From Ray.Gans at Sun.COM Thu Jan 3 00:32:42 2008 From: Ray.Gans at Sun.COM (Ray Gans) Date: Thu, 03 Jan 2008 00:32:42 -0800 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants Message-ID: As you may have heard, Sun has announced the Community Innovation Awards Program (see http://www.sun.com/software/opensource/awards/) for several open source communities that we sponsor. The goal of this program is to foster and recognize innovation in these communities by offering grants/prizes to new efforts that will have an strong impact. The OpenJDK program will be called the OpenJDK Community Innovator's Challenge Grants (OCICG). We want to encourage developers to collaborate and creatively solve key problems facing the OpenJDK Community, initiate new projects that innovate on the OpenJDK code base, leverage the code for new uses, develop curricula and training, and port the code to new platforms, all to further the objectives of the OpenJDK Community in developing and disseminating compatible, free software implementations of the Java SE platform. We'd also like your help to make this program effective, valuable and fun for non- Sun participants. To implement this program, Sun will award several large grants to a few projects that can be completed by August 2008. Help us determine how to best select project proposals for consideration and allocate the money (we have approximately $175,000 to distribute). Here is how we're thinking it should work (though this may change based on your feedback): - On January 14, 2008 Sun will kick off the OCICG program and announce the criteria that will be used to select a set number of projects. OpenJDK participants will be encouraged to submit proposals for a project they want to work on. Proposals could be made by groups of individuals, existing F/OSS teams, companies/organizations, Java User Groups, etc. - Proposals will be accepted until March 3, 2008. At this time the proposals will be judged by a team of people (we're thinking 2 from Sun and 3 from outside Sun). We're also thinking of accepting only seven or less proposals. - Accepted proposals will be announced on March 17 ,2008 with all project work to be completed by August 4, 2008. - Awards will be delivered to completed projects on August 18, 2008 with cash amounts determined by the judges. We're thinking that the most valuable projects should be awarded a larger prize than others ? though all completed projects will be given a cash award. Note that no money will be available until August and all awards must be distributed at that time. Obviously, judges and Sun personnel will be ineligible for any cash awards. Scope and Constraints - The program will begin in January 2008 and end in August 2008. - Since this program is technically an international contest, there are strict rules by which it must be run. For example, participation will be restricted to countries that allow these kinds of contests. We would like to make the program open to as many countries as possible, however, since every country has different laws and requirements, we cannot accommodate everyone. We won't have the exact country list until mid-January. Other rules may also apply that limit what can and can't be done as part of this program. - Projects can only have limited dependence on Sun involvement/ participation. This is for fairness across all projects. Likewise, projects cannot require a commitment by Sun for significant time/ effort for success since we cannot guarantee adequate Sun resources will be available -- for example, a project to build a better bug database for OpenJDK, while very useful, would require heavy involvement by Sun personnel to integrate it with Sun's internal bug management systems. - All project code (if any) must be contributed to Sun under the Sun Contributor Agreement (see http://www.sun.com/software/opensource/ sca.pdf). We are interested in what the OpenJDK community thinks about the OCICG. You can help by providing input on any of the following questions (and whatever else you'd like to comment on). - What kind of projects do you think would be valuable to the OpenJDK community? - What selection criteria should be used to choose the best proposals? - How many proposals should be accepted? - Do you think the OpenJDK community at large should have any input into the proposal selection process? - Who you think would make good objective judges for the program and why? - What thoughts do you have about how the proposal selection process should be handled? - Do you think the OpenJDK community at large should have any input into selecting projects that really excel (and be awarded larger prizes)? - What criteria should be used to determine the payout for cash awards? - How should abandoned or non-completed projects be handled and what should constitute a "completed" project? - How should