|
|
|
|
|||||
|
||||||
Subject: Re: [Nml] Cap'n tips
From: Rick Emerson (rick@XXX.XXX)
Date: Mon Feb 15 1999 - 00:39:03 EST
Steven D. Tripp writes:
> > One quirk I did see repeatedly is a loss of centering when using
> > "intellegent zooming". This feature which is very useful keeps
> > selecting the most appropriate scale chart for the area being
> > displayed. If, for example, you're looking at a 1:40000 chart and
> > want to zoom down to get a close look at a feature and a 1:20000 chart
> > covers the area, this chart is brought up automatically. Conversely,
> > zooming out will invoke the next smaller scale chart (e.g., 1:80000).
> > When zooming in, however, occasionally the center of the new, larger
> > scale (smaller area) chart is no longer anywhere near the center used
> > with the previous chart. I was looking at the Portland, ME area and
> > zoomed in to look at Peaks I but found myself looking at a spot well
> > to the southwest of there. I'm not sure *why* this happens and will
> > accept it may well be an error on my part but it can be frustrating on
> > occasion.
>
> I'm not sure how the Cap'n works, but some of these nav programs do raster maps
> in this way: They break up maps into small rectangles for storage and load
> several (9?) pieces and join them seamlessly on screen. When you zoom in or
> out, it takes the clickpoint and tries to find a rectangle at the higher or
> lower scale and then loads it at the center and loads its surrounding pieces.
>
> If your click occurs right on an (invisible) boundary the computer won't know
> which rectangle to load to the center. On the other hand maybe the programmer
> made a mistake in labeling the rectangles.
The explanation is simpler than that. When The Cap'n shifts from a
1:80000 to a 1:40000 chart, it centers the image on the cursor whose
position can be off-center from simple program use. The result is the
offset, which is minor at 1:80000, is now much greater at 1:40000. I
don't know the internals of how the shift in chart images is done, but
I'd opt for passing the co-ords for the prervious image's center
instead of basing the new image on the cursor location which may have
shifted. In effect, there are two events happening here, a "magnify"
event and a "center image on cursor" event and the latter, IMHO,
should not happen with a magnify event. Nautical Technology sees it
otherwise.
The quilting you described above is used in Jeppeson's charting
package and the results, even on a PII 266, were unacceptably slow and
visually unattractive.
> Incidently, I am rewriting my Mac navigation program for Windows and UNIX and
> would be interested in suggestions for features.
A clean way to show running fixes on the chart would be nice.
Rick
S/V One With The Wind, Baba 35
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=
=-= TO UNSUBSCRIBE, send this message to majordomo@XXX.XXX: =-=
=-= unsubscribe navigation =-=
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--=-=
|