Component Libraray For New Spring Desktop Project

[magpor]

Hi

I was wondering if any one has thought about starting a component project using the new Spring Desktop framework. Or I acctualy don't care if it is Spring RCP or Spring Desktop I guess a open source component library project aimed at implementing high quality Swing components for the most used components like a Microsoft Outlook look-a-like, PDF Component, Image Viewer etc. I know there are component slike these out there but they are normaly not that good and needs integration with Spring RCP/Desktop. I would want a component libarary based on Spring Desktop. It should offcause run standalone also but should be built around the Spring RCP/Desktop concepts.

I would love to get into such a project, maybe give the guys at Jide a bit of a competition, even though they are REALLY good I have to say


By the way does anyone know about a Microsoft look-a-like component built in Swing,w chih I can use in an appication I'm doing right now.

Cheers
Magnus

[LievenDoclo]

Hmm, I'd rather try and integrate all Jide components into RCP. I know they're not free, but I believe it'll be of great value and will help to spread Spring Desktop (or RCP for the moment) usage.
There is a project on Jide-RCP integration, but it seems to be a bit stagnant.

[rdawes]

Have you looked at swingx?

[LievenDoclo]

Yes I have. Before you get me wrong. I'm no evangelist for Jide. I do like the way they build up their component and the sheer size of the amount of components they have.
I've used JXTable, JXTreeTable, JXErrorDialog before. And they're great to work with. And we should integrated SwingX components into Spring RCP. I just don't think that'll be enough. SwingX is still quite limited.

[magpor]

Hi

I agree Jide is good, I use it alot but the point of open source is that it is open. It gives us way of learning from more experienced people. Personally I feel Im getting to the point where I can be considered experienced so I think it would be a learning experience and make Spring Desktop evnt more attractive.

My company (So4it) also build alot of its business around rich clients so we need to do it anyway since Jide is not the answer to everything we like to do it open source to give back something.


Cheers

Magnus

[LievenDoclo]

You can always try to persuade the Jide people to put more of their stuff in their open-sourced project...

[jwray]

Hi,

As the author of most of the Jide integration code I thought I'd add my comments to this.

The integration project hasn't had much activity of late mainly because it is basically working. There have been some small changes, mainly due to changes in the underlying APIs of either JIDE or Spring RCP, but the goal of the original project has been met, and the code base is being used successfully in multiple places.

As for integration of more components - what are you expecting? I use most of the JIDE components just fine in a number of rich client projects without any Spring specific integration, it is just not needed. An exception might be binding/binders for data entry components. The few that are in the project are the ones I have needed but I'd be happy to see more contributions for others.

Another potential integration point is with the action framework. I have some simple code along these lines that I've been waiting on committing since it depended on changes to the RCP codebase. These changes have been made so I'll add these soon.

If you have any other suggestions about useful ways to integrate more components I'd be glad to hear them.

Jonny

Hmm, I'd rather try and integrate all Jide components into RCP. I know they're not free, but I believe it'll be of great value and will help to spread Spring Desktop (or RCP for the moment) usage.
There is a project on Jide-RCP integration, but it seems to be a bit stagnant.

[magpor]

Hi

Once again I agree jide is great but its not free and even though my company can afford to use jide, and we will, I think the core of Open Source is to have it open and also to learn from each other. Im not suggesting we should start implementing Jide but at least give people an idea of how to write Swing component and maybe look into if there is some component integration that needs to be done with the new Spring Desktop.

Just a thought and since we will implement component like outlook etc I just thiught someone lese had thiught about it but it seems people are satisfied with wthat exists in the Open SOurce world already so be it.


Cheers
Magnus

[jwray]

Hi Magnus,

I actually agree with you - there is a dearth of good quality components in the open source world, that's why I ended up going the JIDE route. In addition, I personally have never implemented a custom swing component and it wouldn't be a trivial matter for me to learn how. The existence of a high quality library of open source examples would seriously reduce the learning curve there.

I was more replying to the suggestion that i) the jide integration project was dead and ii) that to use third party components some explicit integration with Spring RCP is needed, when a lot of the time it isn't.

Jonny

Hi

Once again I agree jide is great but its not free and even though my company can afford to use jide, and we will, I think the core of Open Source is to have it open and also to learn from each other. Im not suggesting we should start implementing Jide but at least give people an idea of how to write Swing component and maybe look into if there is some component integration that needs to be done with the new Spring Desktop.

Just a thought and since we will implement component like outlook etc I just thiught someone lese had thiught about it but it seems people are satisfied with wthat exists in the Open SOurce world already so be it.


Cheers
Magnus

[LievenDoclo]

jwray, I never suggested the JideRCP project was dead, I said stagnant (as in not much movement in CVS). I do understand you put stuff in on a need-to-have basis, and I'm really impressed on the work so far. So my apologies if you interpreted it that way.

As for the need to integrate, it helps a lot if people don't need to learn the inner workings of a component, when you're able to use an abstraction.
For example, it's a lot easier if you create a component which combines JidePivot with GlazedLists' EventTableModel so you can work with a list of items, nested properties and such, and then let users use that component instead of browsing through Jide's Javadocs (and without source support) and make the users use the 'raw' component.

As for open source, I'll be the first to use an open source component. However, I'm more persuaded to use Jide for commercial grade application, since they're just so high-quality. It's the same with DevExpress for .NET and Delphi. I do use JXTable, JXTreeTable, other SwingX components and other different Swing components, but in corporate world, it's often easier (and cheaper) to buy a well-documented, high quality component than to use and adapt an open-source component to fit your needs (perhaps even implement some features of the bought component that you really need).