RE: MI Problem discovered w/Bounds Clause

1999-01-17 Thread Jacques Paris

I can only repeat what I have written here already but maybe too long ago.
The cosmetic layer is in the MapInfo default "projection" : lon/lat in
degrees not bounded. (see the discussion on internal precision way back) And
there is nothing to do about it.

Steve's advice not to use the cosmetic layer as a way to transfer/manipulate
graphic objects in situations other than unbounded lon-lat is perfectly
sound. But it may arrive too late (even when I found that 2/3 years back it
was too late) because some mbx utilities dealing with editing/ manipulating
graphic objects commercially or freely available have been built around the
use of the cosmetic layer to hold temporary objects.

I am not even sure that an object copied from a "projected" layer onto the
cosmetic layer and back to the original layer will be purely identical to
itself in all circumstances; one must consider the roundings that occur when
converting between coordinate systems.

To those using such utilities, I would advise them to have them checked at
least on the specific point of using the cosmetic layer if they work on maps
that require higher than default precision.

I take that opportunity to extend a warning not to place in the same mapper
layers with different bounding clauses (they may have the same "precision"
as expressed by the "widths" of the bounds but different bound values, or
origins if you prefer); if you do it, when working with graphic objects, you
may have lots of small problems, like slivers.., because of the lack of
coincidence of the internal coordinate grids.

Jacques Paris

e-mailalternate
 [EMAIL PROTECTED]   [EMAIL PROTECTED]

paris PC Consult (mainly MapInfo app.)
 www.total.net/~rparis/gisproducts.htm

-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED]]On Behalf Of Steve Wallace
Sent: November 18, 1999 7:46 PM
To: MapInfo-L
Subject: MI Problem discovered w/Bounds Clause


Something that might interest those of you who use the MapInfo "Bounds"
clause to increase the precision of your maps -- don't use the Cosmetic
Layer to transfer data around. I had been wondering why some of my
points/lines would move around from time to time. The distances were very
small, to the point where some might argue, "Why does it matter?" But the
problem is, the Object Processing notices these small differences, leaving
you will all kinds of unwanted spikes and slivers when you split or erase
objects.

The problem is that I, like many of you, use the Cosmetic Layer when
editing maps. In fact, when I first noticed the problem was going through
the prescribed method taught by MapInfo for fixing self-intersecting
polygon errors flagged by MapCheck. When I copied a string of nodes from my
bounded map to the cosmetic layer, then transformed to a region and went
back to process this against my original, the region in the cosmetic layer
was slightly offset from the original.

This is because the Cosmetic Layer is unbounded -- you lose any increased
precision from your bounded objects when you paste them info the cosmetic
layer.

WORK-AROUND: If you work with bounded data, create a blank layer with the
same bounds clause and use it like you would your cosmetic layer when you
need to put/process objects in a temporary layer.

This certainly explains why I have so many unwanted artifacts in my maps...



Steve Wallace
GIS  Market Information Manager
Florida Farm Bureau Insurance Companies

--
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]

--
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]



RE: MI Problem discovered w/Bounds Clause

1999-01-17 Thread Steve Wallace

At 04:37 PM 11/19/1999 +1000, Robert Crossley wrote:
Do you know if the object variables also unbounded as well? It could
explain some of the  problems with object processing that I have been
having with finding coincident points.   Could also be the cause of the
GPFs I've been getting with merging using projected data into a variable.


Good question. I'm not sure how MapInfo handles object variables in 
MapBasic. Since there is no map at the time, thus no projection clause or 
bounds, I can't see how they would be affected.



Steve Wallace
GIS  Market Information Manager
Florida Farm Bureau Insurance Companies

--
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]



RE: MI Problem discovered w/Bounds Clause

1999-01-16 Thread Robert Crossley

Do you know if the object variables also unbounded as well? It could 
explain some of the  problems with object processing that I have been 
having with finding coincident points.   Could also be the cause of the 
GPFs I've been getting with merging using projected data into a variable.

Rob.



-Original Message-
From:   Steve Wallace [SMTP:[EMAIL PROTECTED]]
Sent:   Friday, November 19, 1999 10:46 AM
To: MapInfo-L
Subject:MI Problem discovered w/Bounds Clause

Something that might interest those of you who use the MapInfo "Bounds"
clause to increase the precision of your maps -- don't use the Cosmetic
Layer to transfer data around. I had been wondering why some of my
points/lines would move around from time to time. The distances were very
small, to the point where some might argue, "Why does it matter?" But the
problem is, the Object Processing notices these small differences, leaving
you will all kinds of unwanted spikes and slivers when you split or erase
objects.

The problem is that I, like many of you, use the Cosmetic Layer when
editing maps. In fact, when I first noticed the problem was going through
the prescribed method taught by MapInfo for fixing self-intersecting
polygon errors flagged by MapCheck. When I copied a string of nodes from my 
bounded map to the cosmetic layer, then transformed to a region and went
back to process this against my original, the region in the cosmetic layer
was slightly offset from the original.

This is because the Cosmetic Layer is unbounded -- you lose any increased
precision from your bounded objects when you paste them info the cosmetic
layer.

WORK-AROUND: If you work with bounded data, create a blank layer with the
same bounds clause and use it like you would your cosmetic layer when you
need to put/process objects in a temporary layer.

This certainly explains why I have so many unwanted artifacts in my maps...



Steve Wallace
GIS  Market Information Manager
Florida Farm Bureau Insurance Companies

--
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]
--
To unsubscribe from this list, send e-mail to [EMAIL PROTECTED] and put
"unsubscribe MAPINFO-L" in the message body, or contact [EMAIL PROTECTED]