commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Luc Maisonobe <Luc.Maison...@free.fr>
Subject Re: [math] [sandbox] merging Apache Commons BSP into Apache Commons Math
Date Sat, 14 May 2011 17:57:37 GMT
Le 14/05/2011 16:03, Phil Steitz a écrit :
> On 5/14/11 2:08 AM, Luc Maisonobe wrote:
>> Hello,
>>
>> Some weeks ago, I have imported in the sandbox a new component,
>> Apache Commons BSP which implements Binary Partitioning Trees (see
>> the thread about this creation here:
>> <http://markmail.org/thread/mcg23trgtl472ica>).
>>
>> Thinking further about it, I would like to directly merge it into
>> [math]. We are creating a new major release for [math], so it may
>> be a good time to do so. Also as I changed the code to put it in
>> sandbox, in fact I already did all the necessary work to have a
>> working implementation, with tests and clean reports from
>> checkstyle and findbugs. The implementation is complete for
>> dimensions 1, 2 and 3 in Cartesian space. I also need an
>> implementation on spherical geometry but I can do that regardless
>> of the component hosting this package. So the component could be
>> promoted and it really makes sense to have it inside [math].
>>
>> It could be put as a bsp package at top level (alongside with
>> geometry, analysis, ode ...).
>>
>> Does this suggestion makes sense ?
>
> +1 to integrate this code into [math].  I don't think we need
> anything beyond lazy consensus to do this, since we are not creating
> a Commons proper component here.  Lets just give others a little
> while to weigh in.
>
> I am on the fence re top level package vs subpackage of geometry,
> leaning toward the latter.  Why would you consider it as not a
> natural part of geometry?

We could put it there. In this case, we should probably move the current 
classes that are 3D specific in a sub-package too.

The current layout in [math] is one o.a.c.math.geometry package with 6 
3D specif classes. The current layout in [bsp] is one 
o.a.c.bsp.partitioning package that is dimension-independent, one 
o.a.c.bsp.utility package, three dimension-specific euclidean packages 
o.a.c.bsp.euclidean.oneD, o.a.c.bsp.euclidean.twoD, 
o.a.c.bsp.euclidean.threeD.

We could import bsp by changing o.a.c.bsp into o.a.c.math.geometry, 
putting the existing math 3D classes into the new 
o.a.c.math.euclidean.threeD package. This would open the opportunity to 
add 2D classes (some people have asked me if we would consider adding them).

The next step I mentioned in my original post would be to add later on 
o.a.c.math.sphere.oneD (for modeling planar angular sectors) and 
o.a.c.math.sphere.twoD (for modeling geographic data).

Luc

>
> Phil
>> Luc
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>> For additional commands, e-mail: dev-help@commons.apache.org
>>
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Mime
View raw message