Many dev-java packages don't have there files specified in SRC_URI, thus have empty digests. It makes impossible to detect download corruptions, and creates false positive for distfiles cleaning scripts. I understand it's related to there fetch restriction, but I don't think it is the right way to do it (correct me if i'm wrong). I think a better example is sun-jdk for instance, where file is in SRC_URI, and information about downloading it is in pkg_nofetch() function. Digests are okay in this case. This is the ebuilds list: ibm-jdk/ibm-jdk-1.4.1.ebuild ibm-jre/ibm-jre-1.3.0-r3.ebuild ibm-jre/ibm-jre-1.3.1-r2.ebuild ibm-jre/ibm-jre-1.4.1.ebuild infobus/infobus-1.2.ebuild jdbc2-oracle/jdbc2-oracle-8.1.7.ebuild jdbc2-oracle/jdbc2-oracle-9.0.1.4.ebuild jdbc2-oracle/jdbc2-oracle-9.0.1.ebuild jdbc2-oracle/jdbc2-oracle-9.2.0.1.ebuild jdbc2-oracle/jdbc2-oracle-9.2.0.3.ebuild jdbc3-oracle/jdbc3-oracle-9.2.0.1.ebuild jdbc3-oracle/jdbc3-oracle-9.2.0.3.ebuild jrockit/jrockit-8.1.ebuild sun-j2ee/sun-j2ee-1.3.1.ebuild sun-j2ee/sun-j2ee-1.3.1-r1.ebuild sun-j2sdk/sun-j2sdk-1.4.1.ebuild sun-j2sdk/sun-j2sdk-1.4.1-r1.ebuild sun-j2sdk/sun-j2sdk-1.4.2.ebuild I will attach a patch for the first one to make clear what kind of modification i think should be applied. (i may submit more patches later to help on this boring task if you agree on the principle.) Reproducible: Always Steps to Reproduce:
Created attachment 25288 [details, diff] ibm-jdk-1.4.1.patch A patch example.
im working a lot on ibm-jdk to have it well supported on PPC and x86 and have the digests of it fixed.
Note to self, took care of everything but jdbc and ibm, those are strider's :P (I fear his ebuildfoo in jdbc src_uri ;P)
ibm is cleared out now, only jdbc crap left
jdbc is not my package, ask absinthe, i think he put it in portage.
Fixed this for dev-java/jrockit. Also, renamed that package to dev-java/jrockit-jdk-bin, as it's full JDK.
i fixed the jdbc's ones :\ all gone now
