[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [Newsgroup Home]
[news.eclipse.webtools.jsf] Re: [Announce] Texo project proposal: EMF technology for web application development environments

Hi Peter,
Thanks for your feedback, it surely is an interesting topic, see my comments below.

gr. Martin

Peter Friese wrote:
Hi Martin,

indeed avery interesting proposal. Here are some thoughts:

1) How about adding RCP support? If the backend architecture is
well-cut, it should be easy to attach an RCP frontend (e.g. using
Spring HTTP remoting).
MT>> Yes adding other target platforms (like rcp) is great. But (if I understand your remark correctly) the idea of the Texo project is not to define a common server side for different ui's. The idea is to generate apps which fit into the standard way of working as is common in the target environment. So a rcp app would/could probably have a different server side then a spring web app. A common server side for different ui's is a pretty large topic (I think e4 is also covering this)
Texo specifically talks about web apps to have a first focus. And to be honest, also because that is were my main expertise is. But I can see (as you) that there are many opportunities there for following a generative approach.


2) To me, the proposal sounds like a generator
"cartridge". I bet there are quite a number of people out there who
would come up with their own cartridges for other technology stacks.
So it might be worthwhile considering to create a higher-level project
for cartridges. Besides, it might be a good idea to mantion in the
proposal that Texo is a cartridge.
MT>> It all depends on the terminology ofcourse. What I understand from cartridges is that they are more like a set of common templates. The Texo proposal both discusses templates as well extending the model with ui-annotations and some tooling around it. The integration with Eclipse is an important topic. So I am not sure if the word cartridge would completely cover the Texo proposal.

Regarding your other remark about creating a higher-level project for cartridges. I can see that there is room for such a project which allows you to easily plugin your own cartridges/templates using for example eclipse extension points. Or is this maybe something which can be done as part of the ecoretools project?

3) Did you consider writing a
textual DSL for Texo? Using Xtext, it is quite easy to create usable
textual DSLs in short time.
MT>> Yes I thought about that. There are some separate ideas to start a project to define a ui-model (and dsl) for describing ui's. My idea for Texo is first to use annotations in the model and implement specific ui-model views in templates (a view is for example a selection grid). When the ui-model initiatives is more clear and the related dsl language has been created then my current idea is to plug this dsl into texo to drive the generation. My expectation is that many templates can be re-used then. So the answer is yes I thought about that but I see this as a next step because others are thinking about a ui-model.

The goal of the Texo project (as I proposed it) is to make a generative and template-oriented development approach available to programmers who are not familiar with model-driven development. Afaics and my experience is that there are still many and many developers out there who have not had any experience with model-driven development (who have never heard of EMF for example). Imho these developers can benefit from using a set of simple-to-use (but function-rich) tools and templates.


Peter



--

With Regards, Martin Taal

Springsite/Elver.org
Office: Hardwareweg 4, 3821 BV Amersfoort
Postal: Nassaulaan 7, 3941 EC Doorn
The Netherlands
Tel: +31 (0)84 420 2397
Fax: +31 (0)84 225 9307
Mail: mtaal@xxxxxxxxxxxxxx - mtaal@xxxxxxxxx
Web: www.springsite.com - www.elver.org