awards be handled for project team members who drop out or are added after a proposal is accepted? Please send your thoughts to discuss at openjdk.java.net. Thanks, The OpenJDK team at Sun -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/2ab7280a/attachment.html From gnu_andrew at member.fsf.org Thu Jan 3 01:23:35 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Thu, 3 Jan 2008 09:23:35 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: References: Message-ID: <17c6771e0801030123l1c6b9af9l735db751b7764743@mail.gmail.com> On 03/01/2008, Ray Gans wrote: > > > As you may have heard, Sun has announced the Community Innovation Awards > Program (see *http://www.sun.com/software/opensource/awards/*) > for several open source communities that we sponsor. The goal of this > program is to foster and recognize innovation in these communities by > offering grants/prizes to new efforts that will have an strong impact. > > The OpenJDK program will be called the OpenJDK Community Innovator's > Challenge Grants (OCICG). We want to encourage developers to collaborate and > creatively solve key problems facing the OpenJDK Community, initiate new > projects that innovate on the OpenJDK code base, leverage the code for new > uses, develop curricula and training, and port the code to new platforms, > all to further the objectives of the OpenJDK Community in developing and > disseminating compatible, free software implementations of the Java SE > platform. We'd also like your help to make this program effective, valuable > and fun for non-Sun participants. > > To implement this program, Sun will award several large grants to a few > projects that can be completed by August 2008. Help us determine how to best > select project proposals for consideration and allocate the money (we have > approximately $175,000 to distribute). > > Here is how we're thinking it should work (though this may change based on > your feedback): > > - On January 14, 2008 Sun will kick off the OCICG program and announce the > criteria that will be used to select a set number of projects. OpenJDK > participants will be encouraged to submit proposals for a project they want > to work on. Proposals could be made by groups of individuals, existing F/OSS > teams, companies/organizations, Java User Groups, etc. > - Proposals will be accepted until March 3, 2008. At this time the > proposals will be judged by a team of people (we're thinking 2 from Sun and > 3 from outside Sun). We're also thinking of accepting only seven or less > proposals. > - Accepted proposals will be announced on March 17 ,2008 with all project > work to be completed by August 4, 2008. > - Awards will be delivered to completed projects on August 18, 2008 with > cash amounts determined by the judges. We're thinking that the most valuable > projects should be awarded a larger prize than others ? though all completed > projects will be given a cash award. Note that no money will be available > until August and all awards must be distributed at that time. Obviously, > judges and Sun personnel will be ineligible for any cash awards. > > Scope and Constraints > - The program will begin in January 2008 and end in August 2008. > - Since this program is technically an international contest, there are > strict rules by which it must be run. For example, participation will be > restricted to countries that allow these kinds of contests. We would like to > make the program open to as many countries as possible, however, since every > country has different laws and requirements, we cannot accommodate everyone. > We won't have the exact country list until mid-January. Other rules may also > apply that limit what can and can't be done as part of this program. > - Projects can only have limited dependence on Sun > involvement/participation. This is for fairness across all projects. > Likewise, projects cannot require a commitment by Sun for significant > time/effort for success since we cannot guarantee adequate Sun resources > will be available -- for example, a project to build a better bug database > for OpenJDK, while very useful, would require heavy involvement by Sun > personnel to integrate it with Sun's internal bug management systems. > - All project code (if any) must be contributed to Sun under the Sun > Contributor Agreement (see *http://www.sun.com/software/opensource/sca.pdf > * ). > > We are interested in what the OpenJDK community thinks about the OCICG. > You can help by providing input on any of the following questions (and > whatever else you'd like to comment on). > - What kind of projects do you think would be valuable to the OpenJDK > community? > - What selection criteria should be used to choose the best proposals? > - How many proposals should be accepted? > - Do you think the OpenJDK community at large should have any input into > the