|Re: [p2-dev] Can I do the Automatic Discovery & Update with Equinox p2?|
I agree with everything Scott says about dynamic network discovery.
On the other hand, if you are talking about "discovering" updates from a known server, this is possible today with p2.
It involves including the p2 bundles in your RCP app, defining in advance what the update server is, and then setting an auto-updates preference so that the updates are presented to the user whenever they are found.
If you are talking about finding updates from a known server, and installing them without bothering the user at all...we don't do that now in the SDK since that almost always involves a restart and the user needs to decide when it happens. But it is possible today with the API, it would just take some custom code to find the updates and directly perform the install without going through the wizard.
So I guess the question...what is the specific scenario?
Scott Lewis <slewis@xxxxxxxxx>
To: P2 developer discussions <p2-dev@xxxxxxxxxxx>
cc: Markus Alexander Kuppe <markus.kuppe@xxxxxxxxxxx>
Subject: Re: [p2-dev] Can I do the Automatic Discovery & Update with Equinox p2?
There is an enhancement request to add ECF's discovery API (and some
discovery provider implementation) to network discover p2 repositories
in the Eclipse ui: https://bugs.eclipse.org/bugs/show_bug.cgi?id=218534
In recent interactions with Susan (leading the p2 UI effort), it seems
likely that this capability will be inserted in 3.5 M5 timeframe.
Here is ECF's discovery API documentation:
I (Scott) have also created a couple of example apps that use discovery
API to publish and then discover available services. I've also worked
on one for p2 discovery and then p2-based remote install scenarios for
So I guess the upshot is: p2 will soon do this (network-based discovery
at least)...but network discovery is not present as of 3.4.X stream.
Qixing Du wrote:
> Hey! After going through a lot of p2 marketing, I am still not clear
> if we can have the Automatic Discovery & Update of RCP application
> with p2.
> Could someone please show me some clues if we can do it? Any hint is
> highly appreciated!
> Thanks so much!
> p2-dev mailing list
p2-dev mailing list
p2-dev mailing list