Community
Participate
Working Groups
Build Identifier: Source for EMF Common, version 2.6.0.v20100914-1218 is missing from the Helios update site at http://download.eclipse.org/releases/helios. Although many other source features with that date exist, the only version available for this feature is 2.6.0.v20100614-1136 (provided you unchecked "Show only the latest version"). I couldn't find it at http://download.eclipse.org/modeling/emf/updates/releases/ either. The consequence for me is that I cannot step through the code while working against version 2.6.0.v20100914-1218. Reproducible: Always Steps to Reproduce: Browse update site at http://download.eclipse.org/releases/helios - the feature is not there.
Actually there's a workaround: I can import org.eclipse.emf.common bundle into workspace from the plug-ins view and then the source is displayed (I guess it is then taken from the running platform instead of the active target platform?). Still the issue remains valid. Oh yes, I forgot to mention that I use update site to create my target platform's contents.
Sorry, my mistake: the source is not displayed by default, but at least I can attach it manually now (it was not possible before I imported the bundle into workspace).
Kenn, Did something go a little off with the SR1 build?
(In reply to comment #3) > Kenn, > > Did something go a little off with the SR1 build? Not that I'm aware of. I suppose this may have to do with the problem we found with Buckminster's calculation of version qualifiers (Michal, can you please confirm/deny this?), but that would surprise me since this was a result of a change to p2 in Indigo (whereas the Helios build should be consuming p2 from Helios)...
> I suppose this may have to do with the problem we found > with Buckminster's calculation of version qualifiers > (Michal, can you please confirm/deny this?) Sorry, I have no idea.
(In reply to comment #5) > > I suppose this may have to do with the problem we found > > with Buckminster's calculation of version qualifiers > > (Michal, can you please confirm/deny this?) > > Sorry, I have no idea. No problem, the question was actually meant for Michal Ruzicka (also copied). ;)
(In reply to comment #4) > (In reply to comment #3) > > Kenn, > > > > Did something go a little off with the SR1 build? > > Not that I'm aware of. I suppose this may have to do with the problem we found > with Buckminster's calculation of version qualifiers (Michal, can you please > confirm/deny this?), but that would surprise me since this was a result of a > change to p2 in Indigo (whereas the Helios build should be consuming p2 from > Helios)... It turns the build is not quite correct. The likely cause is that it was created using Buckminster containing Bug #302804. Details & Workaround: The mentioned bug causes the source feature to have exactly the same presentation name (label) as the binary feature. Indeed if you browse the Helios release repository (http://download.eclipse.org/releases/helios) with the "Install New Software" wizard you'll notice there are actually two "EMF Common" features (with the exact same versions!): EMF Common 2.6.0.v20100914-1218 EMF Common 2.6.0.v20100914-1218 The workaround follows: Find out (by inspecting details of the features via the "More..." link in the "Details" section of the wizard's dialog window) which of the two is the source one and install it. Since a workaround is provided for the current build and any future EMF builds will use a fixed version of Buckminster (which should ensure the issue will not recur) resolving this issue as fixed.
Thanks, it works.
Closing all fixed releng bugs.
