Hi all (and especially the Caliper software architects):

I put in the following comments in a GIS blog (website:
http://www.spatiallyadjusted.com/2006/12/21/fighting-manifold-or-fighting-the-way-ive-learned-gis/).
Feel free to comment on it since I find some rather inaccurate
comments
(from Dimitri) about Maptitude....(you are welcome to respond)

Here's the partial text:



#

Rudy // Mar 31, 2007 at 4:26 pm

I've used all kinds of GIS since the DOS days (15 years and counting) and I
can assure you that by far the best GIS software is Maptitude from Caliper
Corporation. It is a vector GIS software that can also read and handle
raster imagery. It's native database management is vastly superior to any
other GIS (can handle all the TIGER line files of the US in one shot; it can
handle one billion!! records and one million fields — unlike MS access that
allows only 256 fields). It can read oracle and other DBMS without using
programming. An upgrade to TransCAD (another product from Caliper) allows
you to perform redistricting using integer programming and other graph
theoretic methods (invert matrices and store matrices of 25 K in native
format). I can go on and I am happy to debate on this issue. Oh BTY I have
an advanced ArcIMS certificate from ESRI for which my previous employer paid
1000s. I use C++ and R to do programming but was never required to program
for my GIS stuff. Now that's a great software. Isn't it?

Rudy B., PhD
Berkeley, CA
#

82 Dimitri // Apr 1, 2007 at 9:25 am

For all of its limitations I don't think anyone would describe Maptitude as
other than as a competent and pleasant GIS. It is well-matched to its target
market and has a good following.

Despite its many limitations (no IMS, small number of importable formats,
etc.), if you think it is the best GIS ever, well, you are entitled to your
opinion. What makes for "best" in someone's eyes is not necessarily the most
features or the most sophisticated capabilities but the right balance for
that someone's particular needs.

But this bit… "It's native database management is vastly superior to any
other GIS " …is utterly silly, as Maptitude (for all of its other benefits)
is rather well-known for having particularly weak DBMS capabilities.

I grant you that Maptitude may be improving its products and so is getting
better at DBMS, but if memory serves me right (as assisted by a quick review
of the Caliper website), it seems that the following gross DBMS limitations
to Maptitude still apply:

To take the most obvious, Maptitude connects to DBMS using ODBC, a terribly
obsolete way to go. GIS packages with better DBMS capability can connect
using more modern technologies such as OLE DB or ADO .NET. Connecting to SQL
Server using OLE DB is about 600 times faster, read/write, than using ODBC,
so this is a very big deal.

Although Maptitude is said to be able to read "oracle spatial tables" (a
particularly weird way of phrasing the matter, as if there is a host of
"gotcha's" waiting in the wings), there is no mention of any sophisticated
ability to do read/write/edits with many simultaneous users as is normally
desired with Oracle Spatial, nor of projection matching on the fly nor of
any support for GeoRasters, nor of storing formatting and other key drawing
characteristics. All those things are necessary if you are really working
with Oracle Spatial as a fully capable client and not just as some
"read-only" usage of Oracle as a one-way data source. I grant you it is cool
that Maptitude can read Oracle spatial data at all, but to do so in what is
apparently a highly limited way is not the way one wins standing as "vastly
superior."

For that matter, if you really are on top of your DBMS game you'll be able
to read/write/edit not just Oracle spatial but a host of different
geometry-in-DBMS data types and DBMS systems, including, for example OGC WKB
and WKT, ESRI-style geometry and so on in other DBMS packages, such as SQL
Server. Don't see any of that in Maptitude.

There is no trace of spatial SQL within Maptitude, something you'd expect
any GIS laying claim to serious chops in the DBMS world to offer. If you
can't do spatial SQL you're just not in the DBMS game in GIS, not even at
the beginner level let alone as the best.

I'll skip over the hundreds of small, but useful, capabilities that a truly
DBMS-capable GIS has and Maptitude does not (example: right click on a
column and choose "change type" to instantly change type… ) and conclude
with a very telling "big" thing: zero support for 64-bit code and multicore
processors. Modern DBMS is multi-threaded. If you can't run multiple threads
to your DBMS connection you're stuck in the dark ages.

To shift gears away from DBMS to programming, since your post indicated a
certain excitement at not having to do programming to do GIS stuff: not
since the dark ages has any modern, interactive GIS required you towrite
code to do GIS stuff, so don't be too excited that this is the case with
Maptitude.

It's a bit like a country cousin coming to the city and acting astonished
there is indoor plumbing. The indoor plumbing is indeed great, but one
should not betray too much astonishment at encountering such a convenience,
which is taken for granted in modern times. :-)

I don't mean any of the above as a slam at Maptitude because I happen to
like that software and admire it. I especially admire Maptitude because
together with packages like Manifold it helps set the precedent that one can
get truly useful and pleasant GIS for a fraction of the cost of legacy stuff
like ESRI. People in mainstream software markets understand that, but every
bit of re-education helps for those folks still stuck in legacy notions of
price/performance. So we should all praise Maptitude for helping move GIS
into modern notions of price/performance.

But on the way to praising Maptitude there is no need for inaccuracy, and
suggesting Maptitude is the best GIS there is at DBMS is very far from the
truth, hence this correction.

Reply via email to