[postgis-users] Enormous file geodatabase feature class > PostgreSQL/PostGIS? + attribute column standardization app or process?

dnrg dananrg at yahoo.com
Wed Feb 20 11:38:43 PST 2008

Frank, thanks for the info. 

> You may find it helpful to use ogrinfo on your .mdb
> file to determine the name of layers as OGR sees
> them.

So you're saying OGR does not support the file
geodatabase (*.gdb) today.

The personal geodatabase (*.mdb) has a limitation,
last I checked, of 2G--that's all Access supports. My
data is at least 7G, so I may initially have to use
the file geodatabase format, if I have to use an ESRI
product at all.

But I'm sure I can work around the 2G personal
geodatabase limitation(*.mdb). I'm glad OGR can work
with geodatabase mdb files.

> Giving them the benefit of the doubt, I think the
> would like to be open, but it just isn't much
> of a priority.

Sounds good to me. Here's hoping.

Paul, you wrote:

> Tables, they're called tables. Why must
> everything be re-named in ESRI-land? Oh wait,
> I forgot, "GIS is different".

The Obfuscation Approach?

But Paul, "Feature Classes" are double-plus good. I
vid'ee (while I'm mixing dystopian references
here--"Orange" you glad...), charitably, they may have
chosen the term to favor the language of object
orientation than the language of newspeak. Or not.
Table is simpler, but not "table with a binary

> People have loaded the 500M features of UK Ordnance
> Survey into PostGIS. Refractions has 20M feature 
> hydrology databases.  TIGER has been loaded 
> many times (50M features).

Thanks for that. Glad to hear PostGIS and PostgreSQL
are ready for prime time / industrial use.

> Then it won't be particularly concurrent. It's write
> contention that causes concurrency issues, not read
> contention.

Depends on how you define concurrency. Not talking
about contention. It will be concurrent in the sense
of many simultaneous reads (if people like it). But
no, I don't forsee write issues like deadlocks, since
the data will be loaded, in batches, only
occassionally. It's not a transactional system. I
misspoke when I said OLAP--won't have a star or
snowflake schema.

Regarding the geodatabase as an open format / model, 
you wrote:

> That is an issue worth exploring, since the official
> pronouncements have said it would be open, but I
> never really thought it was worth agitating over
> since the format hadn't "arrived". But it's part
> of a shipping product now. Hm. Where's our
> open access?

All your problems will be solved at release n+0.1?

One thing I dislike about ESRI as a software company,
despite liking their products, is how closed /
exclusivist they are with submitting bug fixes /
enhancement requests. Last I checked, the system was
that only one designated person per ESRI customer
number could submit such things.

Why that's the case, I don't know. Seems the opposite
of being customer-facing / customer-centric when they
face only one employee per Customer Number.

Anyone, anywhere using their products should be able
to submit bug reports or feature enhancements,
shouldn't they? I've never heard of another software
company (though they must exist) where only one
blessed employee per customer number could submit a
bug report or feature request. For small GIS shops,
it's a small inconvenience. For large organizations
with thousands of employees, and/or hundreds of actual
GIS users, it seems like they're missing something
with the one blessed person approach.


Looking for last minute shopping deals?  
Find them fast with Yahoo! Search.  http://tools.search.yahoo.com/newsearch/category.php?category=shopping

More information about the postgis-users mailing list