[GRASS-dev] patch for v.voronoi/v.delaunay

[Benjamin Ducke]

So maybe the thing to do for now is to mention all these issues
in the HTML man page. The "BUGS" section is silent about them
right now.
Does the v.voronoi part suffer from the same problems?

Benjamin

Maris Nartiss wrote:
> Hi Benjamin,
> I'm sorry to disappoint You - Your patch does not fix bogous output
> from v.delaunay - there's still a garbage output from 46647 input
> points. Look at this old bug how to reproduce problem on 32bit system
> [1]. Your changes in calling myalloc have no effect, as previously
> allocated memory was increased by 4000*sizeof(*sites) chunks and now
> is assigned in one shot at beginning. It gives only slight memory saving.
> 
> I understand Your (and others) unwillingness to fix this code, still
> shipping broken code is also not a good thing. It MUST be
> fixed (patch/rewrite) or there must be some additional checks like
> if(npoints>MAX_INT) G_error("Blah! Decrease input vector size") to avoid producing bogous results without any warnings.
> 
> Maris.
> 
> 1. http://wald.intevation.org/tracker/index.php?func=detail&aid=335&group_id=21&atid=204
> 
> 2008/2/1, Benjamin Ducke <benjamin.ducke at ufg.uni-kiel.de>:
>> Hi all,
>>
>> As per request, I have patched v.delaunay (source in v.voronoi/)
>> very carefully.
>>
>> The memory handling now seems to work better. At least I was able
>> to triangulate a map with ca 30000 points. Anything more and my
>> system starts thrashing badly (1 GB, with a lot of other apps running).
>> Feel free to test with bigger inputs.
>>
>> The output vector map will now also preserve the 3D coordinates of
>> the original input points. However, the centroids for the triangles
>> are all wrong (placed at z=0.0). I don't see a quick way of fixing
>> this right now.
>>
>> In fact, I'd prefer not to touch the current code for v.delaunay
>> anymore. It is in urgent need of a complete overhaul as it is
>> inefficient, hard to read and still full of site lib artefacts.
>> Better to re-write from scratch with more time on my (or someone
>> else's) hands.
>>
>> Best,
>>
>> Benjamin
>>
>> --
>> Benjamin Ducke, M.A.
>> Archäoinformatik
>> (Archaeoinformation Science)
>> Institut für Ur- und Frühgeschichte
>> (Inst. of Prehistoric and Historic Archaeology)
>> Christian-Albrechts-Universität zu Kiel
>> Johanna-Mestorf-Straße 2-6
>> D 24098 Kiel
>> Germany
>>
>> Tel.: ++49 (0)431 880-3378 / -3379
>> Fax : ++49 (0)431 880-7300
>> www.uni-kiel.de/ufg
>>
>>
> 

-- 
Benjamin Ducke, M.A.
Archäoinformatik
(Archaeoinformation Science)
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
Johanna-Mestorf-Straße 2-6
D 24098 Kiel
Germany

Tel.: ++49 (0)431 880-3378 / -3379
Fax : ++49 (0)431 880-7300
www.uni-kiel.de/ufg