Eclipse Community Forums: Oomph » Target Platform applied every startup

Help

Home

Home » Eclipse Projects » Oomph » Target Platform applied every startup(Target Platform applied every startup)

Show: Today's Messages :: Show Polls :: Message Navigator

Target Platform applied every startup [message #1774543]

Tue, 17 October 2017 02:21

Chris Lake

Messages: 10
Registered: November 2015

Junior Member

We use and apply a custom Target Platform using pde:TargetPlatformTask and it works well.

The only slight annoyance is that it applies itself at every startup. It's not something that changes often so it seems pointless (as in the devs have been whinging at me!).

I thought I could restrict the triggers, but I'm only allowed "MANUAL" which mean it won't get applied on initial setup.

I had a look through the code and I think I can see why it occurs:

  public boolean isNeeded(SetupTaskContext context) throws Exception {
  ...
    if (targetDefinition == null || !targetDefinition.isResolved() || !isActivate() || activeTargetDefinition == null
        || !targetDefinition.getHandle().equals(activeTargetDefinition.getHandle()))
  ...
  }

Watching the startup, the target always seems to initially be "unsresolved", but it does get resolved before the setup tasks run.
index.php/fa/31014/0/

So is that check worthwhile? It seems that any target platform will always be needed unless perhaps the check is performed later on.

Chris

Attachment: unresolved.png
(Size: 2.00KB, Downloaded 135 times)
Attachment: resolved.png
(Size: 22.79KB, Downloaded 128 times)

Report message to a moderator

Re: Target Platform applied every startup [message #1774550 is a reply to message #1774543]

Tue, 17 October 2017 06:53

Ed Merks

Messages: 29654
Registered: July 2009

Senior Member

I've been following this Bugzilla https://bugs.eclipse.org/bugs/show_bug.cgi?id=522662 which seems to suggest that for some reason the TP isn't considered resolved on startup even though it seems to have been resolved before shutdown. This appears to be purely a PDE issue, which of course would affect Oomph's task implementation because it checks there the TP is resolved. If it were resolve and the definition didn't change, the task shouldn't perform... So I suspect there is a PDE regression involved and am monitoring the Bugzilla to see how where that ends up...

Report message to a moderator