[postgis-users] ESRI GDB format into PostGIS

[Alan Keown]

> That much?!!
Well, maybe not that much, but a bit. The reason I'm watching is because I'm
hanging out to be able to afford an FME licence.

> Shall we start a survey on what people in this forum think? 

I think that would be a great idea.

I've fully designed and specified four 'enterprise' spatial databases and a
dozen or so large, complex 'application' ones. None have been implemented as
designed because the spatial experts (yes 'ESRIlites') all end up saying
"ArcMap will cark it."  (BTW: I still believe that there is nothing to stop
you implementing a multi-geometry column table in a geodatabase; you just
have to manage it yourself because the ESRI application infrastructure
won't. Which, I think is your point. I say 'believe' because I haven't been
able to get anyone imaginative enough to prove its scalability in a 'real'
environment.)

People like the model simple because it makes the application of it simple.
Lots more work, but simple. ("Waddaya mean 'I should write a query for my
map layer'? Make me a new table!") Gee - I really am getting crankier as I
get older!

So, back to the issue: I spent most of the 90's working on Data Interchange
and ended up completely disillusioned. One interchange format will never
cover all the bases. The dream now is for an Open Spatial ETL suite.
(GDAL-OGR-FDO with muscle.)

None of this helps Ben with his problem but I do think it points to an
answer for your question " why do we waste our time looking over our
shoulders?" - Because they're the elephant in the room.

Cheers, and thanks for your insights on the workspace.xml

AlanK


-----Original Message-----
From: postgis-users-bounces at postgis.refractions.net
[mailto:postgis-users-bounces at postgis.refractions.net] On Behalf Of Simon
Greener
Sent: Tuesday, 2 June 2009 4:53 PM
To: PostGIS Users Discussion
Subject: Re: [postgis-users] ESRI GDB format into PostGIS

Alan ,

> My 3.25 cents (the $AUS is rising)

That much?!!

> "Geometry is just another data type" - one of my favourite mantras for
> years.

Yep, mine too. But I've been a database guy since my Computer Science degree
days: now well over 25 years ago!

> I'm not so sure that ESRI "forces" the divide between spatial and
attribute
> but they are driving the spatial - it's the "be-all-and-end-all" of what
> they do according to Jack.

What I meant to say is that it seems to me that if you swallow the
Enterprise GeoDatabase thing and attempt to implement
it then you will come up against the IT world which doesn't think in such
terms or use such narrow, stovepipe, technology.
My last paid employer's developers used ERWin to develop around 5 database
applications (property rights, asset management,
conservation, inventory and supply chain) based on Oracle + spatial. The
datamodels had all sorts of modelling that ESRI simply
has never "allowed" over the years: multiple spatial columns per table;
multiple spatial objects per column; circles etc.

What ESRI is saying to IT Departments is that that approach is wrong: if the
datamodel contains spatial data it is a GeoDatabase and
must be modelled in Visio and deployed into their database, middle-tier and
client-tier technologies.

As if.

I have asked ESRIlites over the years how many GeoDatabases are fully
designed, specified and implemented? Shall we start a survey on what people
in this forum think? 

> It is sooo tempting to fiddle the data model to make the application
> modelling easier. ESRI caught this disease with INFO and, it seems, just
> can't quite shake it off. But, as you say Simon, why should they? It works
> for them.

Sure it works for them and the audience they have captured with their
education, sales and marketing efforts. But I can't help think that to the
bigger, broader, IT world they are a niche player. We worry about them
because we ply our trade in a market segment dominated by the ESRI
zeitgeist.

> The first GIS I worked with was GeoVision (a geometry engine sitting on
top
> of Oracle). When I first encountered ArcINFO I was horrified at the INFO
> "database" but bewitched and beguiled by the Arc geometry engine.

Mate, me too! I hated INFO (preferred Genamap's theoretical and flawed
approach) and have come over 25 years of use to respect the algorithmns in
the Arc fortran code.

> So, I'm interested Simon: "Do you think the ESRI Workspace XML has any
> future as a 'logical interface'?

I know nothing about the Workspace XML format. My experience has been more
with the XML Schema of the GeoDatabase. Now with this format the problem is
that everything inside the XML is couched in ESRI terminology and speak.
Also, I am not convinced that the format is as open as they seem to imply.
Try and get a hold of the XSD. There is a legal gray area around this format
that scares me. Also there is no evidence to say that ESRI will play fair:
the shapefile format is OPEN but they never published the spatial index file
formats (Gee, I wonder why?). Finally, very few ESRIlites know how to
generate such a document such that it contains:

1. Schematics;
2. Instance Data
3. Schematics and Instance data.

Frankly, ESRI never plays fair: why do we waste our time looking over our
shoulders?

S
-- 
SpatialDB Advice and Design, Solutions Architecture and Programming,
Oracle Database 10g Administrator Certified Associate; Oracle Database 10g
SQL Certified Professional
Oracle Spatial, SQL Server, PostGIS, MySQL, ArcSDE, Manifold GIS, FME,
Radius Topology and Studio Specialist.
39 Cliff View Drive, Allens Rivulet, 7150, Tasmania, Australia.
Website: www.spatialdbadvisor.com
  Email: simon at spatialdbadvisor.com
  Voice: +61 362 396397
Mobile: +61 418 396391
Skype: sggreener
Longitude: 147.20515 (147° 12' 18" E)
Latitude: -43.01530 (43° 00' 55" S)
NAC:W80CK 7SWP3
_______________________________________________
postgis-users mailing list
postgis-users at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users