From mark.reinhold at oracle.com Fri Feb 7 03:39:11 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Fri, 07 Feb 2014 11:39:11 +0000 Subject: Proposal: Drop Stripped Implementations from Java SE 8 Message-ID: <20140207113911.538612@eggemoggin.niobe.net> In order to preserve compatibility and guard against fragmentation, the Stripped Implementations feature of Java SE 8 [1] requires some nontrivial changes to the TCK license. I've been working with Oracle's legal department on these revisions for some time now. We have an initial draft but at this point, unfortunately, I don't think there is sufficient time for members of this Expert Group, members of the JCP Executive Committee, and other interested parties to review and comment on these changes. I therefore propose to drop the Stripped Implementations feature from Java SE 8. This will only require changes to the Specification and to the TCK rules -- no change to the RI, or to actual TCK tests, is needed. I remain convinced that Stripped Implementations are important to the future of the Platform. Once Java SE 8 is out the door I'll look into how we might add this feature to the Platform in a release prior to Java SE 9. If you have any comments on this proposal, please let me know no later than 17:00 UTC next Tuesday, 11 February. Thanks, - Mark [1] http://cr.openjdk.java.net/~mr/se/8/java-se-8-fr-spec-01/#s9 From aph at redhat.com Sat Feb 8 02:54:52 2014 From: aph at redhat.com (Andrew Haley) Date: Sat, 08 Feb 2014 10:54:52 +0000 Subject: Proposal: Drop Stripped Implementations from Java SE 8 In-Reply-To: <20140207113911.538612@eggemoggin.niobe.net> References: <20140207113911.538612@eggemoggin.niobe.net> Message-ID: <52F60CFC.1050101@redhat.com> On 02/07/2014 11:39 AM, mark.reinhold at oracle.com wrote: > In order to preserve compatibility and guard against fragmentation, > the Stripped Implementations feature of Java SE 8 [1] requires some > nontrivial changes to the TCK license. > > I've been working with Oracle's legal department on these revisions > for some time now. We have an initial draft but at this point, > unfortunately, I don't think there is sufficient time for members of > this Expert Group, members of the JCP Executive Committee, and other > interested parties to review and comment on these changes. > > I therefore propose to drop the Stripped Implementations feature from > Java SE 8. This will only require changes to the Specification and > to the TCK rules -- no change to the RI, or to actual TCK tests, is > needed. > > I remain convinced that Stripped Implementations are important to the > future of the Platform. Once Java SE 8 is out the door I'll look into > how we might add this feature to the Platform in a release prior to > Java SE 9. > > If you have any comments on this proposal, please let me know no later > than 17:00 UTC next Tuesday, 11 February. Is it necessary to pull Stripped Implementations from the spec when only the TCK licence requires it? Or is the problem that, logically speaking, implementations can't be compliant with the SE spec without also being compliant with the TCK licence? In other words, couldn't we handle this without amending the SE spec? Andrew. From jochen.wiedmann at gmail.com Sat Feb 8 08:48:16 2014 From: jochen.wiedmann at gmail.com (Jochen Wiedmann) Date: Sat, 8 Feb 2014 17:48:16 +0100 Subject: Proposal: Drop Stripped Implementations from Java SE 8 In-Reply-To: <20140207113911.538612@eggemoggin.niobe.net> References: <20140207113911.538612@eggemoggin.niobe.net> Message-ID: Mark, I hsve just read for the first time about stripping. So, bear with me, if that comment comes comparatively (and probably: too) late. Nevertheless, I'd like to add the following notes (perhaps with Java 9 in mind): 1.) If you are going to permit stripped implementations, I see no sense in the following part A Stripped Implementation, once created, cannot be changed. No further elements may be added or removed. What's the point? The actual result is not different from performing both steps in the first place. 2.) If stripped implementations are permitted: What's the point in rejecting repackaging, like some Linux distributions (Debian, Ubuntu, or the JPackage project come to my mind) did in the past? I admit that this point is no longer as important as it was (with OpenJDK and IcedTea becoming very mature and floating all around). Nevertheless, from the view of practicality,Java is still lacking on certain platforms. To give but one example: As a professional I am bound to have Java 6 available (it has already been dropped from Fedora, and fixing that is quite hard, compared to just extracting and repckaging the official Oracle binary packages) and this tends to become harder and harder. At least my impression is: Providing a repackaged (but complete) implementation is the lesser evil, compared to a stripped implementation, of which we usually won't know precisely what's in it, and what's missing. Thanks for your attention, Jochen On Fri, Feb 7, 2014 at 12:39 PM, wrote: > In order to preserve compatibility and guard against fragmentation, > the Stripped Implementations feature of Java SE 8 [1] requires some > nontrivial changes to the TCK license. > > I've been working with Oracle's legal department on these revisions > for some time now. We have an initial draft but at this point, > unfortunately, I don't think there is sufficient time for members of > this Expert Group, members of the JCP Executive Committee, and other > interested parties to review and comment on these changes. > > I therefore propose to drop the Stripped Implementations feature from > Java SE 8. This will only require changes