AW: AW: providin a dtd to hivemind xml files

On 2/27/06, Domsch, Christian <top-logic.com">Christian.Domschtop-logic.com> wrote:

Hi Hugo,

Your are right, that this component should be IDE aware in the sense,
that is has an api that future IDE implementations could use. I had this
in mind, but since my main knowlegde lies in swing implementation, this
was my first thought. I do have some knowlegde in eclipse plugins
building, and from this I gather that it shouldn't be too hard to write
the editor in that way, that the components would be reusable.

And I think that would be a great start for the community to skip in and
do the DIE stuff.

I will create a sourceforge project and keep this list informed.

To Andrea:

I have to disagree. Although normal xml editing shouldnt scare any
decent developer, the main disadvantage in the hivemodule writing lies
not in how to do xml tags and keep them correct, but in the knowlegde of
how the contributions to an excisting configuration point should look
like. In my last company I was the development lead of a very big swing
ui (1.5 million lines of code) and we had a really big hivemind backbone
(I still wonder, if our project is the biggest use of hivemind) with >
110 services, > 60 configuration points and > 50 schemata. And the
project was split into round about 30 libraries.
Now some developer had to use some functionality in our baseframework,
he had to know the exact schema of the configuration point, which meant
digging into the source code. (Ok, we could have used the hivedocs more
excessivly, but the generation of the hivedocs was not integrated in our
continous build environment). So this kind of editor, we are talking
about would have greatly improved their work, because the editor would
take over the part of knowing how those contributions shoudl look like.
This has nothing to do with having the knowledge of writing valid xml.

Greetings,

Christian.

---------------------------------------------------------------------
To unsubscribe, e-mail: jakarta.apache.org">hivemind-user-unsubscribejakarta.apache.org
For additional commands, e-mail: jakarta.apache.org"> hivemind-user-helpjakarta.apache.org

On 27/02/06, Martin Pike <yntegral.com"> mpyntegral.com> wrote:

I am looking to create a hybrid Java/Jython application. Java at the backend and Jython used for scripting components together and web application development. I have used HiveMind to configure Java services and would like to be able to do something similar with Jython ones also. I do not want to compile the Jython to Java using Jythonc,&nbsp; but would be interested in a strategy such as mapping Jython scripts using a generic factory for instantiating jython classes.

Does anybody have experience of using Jython with HiveMind or have any suggestions as to a direction I might take to accomplish this?

Regards,

Martin Pike

Andrea Chiumenti wrote:

mail.gmail.com" type="cite">Christian:
You convinced me, also because I stopped a minute and thought a bit about a previous employment for a customer (one of the big I can't make the name )
So, yes it should be extremely usefull!

On 2/27/06, Domsch, Christian <top-logic.com" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">Christian.Domschtop-logic.com > wrote:

Hi Hugo,

Your are right, that this component should be IDE aware in the sense,
that is has an api that future IDE implementations could use. I had this
in mind, but since my main knowlegde lies in swing implementation, this
was my first thought. I do have some knowlegde in eclipse plugins
building, and from this I gather that it shouldn't be too hard to write
the editor in that way, that the components would be reusable.

And I think that would be a great start for the community to skip in and
do the DIE stuff.

I will create a sourceforge project and keep this list informed.

To Andrea:

I have to disagree. Although normal xml editing shouldnt scare any
decent developer, the main disadvantage in the hivemodule writing lies
not in how to do xml tags and keep them correct, but in the knowlegde of
how the contributions to an excisting configuration point should look
like. In my last company I was the development lead of a very big swing
ui (1.5 million lines of code) and we had a really big hivemind backbone
(I still wonder, if our project is the biggest use of hivemind) with >
110 services, > 60 configuration points and > 50 schemata. And the
project was split into round about 30 libraries.
Now some developer had to use some functionality in our baseframework,
he had to know the exact schema of the configuration point, which meant
digging into the source code. (Ok, we could have used the hivedocs more
excessivly, but the generation of the hivedocs was not integrated in our
continous build environment). So this kind of editor, we are talking
about would have greatly improved their work, because the editor would
take over the part of knowing how those contributions shoudl look like.
This has nothing to do with having the knowledge of writing valid xml.

Greetings,

Christian.

---------------------------------------------------------------------
To unsubscribe, e-mail: jakarta.apache.org" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">hivemind-user-unsubscribejakarta.apache.org
For additional commands, e-mail: jakarta.apache.org" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)"> hivemind-user-helpjakarta.apache.org