To track progress on removal due to security and EOL.
Status so far (searched by qgrep -H '=virtual/jdk-1.4*' so might missed something). Last rites of stuff that depends on 1.4 unconditionally (till 03 Nov): dev-java/axion dev-java/crimson -> dev-java/xmojo-bin (rdep of crimson) dev-java/tagunit (Rev)bumped to ~arch to remove java5 useflag (making its behavior unconditional) where !java5 means dep on 1.4 (can ask for stable on 03 Nov): dev-java/jdbc-jaybird-2.1.5-r1 dev-java/jdbc-postgresql-8.3_p605 dev-java/tomcat-servlet-api-5.5.27-r1 www-servers/tomcat-5.5.27-r4 Note: there's more stuff with java5 useflag but deps include "!java5? ( >=jdk-1.4 )" which means that 1.5 is always fine there and flag can be removed later. Stuff bumped by ali_bush, can ask stable on 18 Oct 2009: dev-db/octopus-3.0.1-r2 dev-java/openjms-0.7.6.1-r7 The rest of stuff depending on 1.4 unconditionally: dev-java/xindice (breaks on 1.5 xml api, might try bumping to 1.1 or masking) dev-java/jessie (breaks on 1.5 crypto api, dependency of gnu-classpath-inetlib, which is dependency of gnu-javamail) dev-util/eclipse-sdk:3.2 - the only stable version, nightmare www-server/axis (should look on the 1.4 bump bug/overlay)
To the 'revbumped to remove java5 flag' list: dev-java/backport-util-concurrent-3.0-r1
- bumped gentoo-classpath-inetlib to drop jessie dependency (any jdk 1.4+ provides this stuff already), so in 30 days it can go stable and jessie go away - last riting dev-db/xindice unless someone steps up and tries to properly bump it etc., nothing uses it - axis-1.4 seems to break mx4j, maybe 3.0.2 would be compatible but it fails to compile so far here
All last rites removed, all bumped stuff filled for stabling. Axis was bumped, which leaves eclipse-sdk the last thing to be dealt with :/
dev-java/jessie last rited
The 1.4 jre's are p.masked (why haven't we think of that earlier?) =app-emulation/emul-linux-x86-java-1.4* =dev-java/sun-jre-bin-1.4* =dev-java/ibm-jre-bin-1.4* dev-java/blackdown-jre
+# Vlastimil Babka <caster@gentoo.org> (17 Nov 2009) +# Removal for EOL and security, bug #287615 +=app-emulation/emul-linux-x86-java-1.4* +=dev-java/sun-jdk-1.4* +=dev-java/sun-jre-bin-1.4* +=dev-java/ibm-jre-bin-1.4* +dev-java/blackdown-jre +dev-java/blackdown-jdk
I have a problem with this removal. I maintain a java application, which needs sun jdk 1.4. I don't expect you will suspend removal from the main tree just for me :-), but how jdk-1.4 will be available for installation after that? Thanks. Oleg
You can still install it manually from Sun or another provider.
The plan is to keep it in java-overlay, and not stable. How's that?
(In reply to comment #10) > The plan is to keep it in java-overlay, and not stable. How's that? > Yes, this option is ok. Thanks. Oleg
Removed (java junkyard): dev-java/blackdown-jre =dev-java/sun-jre-bin-1.4* Moved to java-overlay, unstabled, marked as build-only: =app-emulation/emul-linux-x86-java-1.4* =dev-java/sun-jdk-1.4* #dev-java/blackdown-jdk Left in p.mask for now: =dev-java/ibm-jre-bin-1.4* IBM seems to security support 1.4 still, but maybe it's not worth the trouble. I will probably remove the JRE anyway and move JDK to java-overlay, but perhaps without marking as build-only (for the security bumped version which is not yet in tree).
Everything that was marked for treeclean was treecleaned.
+# Vlastimil Babka <caster@gentoo.org> (23 Jan 2010) +# Time to go due to EOL and security, bug #287615. +# Will be moved to java-overlay in 30 days. +=dev-java/ibm-jdk-bin-1.4* +=dev-java/jrockit-jdk-bin-1.4* +=virtual/jdk-1.4* +=virtual/jre-1.4*
What are you doing? I understand you are sunsetting java 1.4, but why are you masking these other IBM Java products with reference to their license? They are already fetch-restricted, and the process of manually downloading the binaries requires each user to register with IBM and explicitly agree to the license terms. So what is the purpose of this masking? Are these also being moved to Java overlay? The following is not self-explanatory. !!! All ebuilds that could satisfy "dev-java/ibm-jdk-bin" have been masked. !!! One of the following masked packages is required to complete your request: - dev-java/ibm-jdk-bin-1.6.0.7 (masked by: IBM-J1.6 license(s)) A copy of the 'IBM-J1.6' license is located at '/usr/portage/licenses/IBM-J1.6'. - dev-java/ibm-jdk-bin-1.6.0.6 (masked by: IBM-J1.6 license(s)) - dev-java/ibm-jdk-bin-1.5.0.11 (masked by: IBM-J1.5 license(s)) A copy of the 'IBM-J1.5' license is located at '/usr/portage/licenses/IBM-J1.5'. - dev-java/ibm-jdk-bin-1.5.0.10 (masked by: IBM-J1.5 license(s)) - dev-java/ibm-jdk-bin-1.4.2.13 (masked by: package.mask, IBM-J1.4 license(s)) /usr/portage/profiles/package.mask: # Vlastimil Babka <caster@gentoo.org> (23 Jan 2010) # Time to go due to EOL and security, bug #287615. # Will be moved to java-overlay in 30 days. A copy of the 'IBM-J1.4' license is located at '/usr/portage/licenses/IBM-J1.4'.
