Dear Sanjay, In this case, it is difficult to understand the object (GM_Object of Iso) use for the result of Boolean operations. In a Java hierarchical meaning, we can agree this use, but the two cases you have discussed are problematic.
In geometric modelling, we treat the second case by: the logic difference between A and B raises five components, 2 which came from A, 2 from B, and one component from A and B. We think that each component directly defined from A (resp . B) own the same characteristics than A (resp. B). We have one component with two antecedents (A - B). We would create a complex object to put it, but what about the 4 other components (perhaps have we computed boundaries for this operation?). We think that we just have to store the difference in a complex object with its boundary (which has been probably created). So we take A and B, we make the operation (difference) to create a complex object and we only keep the resulting component, as a complex because we already have its boundary. Perhaps Martin has already seen this problem (perhaps in 2) in forum or OGC conferences? We have seen your mails with Martin in the geotool developpement list. We would like to know when did you plan to commit your project? Indeed, we are waiting for this, with our student, to start the coding phase. It would be interesting to complete your work with methods or classes you dont have enough time to develop. Gilles and Romain Informatique- Synthèse d'images Gesquière Gilles Maître de Conférences LSIS- Université de Provence IUT Arles Rue R. Follereau 13200 Arles France <mailto:[EMAIL PROTECTED]> [EMAIL PROTECTED] <http://g.gesquiere.free.fr> g.gesquiere.free.fr tel: fax: <http://www.plaxo.com/click_to_call?src=jj_signature&To=%2B33+4+90+52+24+10&; [EMAIL PROTECTED]> +33 4 90 52 24 10 +33 4 90 52 24 15 <https://www.plaxo.com/add_me?u=60130157041&v0=1038122&k0=1760460404> Add me to your address book... <http://www.plaxo.com/signature> Want a signature like this? -----Message d'origine----- De : Sanjay Jena [mailto:[EMAIL PROTECTED] Envoyé : mercredi 8 novembre 2006 13:59 À : geotools-devel@lists.sourceforge.net; Prof. J. Roehrig; Romain Raffin; Gilles Gesquière; Martin Desruisseaux; Vincent Heurteaux Objet : Semantic questions Hi GeoTools community, as you know i am involved into an implementation of the Feature Geometry ISO19107. I would like to ask you for your opinion about the semantics of some geomtric objects and results of operations: As stated in the Abstract Specification, Primitives and Aggregates DO NOT contain their boundaries, and Complexes DO contain their boundaries. Set Operations like union, intersection and difference between those objects will usually result in a new object, which type (Primitive, Aggregate or Complex), unfortunately, sometimes is ambiguous. 1) A simple case: Surface A overlays with CompositeSurface B. Of which type will A.union(B) be? (In reality, a part of the object contains its boundary and another part does not). 2) More complicated case: Surface A overlays CompositeSurface B in a way that B.difference(A) splits B in two divided regions. As A doesn´t contain it´s boundary, the two new regions will contain its boundary. Hence, the resulting object should be a Complex. But Complexes (if i understood the definition correctly) can not represent divided regions. Only Aggregates (in that case a MultiSurface) can (but Aggregates do not contain their boundaries, since its contained Primitives don´t contain their boundaries). The type of unambigous operation results will be a design decision (like case 1). Some cases (like 2) bring problems, cause the results might not be representable correctly. I would really appreciate any opinion in relation to this topic, especially about divided regions which contain their boundaries. Thanks. Sanjay -- Sanjay Jena Brasil: Siqueira Campos 128, Ap 906, Copacabana, Rio de Janeiro Cel.: +55 21 81995254 Deutschland: Eitorferstr. 5, 50679 Köln Tel.: +49 221 7885099 Cel.: +49 163 1616196
image001.gif
Description: GIF image
------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________ Geotools-devel mailing list Geotools-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geotools-devel