to the Specification and > to the TCK rules -- no change to the RI, or to actual TCK tests, is > needed. > > I remain convinced that Stripped Implementations are important to the > future of the Platform. Once Java SE 8 is out the door I'll look into > how we might add this feature to the Platform in a release prior to > Java SE 9. > > If you have any comments on this proposal, please let me know no later > than 17:00 UTC next Tuesday, 11 February. > > Thanks, > - Mark > > > [1] http://cr.openjdk.java.net/~mr/se/8/java-se-8-fr-spec-01/#s9 > -- "That's what prayers are ... it's frightened people trying to make friends with the bully!" Terry Pratchett. The Last Hero From mark.reinhold at oracle.com Mon Feb 10 09:21:52 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Mon, 10 Feb 2014 09:21:52 -0800 Subject: Proposal: Drop Stripped Implementations from Java SE 8 In-Reply-To: <52F60CFC.1050101@redhat.com> References: <20140207113911.538612@eggemoggin.niobe.net>, <52F60CFC.1050101@redhat.com> Message-ID: <20140210092152.62562@eggemoggin.niobe.net> 2014/2/7 18:54 -0800, aph at redhat.com: > Is it necessary to pull Stripped Implementations from the spec when > only the TCK licence requires it? Or is the problem that, logically > speaking, implementations can't be compliant with the SE spec without > also being compliant with the TCK licence? > > In other words, couldn't we handle this without amending the SE spec? If the Specification allows Stripped Implementations then the TCK license must also allow them. It might be possible to structure things so that only the TCK license needs to be modified, and we'll explore that option, but at this point we're simply out of time for SE 8. - Mark From mark.reinhold at oracle.com Mon Feb 10 09:44:26 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Mon, 10 Feb 2014 09:44:26 -0800 Subject: Java SE 8 (JSR 337) Final Release: Spec, RI, & TCK updates coming Message-ID: <20140210094426.188846@eggemoggin.niobe.net> An external reviewer spotted a rather nasty bug in the signature of a new method defined in the Lambda (JSR 335) Specification [1]. We decided to fix this in SE 8 because if we don't fix it now then we can never fix it, at least not without breaking compatibility. As a result, new builds of the Specification, RI, and TCK are in the works and will be made available to you shortly for review. Stay tuned ... - Mark [1] https://bugs.openjdk.java.net/browse/JDK-8033590 From mark.reinhold at oracle.com Tue Feb 11 20:54:46 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Tue, 11 Feb 2014 20:54:46 -0800 Subject: Java SE 8 (JSR 337) Final Release: Specification, RI, and TCK updated Message-ID: <20140211205446.566073@eggemoggin.niobe.net> The second draft of the Final Release Specification is available here: http://cr.openjdk.java.net/~mr/se/8/java-se-8-fr-spec-02/ Changes in this version: - Removed the Stripped Implementations feature, per my proposal last week. - Corrected various typographical errors in Annex 1. - Updated Annex 2, the annotated API specification, to be based upon JDK 8 build 129, which is the new candidate RI. - Updated Annex 3: Synced the Java Language Specification with the final version of the Lambda Expressions Specification (JSR 335) and corrected a minor issue concerning the integration of the grammar from the Type Annotations (JSR 308) Specification. Also corrected various typographical errors in both the JLS and JVMS documents. An updated draft Reference Implementation is available here: http://jdk8.java.net/java-se-8-ri/ This is based on JDK 8 build 129; the previous candidate RI was 126. This new RI build includes a fix for serious flaw in a new API [1] as well as various other fixes integrated since build 126 [2]. The draft TCK has also been updated, to JCK 8 build 42. You can access it via the instructions you previously received. At this point there are no known TCK failures, so there is no "known failures" list. Please let me know by 22:00 UTC this Thursday, 13 February whether you think these materials are ready for submission to the PMO, or if you need further time to evaluate them. Thanks, - Mark [1] https://bugs.openjdk.java.net/browse/JDK-8033590 [2] https://bugs.openjdk.java.net/issues/?filter=18854 From aph at redhat.com Thu Feb 13 01:31:27 2014 From: aph at redhat.com (Andrew Haley) Date: Thu, 13 Feb 2014 09:31:27 +0000 Subject: TCK8 looks good on RI Message-ID: <52FC90EF.2080802@redhat.com> Hi, with the newest TCK8, the results look good: compiler: passed: 31698 failed: 0 error: 0 devtools: passed: 19929 failed: 0 error: 0 runtime: passed: 68965 failed: 1 error: 0 Only one failure remains (the same as in previous runs): api/java_awt/Frame/indexTGF_GetSetExtendedState, and this is a known problem with one Linux release. Andrew. From mark.reinhold at oracle.com Thu Feb 13 07:50:56 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Thu, 13 Feb 2014 07:50:56 -0800 Subject: TCK8 looks good on RI In-Reply-To: <52FC90EF.2080802@redhat.com> References: <52FC90EF.2080802@redhat.com> Message-ID: <20140213075056.350086@eggemoggin.niobe.net> 2014/2/12 17:31 -0800, Andrew Haley : > with the newest TCK8, the results look good: > > ... Excellent! Thanks for confirming. - Mark From mark.reinhold at oracle.com Tue Feb 25 14:26:35 2014 From: mark.reinhold at oracle.com (mark.reinhold at oracle.com) Date: Tue, 25 Feb 2014 14:26:35 -0800 Subject: Java SE 8 (JSR 337) Final Release now in FAB Message-ID: <20140225142635.519157@eggemoggin.niobe.net> FYI, I submitted all the materials just in time for the Final Approval Ballot period to start last Tuesday (18 February). It runs through next Monday (3 March). - Mark