This may be an issue on architectures like ppc, for which other java builds are not available (including "icedtea"): # emerge -uavDN world These are the packages that would be merged, in order: Calculating dependencies... done! !!! All ebuilds that could satisfy "dev-java/icedtea6-bin" have been masked. !!! One of the following masked packages is required to complete your request: - dev-java/icedtea6-bin-1.6.2-r2 (masked by: missing keyword) - dev-java/icedtea6-bin-1.6.2-r1 (masked by: missing keyword) - dev-java/icedtea6-bin-1.6.2 (masked by: missing keyword) (dependency required by "virtual/jdk-1.6.0" [installed]) (dependency required by "net-dns/libidn-1.15" [installed]) (dependency required by "net-misc/whois-4.7.36" [installed]) (dependency required by "world" [argument])
(In reply to comment #15) > What are you doing? I understand you are sunsetting java 1.4, but why are you > masking these other IBM Java products with reference to their license? This has nothing to do with 1.4 removal. The IBM licenses were added to EULA group months ago. You can unmask them using ACCEPT_LICENSE. > They are already fetch-restricted, and the process of manually downloading the > binaries requires each user to register with IBM and explicitly agree to the > license terms. Sure, and because of this hassle with registering we don't want portage to select them by default. Sun's license is in EULA group, IBM has to be there too, otherwise it would be selected. Assuming it's easier for user to unmask sun's license than download manually from IBM. > So what is the purpose of this masking? Are these also being moved to Java > overlay? The following is not self-explanatory. 1.4 will be moved to java overlay, 1.5 and 1.6 will stay. (In reply to comment #16) > This may be an issue on architectures like ppc, for which other java builds are > not available (including "icedtea"): > > # emerge -uavDN world > > These are the packages that would be merged, in order: > > Calculating dependencies... done! > > !!! All ebuilds that could satisfy "dev-java/icedtea6-bin" have been masked. > !!! One of the following masked packages is required to complete your request: > - dev-java/icedtea6-bin-1.6.2-r2 (masked by: missing keyword) > - dev-java/icedtea6-bin-1.6.2-r1 (masked by: missing keyword) > - dev-java/icedtea6-bin-1.6.2 (masked by: missing keyword) > > (dependency required by "virtual/jdk-1.6.0" [installed]) > (dependency required by "net-dns/libidn-1.15" [installed]) > (dependency required by "net-misc/whois-4.7.36" [installed]) > (dependency required by "world" [argument]) > Hm this is unfortunate, that portage won't tell you about ibm-jdk-bin in this case. Guess the alternatives order in the virtual will have to be made arch-specific :(
(In reply to comment #17) Ah, I see. Sorry for the irrelevant posts here, then. I saw the reference to this bug and thought it was connected, but the more recent versions being masked for license is a separate issue (and I was ignorant of the "ACCEPT_LICENSE" mechanism). The virtual referring to icedtea (on a machine for which it is not keyworded) is a minor irritation. Thanks.
I've update the virtual/jdk ebuild so that ppc should no longer be affected by this. But it's a workaround for reopened bug 288083.
Thank you, Vlastimil. Also, in line with your suggestion, my problem was resolved by (revisiting the portage man page and then) creating /etc/portage/package.license file containing: dev-java/ibm-jdk-bin IBM-J1.6 Again, sorry for getting off on a sidetrack in this bug. Thanks for your efforts.
(In reply to comment #14) > +# Vlastimil Babka <caster@gentoo.org> (23 Jan 2010) > +# Time to go due to EOL and security, bug #287615. > +# Will be moved to java-overlay in 30 days. > +=dev-java/ibm-jdk-bin-1.4* > +=dev-java/jrockit-jdk-bin-1.4* > +=virtual/jdk-1.4* > +=virtual/jre-1.4* > Moved to java-overlay, jrockit deleted completely (upstream binaries not available). Also removed global java5 use flag, glazedlists has local flag until bug 304697 is done.
java5 flag gone, closing.