Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[location-iwg] Architecture discussion from IRC

Managed to catch up with Andrew on IRC, and he asked me to move (a very interesting) discussion to the email list.


jgarnett: in any case I am keen to be involved in the architecture discussion for LocationTech
[12:31pm] aross42: Great! FYI, we'll have an architecture committee and as a committer you can be elected to it.
[12:32pm] jgarnett: there are kind of two main approaches ...
[12:32pm] jgarnett: a) common interfaces - basing those on spatial standards has worked pretty well. Especially for open source discussions, being able to lean on a definition saves tones of time on email lists.
[12:32pm] jgarnett: problem is it makes the result friendly to a GIS professional, and not especially a Java developer
[12:32pm] jgarnett: b) GeoScript is the other extreme
[12:32pm] jgarnett: where we tone down the GIS dial
[12:33pm] jgarnett: feature becomes record and so on
[12:33pm] jgarnett: this approach had the most opportunities for collaboration (JUMP and GeoTools have discussed this approach before)
[12:33pm] jgarnett: resulting in something friendly to developers, but a bit of a stretch for GIS professionals
[12:34pm] jgarnett: the nice part is that one can layer on top of the other.
[12:34pm] jgarnett: It would be very helpful to involve justin (i.e. jdeolive) the geoscript lead in any discussions
[12:34pm] jgarnett: as I am really keen to hear of his experience.
[12:35pm] aross42: That's a great summary of the two perspectives. I believe b is what some people are particularly excited about.
[12:35pm] jgarnett: and geoscript has a nice "simple" model
[12:35pm] aross42: Naturally a will always have it's place.
[12:35pm] jgarnett: that has been ported into several languages
[12:35pm] jgarnett: well it is pretty clear; the place for (a) is at the interoperability front
[12:36pm] jgarnett: (as an aside GML has the same perspective, it is the mad crazy interoperability solution; which projects are supposed to parse into their internal representation)
[12:36pm] aross42: This is a really interesting discussion. Want to start it on the list? I know others will be interested in weighing in.
[12:37pm] jgarnett: thinking
[12:37pm] jgarnett: I would like to listen a bit
[12:37pm] aross42: At the moment I don't have our bot set up just yet so IRC conversations get lost.
[12:37pm] jgarnett: and that is hard if I start the discussion with a vision such as the above
[12:38pm] jgarnett: If you don't mind I can cut and paste the above discussion to the email list ...
[12:38pm] aross42: I know in past meetings Chris has shared his views on architecture and his hopes of what might come out of collaboration at LocationTech. it's very much on topic.
[12:38pm] aross42: (Chris Holmes that is)
[12:39pm] jgarnett: I understand

Back to the top