[Message Prev][Message
Next][Thread Prev][Thread Next][Message
Index][Thread Index]
Re: Multiple Endpoints
- Subject: Re: Multiple Endpoints
- From: Stuart Booth
- Date: Tue, 23 Sep 2003 23:26:00 +0000
On Tue, 23 Sep 2003 22:27:23 +0100, James <<a
href="/group/xAP_developer/post?postID=uLIiryTzUquBV-v7cYqutk2Ly0PaRjj9F7-15xYo41uiiZvx-vq_KrRUtWHs9gx-6hXc5NY">James@m...</a>>
wrote:
>Stuart,
>
>My vote would be to always go with the naming you use for services. As
>in:
>KCSoft.Slimp3.ServerName:Slimp3name
>In essence your plugin based app is the same as the windows service
>system.
I am very swiftly coming round to this way of thinking. If for some
reason I wanted to run multiple instances of some plugin on the same
machine their address characteristics can be completely customised via
the .config file.
Since I'm very pleased with the way Services work, and the way the
standalone plugin+runtime operates, that would indeed seem to be the
way to go.
Some more tweaking required then.
> The master app keeps an eye on things but is really only a
>device to start/stop and register devices. If you create it as a device
>with subs which then have subs life is just too confusing ;-)
In the case where there were multiple plugins loaded I made the master
application (the runtime as I've been calling it) pump out its own
heartbeats, as well as each of the plugins.
If the runtime doesn't actually load any plugins for some reason
(perhaps the load fails or there aren't any configured or found when a
search is performed), it should still pump out h/bs even though it's
effectively reduced to a passive listener - no different from my
Listener application really.
A runtime with just the 1 standalone plugin generates only the 1 h/b.
But a runtime with 2 plugins generates 3 h/bs, one for the main master
app, and one each for the plugins.
Maybe I've got the 1 standalone plugin case wrong here. I'll have to
dwell on that a little.
>In an upcoming program, to make things easy for the end user I am using
>a schema file that details devices and give frienldy names to elements
>of the schema. To decide what the device is and what posible
information
>could come from it I use the first two elements, and use the remaining
>elements for targeting. So far it is working very well.
I look forward to more details on this so I can understand it better.
Sounds quite flexible though if schema details can be linked together?
>With a spec 1.2 hat on:
>"If a device exists on the xAP network with a logical name a.b.c,
a
>second device may not use the logical name a.b.c.d."
Ahhh, good spot. That puts the nail in the coffin on that particular
technique. That's fine, you and Kevin have got me using subaddressing
for the first time ever.
Oh well, more work. These things never quite seem to end.
S
--
Stuart Booth
xAPFramework.net - a reusable xAP framework for .net
<a href="http://www.xapframework.net/">http://www.xapframework.net/</a>
<a
href="/group/xAP_developer/post?postID=3aGYBZigoHhkHZjYrUEbfDOZggFY8sXc8TJH9QL-GNnvt548lYIAxUL4O4BxS50EANfVZMBcx3q2DFUT5lIm">stuart@x...</a>
xAP_Development Main Index |
xAP_Development Thread Index |
xAP_Development Home |
Archives Home
|