[postgis] Next Level Redux

[Paul Ramsey]

Implementing the Web Feature Server is something I have been thinking
about for a while, and there are two ways of thinking about it. 

One is the way Frank is probably thinking of, which is to use the WFS as
an abstraction layer for a bunch of different data sources. This makes
great sense for his OGR library, which is already an abstraction layer.
A WFS based on OGR gets all the OGR-supported data sets for free.

The other way of thinking about it is as an abstraction for a single
data source: a WFS which serves up all the GIS tables in a PostGIS
database. This seems highly redundant, but there is a reason to grant it
consideration. There are performance advantages to specialization, and
maintenance advantages to abstraction. 

I think a middle ground is probably going to be the best option in the
end: for things like "spatial databases" in general, the data retrieval,
querying and locking problem will have a particular solution. And for
"file based" data sets, the query, retrieval and locking problem will
have a different solution. The two problems can either be stuffed
together into one system or two WFS servers can be built.

I am also interested to hear what people think about WFS architectures:
long-running processes, or one-shot CGI scripts? Probably the best
possible performance would be from an Apache module, but that might be
alot more work than is necessary, given that the parsing overhead for
the client dealing with GML (!) results will probably vastly outweigh
the retrieval/transformation overhead for the server.

Of course, any WFS implementation will have the same problem which
PostGIS has now: no clients! :) 

Rob Hranac wrote:

> 
> I don't know if this is helpful or not, but we seem to have similar goals.
> I can't offer money, but I can offer development services.  I am keenly
> interested in developing an open source WFS with database support and,
> clearly PostGIS is the best candidate for this.  Any thoughts on this?  I am
> not sure if this is what you mean by getting PostGIS up to OpenGIS
> compliance or are you interested in the Simple Features Specification for
> SQL.  Either way, let me know what you think.  I have been talking to Frank
> Warnerdam about this, who I believe is an associate of yours, if I am
> reading things correctly.
> 
> Cheers,
> Rob Hranac
> 
> -----Original Message-----
> From: Paul Ramsey [mailto:pramsey at refractions.net]
> Sent: Tuesday, August 14, 2001 1:05 PM
> To: postgis at yahoogroups.com
> Subject: [postgis] Next Level Redux
> 
> I would be happier flying into the group in my Bell JetRanger with the
> 'Ride of the Valkyries' playing... however, the news is not all that
> triumphant:
> 
> I have done the rounds of (almost) all the Canadian federal contacts I
> have or have had provided, some prominant people in the local provincial
> government, and have not been able to turn up a partner for a proposal
> to GeoInnovations for this funding round. (For those not in on the
> original newsflash, GeoInnovations is a Canadian Federal Gov't program
> to promote geomatics R&D. It provides a 50% cost-sharing of selected R&D
> projects. If we put in a proposal, we would have to either cover the
> remaining 50% ourselves or find a partner to take up some of that slack.
> We were looking to cover 25% ourselves and find a partner for the
> remaining 25%. The total dollars we needed to fund our $CA 120K project
> were therefor about 30K from outside sources. I should also note that as
> a company we are only 5 people, and have very limited resources to
> self-fund this. The only reason we are getting so much work done lately
> is because paying consulting work has been slow and we have idle hands:
> our pain is the community's gain. :) )
> 
> Both Red Hat and Greatbridge are currently hoarding cash. Red Hat to try
> and remain profitable, and Greatbridge to try and remain solvent. The
> upshot is, they were not willing to be one of the project partners. Both
> were open to pursuing GIS extensions in the future in happier times
> however.
> 
> Anyhow, the gate is closing on GeoInnovations, since there are only two
> weeks left until the proposal closing date, so I fear we will not be
> able to take advantage of that particular opportunity. From a magnitude
> point of view (compared to the kinds of dollars dropped on things like
> dead simple Oracle db implementations), getting PostGIS/PostgreSQL up to
> 99% OpenGIS compliance is an astonishly small project: under $100K US
> dollars. Somewhere, somehow, there is an organization with a commitment
> to PostgreSQL that needs/wants that functionality. We will be working
> our way there, however slowly, over the next 12 to 18 months anyways,
> but keep your fingers crossed that someone wants to get there
> faster/sooner.
> 
> From the front,
> Paul
> 
> To unsubscribe from this group, send an email to:
> postgis-unsubscribe at yahoogroups.com
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
> To unsubscribe from this group, send an email to:
> postgis-unsubscribe at yahoogroups.com
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

-- 
      __
     /
     | Paul Ramsey
     | Refractions Research
     | Email: pramsey at refractions.net
     | Phone: (250) 885-0632
     \_

To unsubscribe from this group, send an email to:
postgis-unsubscribe at yahoogroups.com

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/