| Summary: | Broken 1.3M3 target platform (ClassFormatError) | ||
|---|---|---|---|
| Product: | [RT] RAP | Reporter: | Ivan Furnadjiev <ivan> |
| Component: | Releng | Assignee: | Project Inbox <rap-inbox> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | batxut, manuel.woelker |
| Version: | 1.3 | ||
| Target Milestone: | 1.3 M4 | ||
| Hardware: | All | ||
| OS: | All | ||
| Whiteboard: | |||
|
Description
Ivan Furnadjiev
Another variant of the above mentioned workaround is to download the target platform contained in the intro bundle of the tooling and extract the osgi bundle from there. http://download.eclipse.org/rt/rap/1.3/tooling/plugins/org.eclipse.rap.ui.intro_1.3.0.20091110-1744.jar The osgi bundle itself is in the file target/target.zip within the jar. The reason for the different md5 sums is that we used to push all bundles through the pack-200/signing process (so far, we only excluded icu and junit from this procedure). Obviously, the osgi bundle did not survive pack200 normalization this time and I only checked the runtime shipped with the tooling :( Since the non-rap bundles in our target platform are signed already, there is no use in doing that again. This practice even leads to different variants of the same bundle in the same version, which I think must be considered broken. I now excluded all non-rap bundles from pack200 and signing, so we ship the original bundles untouched. Uploaded a repaired zip file for M3, same version, same URL as above. I consider this bug as fixed (see comment #2). Please re-open if someone disagrees. |