proposal selection process? > - Who you think would make good objective judges for the program and why? > - What thoughts do you have about how the proposal selection process > should be handled? > - Do you think the OpenJDK community at large should have any input into > selecting projects that really excel (and be awarded larger prizes)? > - What criteria should be used to determine the payout for cash awards? > - How should abandoned or non-completed projects be handled and what > should constitute a "completed" project? > - How should awards be handled for project team members who drop out or > are added after a proposal is accepted? > > Please send your thoughts to *discuss at openjdk.java.net* > . > > Thanks, > > The OpenJDK team at Sun > > My initial thought is that OpenJDK will suffer from still being a fairly nascent project. Without being disparaging to all the great work done by Sun in releasing the OpenJDK source code (which must have been and still is a huge task), there isn't yet much of an OpenJDK community, other than that which it has adopted from existing projects such as GNU Classpath. Is there really the infrastructure in place to have a timed project take place? That said, I'd love to see it happen and think it would be a great way of fostering such a community if it could be made to happen. As regards projects, I think you have to just give fairly general criteria and see what gets proposed. Only then can you really decide on appropriateness and the exact criteria to use. Again, I think the problem will be that, at least my impression so far, is that there are no really active unfunded projects taking place which really interact with the core OpenJDK code base and bring Sun and external developers together. There is of course a porting effort and the framebuffer toolkit project which are both external, so there may be some interest there. The other projects listed on the OpenJDK site are very much internal Sun projects still; I'm on many of the mailing lists (too many really for most people) and I don't see a lot of traffic on most of them. Certainly nothing like patches, etc. but that's because the public repositories have only just become available. I don't want to sound negative, but when you compare OpenJDK with OpenSolaris, which is also taking part, there's a lot more activity already in existence which will make things a lot easier for them. One solution I think is to give each successful project it's own tree or forest (depending on the scale of the project), just like each team does now. Encourage activity there and interaction between the participants. This is certainly important as you seem to be proposing this to groups rather than individuals, which I think is a plus. Allow frequent commits there (with the appropriate SCA in place of course) and then handle the task of whether or not to feed the result back to the main project in August. When it comes to evaluating success, any successful project needs to have clear defined goals for judging this as part of the proposal. Having the code in a public repository means that both incomplete and complete projects will be accessible afterwards, even if the participants lose interest and disappear. The community can then decide how to handle the resulting work most effectively. As to payouts, it really depends on who you are trying to attract. With Google's Summer of Code, they aim this at individual students where the money is the primary incentive and so the reward is predefined and given in chunks, some of which is awarded before the project even begins (a project that fails midway, for example, still gets the student $2500 out of $4500). My perception of Sun's awards is more of one of bounties for groups of participants, where I think the award can be justifiably set on evaluation of the project and its merit, and then used as a bounty to be awarded only on its successful completion. The scheme would also be helped a lot if the OpenJDK community came up with some suggested projects with appropriate bounties to encourage contributions. It should still be made clear that new proposals are welcome, but this makes it easier for people to get involved who only have a vague idea about what the OpenJDK needs. I hope that's of some help, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/2ce0d147/attachment-0001.html From openjdk at jazillian.com Thu Jan 3 11:38:50 2008 From: openjdk at jazillian.com (Andy Tripp) Date: Thu, 03 Jan 2008 14:38:50 -0500 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: References: Message-ID: <477D39CA.6030003@jazillian.com> An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/8c5e7b42/attachment.html From Tom.Marble at Sun.COM Thu Jan 3 14:44:35 2008 From: Tom.Marble at Sun.COM (Tom Marble) Date: Thu, 03 Jan 2008 16:44:35 -0600 Subject: Source tarballs