----------------------------Original message----------------------------
                       MEMORANDUM
                       Date:   10/11/1993
                       Re:  Terry's properties and other fun
 
>From out on the road, hi folks! I'm here in the beautiful (!!) midwest
for a couple days, after a weekend in Atlanta, and before heading to NY,
VT, and DC.
 
Terry's question is most intriguing. I've only been able to approximate
that hiccup when there is something wrong with either (a) the data set, or
(b) ArcView. However, Terry's description of remediation seems exactly
correct, and baffling. I would be *EXTREMELY* interested to know if anyone
has encountered similarly "reluctant" data sets at any time.
 
Let me handle the second suspicion right away. Occasionally, ArcView gets
a memory migraine and just throws a hissy fit. This is usually after about
two hours of intensive tinkering. The remedy is to quit ArcView, quit
Windows, and reboot. That just clears out left over garbage in memory. At
trade shows, I now reboot after every couple hours.
 
By and large, my first suspicion is always the data files themselves. First
thing is to check the source from which the data were copied; make sure you
have been working with data from the DOS-based CD rather than the UNIX-based
CD. (I've made that mistake a couple times.) Second, if something from a
recopied data set looks goofy, try it from the original, using it straight
off the CD itself. If that works, trash the exisiting data set on the hard
drive and recopy a fresh set from the CD. If it works from the CD and not
from hard drive, there may be a problem with the media where the data is
being stored; it need not be a big problem ... even a singly byte of error
can throw off a system. Norton Utilities can check the media.
 
Meanwhile, I am discovering a new delight and challenge in working with the
1:2M data from ArcUSA. As I commented previously, it can be of extreme help
to have just one state's data available. I encouraged you to work with some
of your local ARC/INFO users to secure that. However, I am also finding it
so handy that *I* am in the process of trying to create single state subsets
for every state. As I get these created, I'll let you know. The whole idea,
of course, is to subset the 100 mb of ArcUSA_1:2M data (or the 100 mb of
decimal degree data), down to a single state, meaning about 3 mb of data for
a typical state. With careful processing, this can be compressed onto a
single disk (New York fits on a single 720-k disk!). So far, I've played
with Indiana, Minnesota, Georgia, Michigan, and New York in this fashion.
I'm discovering the little pitfalls in my procedure and, as I get practiced
at it, I'm improving; I'll write it down as soon as I can for you to share
with local ARC/INFO users.
 
Let me know if there are other special problems, opportunities, or requests!
 
Charlie on the road
----------
Charlie Fitzpatrick
ESRI K-12 Education & Libraries / ArcView
380 New York Street
Redlands, CA  92373-8100
909/793-2853 x.1887
909/793-5953 (fax)
[log in to unmask]