I too
> wonder how many folks would use the DB, it would be something quite helpful.

There are a lot of issues in making the DB really work.
Potentially it could be a great resource getting to the
top of google when people do a search for a plant.
A lot of work needed to get there though.

        Rich
--
Plants for a Future: 7000 useful plants
Web:   http://www.pfaf.org/ same as http://www.comp.leeds.ac.uk/pfaf/
Post:  1 Lerryn View, Lerryn, Lostwithiel, Cornwall, PL22 0QJ
Tel:     01208 872 963 / 0845 458 4719  
Email: webmaster@pfaf.org
PFAF electronic mailing list http://groups.yahoo.com/group/pfaf