> ant jar
Then you could run the test class AddRemoveRelationPanelTest http://tinyurl.com/677s28 Which will give you something like the UI I showed recently here http://www.flickr.com/photos/bblfish/2712928069/ The N3 that drives some of the user interface is here http://tinyurl.com/6aowppThe <urn:java:net.java.sommer.addressbook.ui.editor.PersonInfoEditor> class described in that can be found
here: http://tinyurl.com/6fvdn2The btnk:RelationsEditor which is the factory class AddRemoveRelationPanelFactory - creating components for relations that have specific object editors - can be found here http://tinyurl.com/67kgzn .
The specific object editors are then built by that factory. One of these is the <urn:java:net.java.sommer.addressbook.ui.editor.LocationEditor >
visible here: http://tinyurl.com/6qbnj5The interesting discovery here is that I have found it best to work at a triple level. Perhaps not quite using Sesame Objects, but nearly. The reason is that this gives one a lot of flexibilty.
I suppose I should try to describe the architecture of this in more detail. Perhaps I will blog about it when I am a bit more confident.
Henry On 21 Jul 2008, at 18:22, Leo Sauermann wrote:
Hi guys, anything new? there are amendments to Giovanni Tummarello's conjecture:"thou shalt look at XForms, the forgotten standard promising a better future"Besides that, I still stay to the idea of extending fresnel, the Bizer/Pietriega church has a good evangelion there.... Tummerello's xml frameworks are from the past times of fixed vocabularies... but only god knows what is right.best LeoIt was Story Henry who said at the right time 23.06.2008 21:23 the following words:[note: I cced Jonathan Locke who created Wicket, as he was starting to get interested in semantic web technologies, and as I am mentioning his framework here]As there are many responses, I will group all the responses with the topic of different frameworks here.As I am writing a thick client and if I were going to try to create a easy binding layer like adobe AIR, I would probably try to work on improving JavaFX. Sadly there is no way currently to add @rdf annotations on fields and classes there, or of having fields or classes be URIs as that would make binding to the semantic web incredibly easy, but one could probably hack something together quite well.I am going to spend a few days trying to do something a lot less generic and see if that helps me get what I need. It will be inspired by fresnel. I will leave it to later to see if I can use the fresnel vocabulary. Clearly for editing there are some enhancements required to fresnel anyway. Changing a few URLs in my code should be quite easy once I have something working.I'll then report back with what I have come up with. On 23 Jun 2008, at 17:56, Leo Sauermann wrote:Henry Story wrote:I can't find "the way to tie relations to html fields" in wicket. I only see that their ComponentTags are independent widgets that are tied to values and to value-types, but not to relations - where is this?(( Just a thought:So I wonder if fresnel should not leave the box mode to the appropriate display format. For example, would it not be better to have a way to tie relations to html fields, the way say Apache Wicket does things http://wicket.apache.org/ .(ie somewhere around here? *http://tinyurl.com/6au3ut)*Yes, the clever thing about wicket is that the template is just pure html, with class="classid" and id="AnID" attribute value pairs. The xhtml can be viewed as is by a browser with no magic. The Wicket layer then binds the model to the the xhtml.The advantage is that html developers can produce html, and just need to name the bindings. Java then manipulates the DOM and fills in the fields by linking it to the java layer. Now the trick, would be for those who have some time, to replace the java model with an RDF model.But as I am not currently writing a web client, I can't spend time on this.)))[snip]but I would guess, even better inspiration would be Adobe Air or Microsoft's Silverlight,for example silverlights way of binding data: http://silverlight.net/learn/learnvideo.aspx?video=56926I couldn't find an equivalent for Air, but it should be somewhere here:http://www.adobe.com/devnet/air/flex/quickstart/ such as: MXML - binding controls to data http://livedocs.adobe.com/flex/3/html/help.html?content=mxml_2.htmlimpressive. Of course not useable by me :-) And I would be even more impressed if the whole thing had rdf data bindings...My point here is:if you go for TRUE dynamic user interfaces, the engine (gnogno, fresnel, ) should create a meta-model of the user interface first in a proper GUI representation language, such as MXML or MicrosoftsSiverlight, with data bindings and action bindings, and then let this be rendered..But this is 1000man-year development effort,yes, that's too much work for me :-)so at the end you and I are much quicker by hacking the whole thing with a little fresnel and (if | then | else | switch) hacks at teh right moment, or?any better ideas?Well I am going try doing something very close to this. To start with I will just create relations that will describe in RDF how the pieces fit together. So this will be close to fresnel, but I won't try to be as general as that to start off with. For example I will refer to java classes as urn:java:net.java.dev... URIs.I am thinking of it as properties files on steroids. This will be used by factories to create java objects, which will thenget instance info from the RDF database.I think that should be easy to build. I'll report back on that. Perhaps having done that people will find a way to generalize what I have done. And relate the resources I have created to Fresnel. I do find Fresnel inspiring, and it has helped me think about this. But I do have to get this thing working fast...On 23 June 2008, at 20:01 Axel Rauschmayer wrote:That same day Giovanni Tummarello wrote at 19:30:thou shall not visualize RDF nor EDIT it with an RDF configuration.Thou shall instead look into extending velocity or some other 100man years stable template engine with simple RDF querying and inheritancemechanisms.Velocity is one, wicket is another. Both though are web engines, and I am trying to build a thick client one. I could of course use wicket to create html to build Swing. And I have often wondered if that would not be a lot easier to work with. Perhaps what is needed is compiled xhtml to swing. :-)I agree to some point, but not entirely:- Using Fresnel as a template mechanism: I don't like this either. I'm using my own small templating language (this would be the "velocity" part of your conjecture which I thus agree with) and it is much more usable (for *templating*--see below for editing). - Understanding the Fresnel manual: Is hard. There are a lot of implicit things that make it tricky to figure out what is going on.On the other hand:- I've extended Fresnel to be used for form-based editing [1]. And I'm very happy with it. - Taking the manual as a foundation, discarding some of the more complicated things (additional content, anything but simple selectors, style), and adding what is necessary for editing, I think I now have something that is moderately easy to understand [1]. Well, at least most things are explicit. If you are still not happy with the result, let me know (also let me know why): My understanding of Fresnel (what makes sense, what doesn't) is *still* improving and in the not to distant future that will be reflected in a new version of this document.That is very interesting. There are a lot of pieces there. Mixing wiki syntax and Latex is real wicked though! :-) Adding Eclipse to the mix, and you get the Dr Evil prize, at least from Sun's perspective ;-)[1] "Lightweight data modeling in RDF", Axel Rauschmayer, Malte Kiesel. Download at http://hypergraphs.de/Greetings, Axel
smime.p7s
Description: S/MIME cryptographic signature
_______________________________________________ people mailing list [email protected] http://lists.semanticdesktop.org/mailman/listinfo/people