In-Reply-To: <475DD81E.3040806@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> Message-ID: <477D6553.5000509@sun.com> Andreas Sterbenz wrote: > Xiomara Jayasena wrote: >> >> We really didn't see the need, hence we decided to get rid of them. >> It seems anyone working in JDK 7 may need to become familiar with hg >> -- that said I appreciate your input. Here at Sun we will no longer >> be using the tarred source and expect engineers to do clones and that >> is the same expectation for developers outside of Sun. If many >> people think this is crucial then the decision can be re-evaluated. > > It is still possible to get source tarballs - by going to > http://hg.openjdk.java.net/ . Just click on the zip/gz/bz2 link next to > the desired repository to get the tip (e.g. > http://hg.openjdk.java.net/jdk7/jdk7/archive/tip.tar.bz2) or navigate to > the changeset or tag you like and follow the download link (e.g. > http://hg.openjdk.java.net/jdk7/jdk7/rev/0a5c5386a678). > > The only thing it doesn't do is understand forests, so you have to > download the source for each of the seven repositories in the forest > separately (/if/ you really want them all). I'd like to ask if we (Sun) can reconsider publishing one (1) source tarball for promoted builds. This question has come up in the past in the context of many Free Software distro build daemons which proscribe live Internet access during binary builds. Usually there is a "source package" which is uploaded to the build daemon which specifies any build time dependencies (e.g. specific compilers, header file packages) and includes the upstream tarball(s). This question came up again, today, on IRC in the context of building OpenJDK for stable distros. Developers stated that it is very desirable for non-root users to be able to build OpenJDK on stable distros by making only one simple request to their system administrator for a list of build dependency packages. While the user may have Internet access she cannot specify packages which were not part of the stable release. Without an OpenJDK upstream tarball this means that the typical way of getting the OpenJDK source is with "hg fclone" [0]. Which means that one would need Mercurial [1] with the Forest Extension [2] (and Mercurial needs python 2.4). This is complicated by the fact that the Forest Extension (currently, unversioned) is only supported on Mercurial 0.9.3, 0.9.4, and 0.9.5. If we take such a stable distribution, Debian etch [3] for example, then we find the following: python 2.4.4 http://packages.debian.org/etch/python/python 2.5 http://packages.debian.org/etch/python/python2.5 hg 0.9.1 http://packages.debian.org/etch/mercurial 0.9.5 http://packages.debian.org/etch-backports/mercurial BACKPORT forest (no version number) requires Mercurial 0.9.3, 0.9.4, or 0.9.5. This looks promising, except I learned that the "backports" repository is not considered part of the primary distribution and thus could not be used as a build dependency. The other problem is that getting forest.py configured for the system "hg" (i.e. not requiring every user to fix their ~/.hgrc) would require a new package (i.e. mercurial-forest) which is also not possible because such a package did not exist (and still does not) at the time the stable distribution was frozen. In distributions where hg 0.9.5 is available then there is solution (albeit less desirable) to have the user add forest.py to ~/bin and the path to ~/.hgrc (or equivalent in the build directory). Another possible solution, along the lines of what Andreas suggested, would be to grab the 7 tarballs from upstream and then combine them. If I understand this correctly, however, it would require scripting (or publishing) for each promoted build tag a set of changesets for each subordinate repository (e.g. b24 [4]). Compared to this complexity of these solutions if we simply published one tarball then supporting hg 0.9.5 and the forest extension would not be a build requirement. Please let me know what you think (esp. if I have characterized the solutions accurately). Regards, --Tom [0] http://weblogs.java.net/blog/kellyohair/archive/2007/12/openjdk_mercuri_7.html [1] http://www.selenic.com/mercurial/wiki/ [2] http://www.selenic.com/mercurial/wiki/index.cgi/ForestExtension [3] http://www.debian.org/releases/etch/ [4] http://hg.openjdk.java.net/jdk7/jdk7/log/cfeea66a3fa8 From neojia at gmail.com Thu Jan 3 14:53:21 2008 From: neojia at gmail.com (Neo Jia) Date: Thu, 3 Jan 2008 14:53:21 -0800 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <477D39CA.6030003@jazillian.com> References: <477D39CA.6030003@jazillian.com> Message-ID: <5d649bdb0801031453j4801ba55r7bfc844c7e698780@mail.gmail.com> On Jan 3, 2008 11:38 AM, Andy Tripp wrote: > > Hi Ray, > Here are my thoughts... > > Ray Gans wrote: > > > Proposals could be made by groups of individuals, existing F/OSS teams, > companies/organizations, Java User Groups, etc. > Presumably also individuals, right? > > > - Proposals will be accepted until March 3, 2008. At this time the proposals > will be judged by a team of people (we're thinking 2 from Sun and 3 from > outside Sun). Make this "group of 5" be an OpenJDK group, with all (or most) > discussions on an open mailing list. > > > > Note that no money will be available until August and all awards must be > distributed at that time. > It might be worthwhile to have at least one fairly formal milestone for > each project, so the project developers can be sure they're "on the right > track". > > > > - Projects can only have limited dependence on Sun > involvement/participation. This is for fairness across all projects. > Likewise, projects cannot require a commitment by Sun for significant > time/effort for success since we cannot guarantee adequate Sun resources > will be available -- for example, a project to build a better bug database > for OpenJDK, while very useful, would require heavy involvement by Sun > personnel to integrate it with Sun's internal bug management systems. > > Each proposal should be required to spell out exactly what it requires from > Sun. Using your bug database project example, it might require a snapshot of > the current Sun bug database as simple comma-separated values. One snapshot > at the start of the project and then another snapshot in August to cut over > to the new database. > > > > > > > > - What kind of projects do you think would be valuable to the OpenJDK I can provide something I have when I hack the jdk especially GC part, which might help somebody to start their jdk projects. Where should I put them? It is on my personal wiki now. Thanks, Neo > community? A "How to Hack the OpenJDK" book. > A better bug database (obviously, you've already thought of that :) ). > Massive improvements to the build system (Kelly O'Hair on steriods). > A remote build system that lets a developer make changes locally and submit > changes to a server and get back executables for his platform, without him > having to know anything about how the build works. > A version of Android based on OpenJDK. > > > - What selection criteria should be used to choose the best proposals? The > group should just have some general criteria statement like "Projects that > advance the goals of the OpenJDK project" and then reference a link that > spells out the goals of OpenJDK. Unfortunately, the only such like I can > find is this: http://www.sun.com/software/opensource/java/faq.jsp#e > So maybe spell out the goals yourselves: "increased adoption and increased > innovation" > > > - How many proposals should be accepted? Seven or less sounds good. > > > - Do you think the OpenJDK community at large should have any input into > the proposal selection process? Other than to allow anyone to participate in > a mailing list discussion, no. > > > - Who you think would make good objective judges for the program and why? > Inside Sun, someone like you and a Mark Reinhold/Danny Coward type of > person. Outside Sun, you just want to be sure to pick people who want to > advance OpenJDK itself, as opposed to pushing some social agenda or some > OpenJDK alternative. > > > - What thoughts do you have about how the proposal selection process should > be handled? After the proposal submission period ends, publish the > proposals, allow a couple weeks of discussion on a mailing list, > and then the "group of 5" votes. If greater than 7 submissions, each person > may only vote for 7 proposals and > the 7 proposals with the most votes win. If 7 or less, A yes/no vote on > each. To break down the money awards, > after discussion, any of the 5 can submit a proposal for how to split the > awards. > The first such proposal to get 3 yes votes wins. > > > - Do you think the OpenJDK community at large should have any input into > selecting projects that really excel (and be awarded larger prizes)? Again, > hold the discussions on open mailing lists and allow anyone to comment. > > > - What criteria should be used to determine the payout for cash awards? Use > some vague wording such as "Award amount will be based on potential value to > the OpenJDK community". > And I would keep it to being based on value, not difficulty. No need for a > "degree of difficulty" modifier. > > > - How should abandoned or non-completed projects be handled and what should > constitute a "completed" project? I presume accepted project participants > will have to sign something simple that says they will do their best to > complete > the project by a certain date. It should also mention that if they can't > complete the project, they must post a notice > on the mailing list. > > > - How should awards be handled for project team members who drop out or are > added after