logo separator

[mkgmap-dev] [Patch v1] improve LinePreparer to reduce img size

From Gerd Petermann GPetermann_muenchen at hotmail.com on Tue Jul 19 13:26:31 BST 2016

Hi Andrzej,


I did not find a simple way to predict the bit stream length

without repeating lots of code lines.

On the other hand, I also did not see an effect on the run time,

so the simple iteration looked good enough to me.

If you find different results, I'll try again.


In the meantime I was working on other improvements

in the existing filters, found some more special cases which

can be improved, but the effect will probably be smaller.


Gerd


________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Andrzej Popowski <popej at poczta.onet.pl>
Gesendet: Dienstag, 19. Juli 2016 13:09:15
An: mkgmap-dev at lists.mkgmap.org.uk
Betreff: Re: [mkgmap-dev] [Patch v1] improve LinePreparer to reduce img size

Hi Gerd,

this is great improvement, thanks! I have tested your patch on a road
map and get about 4% decrease of map size. I'm compiling topo map now, I
think it is over 10% size reduction.

I would like to suggest some optimizing of code. I think, you can get
optimum bit length directly in procedure calcDeltas from LinePreparer.
Just instead of only looking for min/max bit length, create full
profile: number of legs vs bit length, so you can estimate full bit size
like you have done it in your previous posts, without actually
processing a line. I hope this makes sens, I don't know your code so well.

--
Best regards,
Andrzej
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mkgmap.org.uk/pipermail/mkgmap-dev/attachments/20160719/cdf13b93/attachment.html>


More information about the mkgmap-dev mailing list