Hi all,

Attached is 2nd version of the patch.

The improvements are a bit better, although I've removed one

change in PolygonSplitterBase which reduces the number of splits

in rare cases (e.g. when the map contains large sea areas).

I think this change is okay, but it is not related to the LinePreparer

and thus should be handled separately.

A new binary containing the attached patch(es) is here:

http://files.mkgmap.org.uk/download/302/mkgmap.jar

If I here no complains I'll commit both patch next wednesday.

Gerd

Hi all,

during the last days I tried to find more about the way how mkgmap stores

polygons, esp. why we have some limits. While doing that I stumbled over an

old comment in LinePreparer.java:

        // I don't care about getting the smallest possible file size so
        // err on the side of caution.

which made me curious because I do care ;-)

I found out that Garmin offers a "trick" in the delta encoding which allows

to reduce the needed bytes in some cases, the imgformat-1.0.pdf

explains this in detail, search for " only the sign bit is set ".

This "trick" helps esp. well when coastline polygons or rivers are stored.

I can explain this more detailed if somebody likes to know more,

for now I'd like to hear if the trick causes any problems, I did not find any

so far.

A binary with this and a slightly improved overview_levels_v2.patch

can be found here:

http://files.mkgmap.org.uk/download/301/mkgmap.jar

Depending on the data a single tile with size ~5MB may be reduced by 50 - 500+kB.

Gerd

<reduceImgSize_v1.patch>

_______________________________________________
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev