From: Barend Gehrels (barend_at_[hidden])
Date: 2008-01-31 11:54:59
Thanks for your reaction. I will consider the separation of core
geometry and more GIS-like geometries, and will study on your idea of
We had the feeling that a library user or implementer can always
speciliaze for other points than x-y, so it would work, but indeed if
you have to redo all algorithms for latlong it is quite a job.
Your idea is interesting and I will think about it. I cannot imagine it
all working like that, at this moment, but give me some time to look at it.
It is certainly food for thoughts!
Regards, Barend (it is "Barend").
Hervé Brönnimann wrote:
> On Jan 30, 2008, at 5:23 AM, Barend Gehrels wrote:
>> Dear Hervé,
>> Thanks for your extensive review on my preview...
>> Maybe I should have stated more clearly that I'm a geographer and
>> this library from a GIS perspective. We (at least in my company)
>> normally do not distinguish between polygon types, for example.
>> Therefore, as in my other mail, we might change the namespace to
>> "geospatial" or to "geospatial2d" to make this more clear.
> I agree that names of entities of your library (never heard of a
> linestring outside GIS, for instance) is slanted towards geospatial.
> However, many of the algorithms such as intersection, distance,
> convex hulls, etc. are much more general than that. I wonder if you
> could isolate the geometry part better without geospatial in mind,
> and perhaps have your own geospatial layer on top (in or out of
> boost). After all, it's it's truly generic, you shouldn't have
> problems / performance loss using it in your context. But I think
> it's not a good idea to craft a general-purpose geometry with names
> that belong to a particular domain.
> Hope those considerations give you food for thoughts,
> Hervé Brönnimann
> Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Barend Gehrels, Geodan Holding b.v., Amsterdam, The Netherlands, www.geodan.nl
Boost list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk