summaryrefslogtreecommitdiff
path: root/starmath/Module_starmath.mk
diff options
context:
space:
mode:
authorStephan Bergmann <sbergman@redhat.com>2014-02-11 14:10:43 +0100
committerStephan Bergmann <sbergman@redhat.com>2014-02-11 14:21:07 +0100
commit6c697a99c90a40492eb8c16466a2082cc2be927e (patch)
treec3f8e543ec1eaf0ee586900e5b378c543bccce02 /starmath/Module_starmath.mk
parentb1add8f4acc231875d08e3edcf318865f2630e78 (diff)
Related fdo#74597: Java on Mac doesn't use javaldx/library path fiddling
...so no need to return non-empty SunInfo::getLibraryPaths (for the Oracle JRE, cf. how this had always returned an empty set for the Apple JRE in otherjre.cxx). This fixes the bug that selecting the Oracle JRE in the Advanced options pane didn't stick, because jfw_areEqualJavaInfo was called on a JavaInfo read back from XML where nRequirements had been cleared (in CNodeJavaInfo::loadFromNode) and a JavaInfo where non-empty SunInfo::getLibraryPaths -> VendorBase::needsRestart() == true -> nRequirements == JFW_REQUIRE_NEEDRESTART, so always returned false. (There's multiple errors here. It is questionable whether jfw_areEqualJavaInfo should take nRequirements into account, at least when used from SvxJavaOptionsPage. And VendorBase::getLibraryPaths appears to be only used to compute VendorBase::needsRestart, ignoring the actual path data. But leave that for future clean-up.) Change-Id: I5f004434b1cb7e909a153c43c474a2317f24b0c1
Diffstat (limited to 'starmath/Module_starmath.mk')
0 files changed, 0 insertions, 0 deletions