|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | RE: GISList: OGC and Standards, - a response |
| Date: |
01/07/2003 11:20:54 AM |
| From: |
sonny |
|
|
Dimitri,
Its quite obvious you have your opinions. A point by point rebuttal is not
productuve given the positions you have taken, so I'll exercise my option to
express my opinions and try to clarify some points.
1. I am actually quite familiar with your product and have purchased it and
subsequent upgrades since version 3.5 or so. I find it easy to use and
quite effective for the price. However, it would be easy to say that
Manifold's .map is reminiscent of MS Access .mdb and all the associated
problems with holding everything into single file. It would also be easy to
say the having to import a .map file to extract a component such as a script
to reuse it in another .map file is stupid. I can also say that Manifold's
documentation is quite lacking in examples when it comes to automating
processes. My point is that all software has flaws or annoyances. The
implementation of GML is not optimized for compact data storage, in fact
most vendors use it as a lingua franca between systems and not as native
objects. Because GML is XML, we can integrate easily into business systems,
EAI, and modern information buses such as Tibco. Having designed and
deployed many enterprise-wide systems, I can say that the proprietary data
formats employed by GIS systems are what drives the cost of integration.
GIS systems are that 20% of the system that ends up costing the most to
integrate.
2. I have read Manifold's position paper and it makes sense if the entire
geospatial market is just for GIS specialists. However that market is
saturated with lots of powerful desktop products, Manifold included.
Companies such as CDA (Manifold), Caliper, etc. combined don't even make a
1% dent in Daratech or Gartner market reports. There will always be a place
for desktop GIS, whether Manifold can beat the marketing machines of
established vendors remains to be seen.
The next market for GIS is what I euphemistically call "the stuff that just
works," although its more frequently called "commodity goods and services."
The future of GIS is not on the desktop or "low performance web interfaces,"
the future is at the backend. Business and consumers want to buy stuff that
just works, instead of learning yet another interface on top of a powerful
GIS. Most people, myself included, just don't care if you use a custom
forward a-star algorithm or simulated annealing to route them from point A
to point B. All that they care about is if they can follow the directions
and that they get there with a minimum of confusion. Because we build our
software around published specs, other software vendors (e.g. routing
software vendors) can integrate with our software, or we can integrate with
their software without resorting to low level API calls.
3. "No files" is not just marketing hype, it refers to the philosophy that
people want machines to provide them with what they want instead of turning
them into expert widget pushers. This is just more of the "stuff that just
works" philosophy. Of course all systems use files, whether they are data
sources, temporary work files, or segments in memory. However, why should
people have to deal with all the data when they only want part of the data?
4. Microsoft is an OGC principal member and has made Terraserver available
through a WMS interface. So there is no need to join MS, they already are
part of OGC.
5. OGC compliant software is not the Net PC, its software that complies to a
standard that promotes interoperability. The Net PC was a piece of hardware
in search of content. OGC compliance is just the opposite. From my client
software, I can access ESRI, Intergraph, Autocad, MapInfo, Cubewerx, Ionic
(here's the commercial plug), etc servers that expose their data through OGC
interfaces. I can request maps and feature data for inclusion into my
application without file transfers of data layers or reprojection. I can
also request portrayal rules and symbol libraries from other remote sources,
i.e. the same data can be portrayed as a USGS map or a MIL 2525b map by
pointing the software to the appropriate service URI. Other vendors can
also do the same depending upon the capabilities of their software.
6. I am really amazed at the claims of US government holding data hostage.
In my experience, the US is the only country that has consistently made
geospatial data available for the cost of downloading. This data is
produced to meet production requirements that may or may not take into
consideration commercial GIS systems. It is made available on an as-is
basis. I don't think anyone is playing gatekeeper here. Yes the formats are
obtuse, but the information is there to make use of the data. Everybody
feels this pain.
7. I would probably say that most if not all the active participants play by
the same "real-world" rules. I think that the most pleasant surprise when I
joined OG
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
