[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [udig-devel] uDig extension presentation on the 2nd Open Source GIS Meeting in Girona (Catalonia, Spain)
|
Javier Sánchez Ramírez wrote:
- One question I had was about the use of SOAP between the client and
the Oracle Web Services platform - is this simply for security?
We´ve been developing the Web Services SOAP Platform and the uDig extension at the same time. There were several reasons to face a SOAP platform implementation:
* By the time We have already chosen Oracle to be the Spatial Repository (mainly for its time-management capabilities), other standards such as WFS was not supported by Oracle yet.
* Every client/server communications had to be completely Web Oriented, including whatever alphanumeric request. This web oriented requirement for the client app. made us choose uDig.
* Interoperability reasons: uDig was not going to be the only client. In fact were also developing .Net and Open Layers clients.
* And not only security, but the fact that Ad-hoc logical requirements could be solved via PL-SQL programming.
That is all cool; I have used Spring to go from uDig to a web
application before. If you were starting from scratch there are several
good standards around these days beyond as web 2.0 stuff starts to come
online. Personally I find that I can break SOAP with sufficient data; it
will be really cool for me to watch your technology scale :-)
- You left the Vector image display blank; was that intentional?
* Not really, just lazy translating may be ;-). For the vector image we implement WKB format: feature(uDig) <-> WKB(web) <-> sdo_geometry(ora). As far as I remember GML, was not be parshed either way (GML<->sdo_geometry) when we started the project, and we implemented WKB. At the same time, every Data Base Version has it authorized geographical area assigned, this means, that a certain user will download vector images only within its assigned area, not the hole DB.
Okay that is fun; WKB is a cool choice to run on the wire.
- Can I confirm that the Bi-temporal management is the idea of allowing the map to display information at a specific point in time (in addition to the the usual point in space).
That is the point & something else! Bi-temporal management means valid-time + transactional-time management.
The user can change the display time, and new records can be inserted with specific valid-time, by means of valid-time management. Historical Data Base changes (even those considered as corrections) are stored in the Data Base by means of transactional-time management. Oracle versioning enablement capabilities already implements it.
Finally we are implementing a simplification of this with the following general idea: The user can set the working date (today by default), and this is the visualization display time and data query results time (past, present or future). At the same time, this user-defined date, becomes the initial date for every new Data Base record inserted (end time is set to null). Whenever you update a certain geometry, a new version of the record is created from the established date onwards, and the previous version sets its end-time to the established date.
At the same time, the user can manage the Oracle Workspace Manager API by means of the new Workspace View, in order to create new data base versions, set the workspace version to work with, invoke merge and refresh operations, etc.
Nifity. I jotted down some notes in the "HACK" space about one idea for
making "time" available in the uDig user interface. I would be
interested in your feedback. I will hunt down the link and send it to
you in a bit.
Thanks for your feed-back and added info, and sorry about my delay.
That is fine; development pace depends on the project - it matters that
we communicate at all (rather than how quickly we do it).
Jody