Bugzilla – Full Text Bug Listing
| Summary: | product bundle has 1.6 BREE | ||
|---|---|---|---|
| Product: | [Technology] EPP | Reporter: | David Williams <david_williams> |
| Component: | javascript-package | Assignee: | Project Inbox <epp.packager-inbox> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | blocker | ||
| Priority: | P3 | CC: | david_williams, mknauer |
| Version: | 1.3.0 | ||
| Target Milestone: | 1.3.0 | ||
| Hardware: | PC | ||
| OS: | Windows 7 | ||
| Whiteboard: | |||
|
Description
David Williams
I checked back to RC1, and apparently this problem has existed for a while, if not from the beginning. Lucky I was trying variations on my usual test pattern, I guess. I would like this fixed for the release ... even though I suspect 95% (or more?) of users do use java 1.6, it just seems wrong to constrain it when there is no reason to. I spot checked a few other packages (cpp, java, in addition to jee) and they either did not have any BREE (which is correct) or they had a lower leveled one (which isn't quite right, since this product bundles contain no java code) but, that is not as big a deal since we advertise that "at least 1.5 is required". Yepp, I agree, this is a blocker and I don't know how that value went into the bundle. I removed the value from the manifest and will start a rebuild of the JavaScript package soon. A new JavaScript package rebuild is under way: http://build.eclipse.org/technology/epp/epp_build/36/download/20100618-0543 But it would be interesting to use the old package with the error, add this p2 repository URL and start an update. http://download.eclipse.org/technology/epp/packages/helios/RC5.207 My expectation is that this fixes the error as well. Out of curiosity... and I haven't tried or even verified this: This is the product bundle, i.e. it doesn't contain Java code or any other binaries, just branding information (icons, plugin_customization.ini, etc.). If it was running in a 1.5 JVM with its old BREE >=1.6 settings, I would expect that this bundle ends in the state RESOLVED but not RUNNING. Isn't RESOLVED enough for such a product branding bundle? In my theory it should be working, but as I said: I never tried it out and I still think that it qualifies a rebuild of the JavsScript package. (In reply to comment #4) > Out of curiosity... It definitely did not work. That's how I found it. Just happened to try a 1.5 vm I had laying around for other reasons. The log was as below. The end result was that only the resource perspective displayed. I actually didn't look too much further ... I suspect most of the actual function would still work ... just not "look like" the JavaScript IDE you'd see with 1.6. !SESSION 2010-06-18 01:55:29.846 ----------------------------------------------- eclipse.buildId=I20100608-0911 java.version=1.5.0_22 java.vendor=Sun Microsystems Inc. BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=en_US Framework arguments: -product org.eclipse.epp.package.javascript.product -showlocation Command-line arguments: -os win32 -ws win32 -arch x86 -product org.eclipse.epp.package.javascript.product -showlocation -debug -consolelog -console -data ./workspace !ENTRY org.eclipse.equinox.app 0 0 2010-06-18 01:55:30.953 !MESSAGE Product org.eclipse.epp.package.javascript.product could not be found. (In reply to comment #3) > But it would be interesting to use the old package with the error, add this p2 > repository URL and start an update. > > http://download.eclipse.org/technology/epp/packages/helios/RC5.207 > > My expectation is that this fixes the error as well. Not sure, maybe? Or maybe I did it wrong? I selected one thing "JavaScript IDE Developers" (did not select the EPP feature line?) ... it said it would perform and update ... and it did, taking forever :) .... but still didn't work quite right and gave some different error messages in log. Also, minor note, when updating from the repo, something had an "inconsistent license" (or, SUA) so it asked be to read two of them to accept. The inconsistent one had caps, and June, 2010: ECLIPSE FOUNDATION SOFTWARE USER AGREEMENT\n\ June, 2010\n\ The log, after my update, was !SESSION 2010-06-18 02:19:55.790 ----------------------------------------------- eclipse.buildId=I20100608-0911 java.version=1.5.0_22 java.vendor=Sun Microsystems Inc. BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=en_US Framework arguments: -product org.eclipse.epp.package.javascript.product -showlocation Command-line arguments: -os win32 -ws win32 -arch x86 -product org.eclipse.epp.package.javascript.product -showlocation -debug -consolelog -console -data ./workspace !ENTRY org.eclipse.osgi 2 0 2010-06-18 02:19:58.364 !MESSAGE One or more bundles are not resolved because the following root constraints are not resolved: !SUBENTRY 1 org.eclipse.osgi 2 0 2010-06-18 02:19:58.364 !MESSAGE Bundle reference:file:plugins/javax.mail.glassfish_1.4.1.v201005082020.jar was not resolved. !SUBENTRY 2 javax.mail.glassfish 2 0 2010-06-18 02:19:58.379 !MESSAGE Missing imported package javax.activation_0.0.0. !ENTRY org.eclipse.osgi 2 0 2010-06-18 02:19:58.379 !MESSAGE The following is a complete list of bundles which are not resolved, see the prior log entry for the root cause if it exists: !SUBENTRY 1 org.eclipse.osgi 2 0 2010-06-18 02:19:58.379 !MESSAGE Bundle javax.mail.glassfish_1.4.1.v201005082020 [290] was not resolved. !SUBENTRY 2 javax.mail.glassfish 2 0 2010-06-18 02:19:58.395 !MESSAGE Missing imported package javax.activation_0.0.0. The new build of the JavaScript packages is available: http://build.eclipse.org/technology/epp/epp_build/36/download/20100618-0543 I will have a look into the other issues (update+license, see comment #6) later today (or this evening). The blocking part of this is fixed with the new build. I did see another bug, bug 317268, but I think that's minor. I've asked some others to look into it on Friday, to make sure its not a sign of something bigger that's wrong, but will give conditional approval ... conditional on none of these experts finding anything worse than a simple handler conflict. Much thanks for the quick turn around. |