At 03:20  30/3/01 +0200, Stephen McConnell wrote:
>My view (and this is not a developer view) is that I would like to see a
>similar pattern to the Loggable, Configurable, Contextualizable, Startable
>... patterns.  What I imagine is a service (not necessary a component) that
>supports a lifecycle interfaces like "Installable", "Upgradable",
>"Degradable", and Removable.  How these interfaces are implemented is anther
>question and that where we get into the question of how assembly
>configurations are built and where the default configuration come from.

What I was thinking was modeling it more like Deployer class. So you would
have a

interface Installer
{
  void install( String name, URL location )
    throws InstallationException;
  void uninstall( String name )
    throws InstallationException;
}

This would install components into a container. Installation would involve
essentiall doing most of the work Deployer does now except-
* It does not mark the component as "ready to be run"
* It does not interpret any config files

>For reference, what are the issues you referring too ?

A few implementation and a few design;

implementation
* I didn't really understand JNDI till a bit back
* too lazy to implement DirectoryContext
* Can not hijack global NamingManager as hosted apps may need it

design
* not sure how far to go. ie should we continue to have BlockEntrys or
should everything be put in directory - or should we just mount the
BlockEntry in directory
* not sure on how to design namespace. ie Functional vs structural or both
ie
Functional:
app1/blocks/block1/Configuration
app1/blocks/block1/ComponentMapping
app1/blocks/block1/...

Structural:
app1/configuration/blocks/block1
app1/componentmapping/blocks/block1
app1/logs/block1/category
app1/policy/block1/category

or a combined approach where underlying Structural but have a Functional
line-up that referrs to the correct items.

Cheers,

Pete

*-----------------------------------------------------*
| "Faced with the choice between changing one's mind, |
| and proving that there is no need to do so - almost |
| everyone gets busy on the proof."                   |
|              - John Kenneth Galbraith               |
*-----------------------------------------------------*


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to