a proposal is accepted? Ugh. That's a tough one. I guess that > this thing that accepted projects would have to sign should designate a > single > entity (person or corp) that will receive the money, and that Sun will > award the money to that entity. Who gets > added or deleted or whatever from the project team is then none of Sun's > business. You sure don't want to > try to micromanage that. > > So, this "You've been accepted" document might contain: > * notice that you've been accepted > * expectation that you'll finish by $DATE, that it will be reasonable > quality, and it will be substantially what's in the proposal. > * promise that Sun will pay the amount of $AMOUNT on date $DATE to $ENTITY > upon successful completion > * description of any milestones > * mention that ultimately, "successful completion" will be determined by > this "group of 5". > * mention that Sun stays out of issues about who is on the project > * mention any help that Sun agrees to (e.g. one bug database dump at start, > another at end) > * any other required legal mumbo jumbo > > > > > Please send your thoughts to discuss at openjdk.java.net. > > > Thanks, > > > The OpenJDK team at Sun Good Luck with it! > Andy > > -- I would remember that if researchers were not ambitious probably today we haven't the technology we are using! From Xiomara.Jayasena at Sun.COM Thu Jan 3 15:49:29 2008 From: Xiomara.Jayasena at Sun.COM (Xiomara Jayasena) Date: Thu, 03 Jan 2008 15:49:29 -0800 Subject: Source tarballs In-Reply-To: <477D6553.5000509@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> Message-ID: <477D7489.9000205@sun.com> Hi Tom, Happy New Year! Tom Marble wrote: > > I'd like to ask if we (Sun) can reconsider publishing one (1) > source tarball for promoted builds. > Yes it is definitely being considered -- as it makes sense, especially if it is difficult for people to pull down the 7 tar balls that currently exist now. There are a couple of ways to go about this so that will need to be sorted out. Best regards, -Xiomara -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/197c2a88/attachment.html From Weijun.Wang at Sun.COM Thu Jan 3 15:53:55 2008 From: Weijun.Wang at Sun.COM (Max (Weijun) Wang) Date: Fri, 04 Jan 2008 07:53:55 +0800 Subject: Source tarballs In-Reply-To: <477D7489.9000205@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> <477D7489.9000205@sun.com> Message-ID: <04B63CDE-8044-42A5-9B32-0FD6B8CC391C@sun.com> Is it easy to hack the hg HTTP server to let it understand forest? Max On Jan 4, 2008, at 7:49 AM, Xiomara Jayasena wrote: > > Hi Tom, > > Happy New Year! > > Tom Marble wrote: >> I'd like to ask if we (Sun) can reconsider publishing one (1) >> source tarball for promoted builds. > > Yes it is definitely being considered -- as it makes sense, > especially if it is difficult for people to pull down the 7 tar > balls that currently exist now. There are a couple of ways to go > about this so that will need to be sorted out. > > Best regards, > -Xiomara > > From landonf at threerings.net Thu Jan 3 16:14:00 2008 From: landonf at threerings.net (Landon Fuller) Date: Thu, 3 Jan 2008 16:14:00 -0800 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: References: Message-ID: <55C4C4C6-6E88-482F-98D1-31A0D6399EA9@threerings.net> On Jan 3, 2008, at 00:32, Ray Gans wrote: > - What kind of projects do you think would be valuable to the > OpenJDK community? I'd love to see more community development on porting work -- including Mac OS X and BSD support (perhaps this is obvious, coming from me). Speaking just for Mac OS, it still needs work on multiple fronts (via the OpenJDK Porting Group and FreeBSD Java Project): - Merging into OpenJDK (part of the larger BSD porting work). - Support for hotspot x86_32 16-byte stack alignment (vs. using - fstack-realign work-around) - x86_64 16-byte stack alignment fixes - Integration of PowerPC support? (gbenson's?) - Native AWT implementation I'm not sure if this is something both Sun and the larger community are interested in =) -landonf -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/f03cbfc1/attachment.html -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://mail.openjdk.java.net/pipermail/discuss/attachments/20080103/f03cbfc1/attachment.bin From David.Herron at Sun.COM Thu Jan 3 19:18:39 2008 From: David.Herron at Sun.COM (David Herron) Date: Thu, 03 Jan 2008 19:18:39 -0800 Subject: Project Proposal: Haiku port of OpenJDK In-Reply-To: <477D29FF.2070605@gmail.com> References: <477D29FF.2070605@gmail.com> Message-ID: <477DA58F.3030405@sun.com> Yes! Yes! Yes! Yes! Yes! Yes! Yes! Yes! Yes! Yes! Dalibor Topic wrote: > I hereby propose, on behalf of the Haiku OpenJDK port developers, > the creation of a new project to create and maintain a port of OpenJDK > to the Haiku operating system. > > For existing discussion of the proposal, I'd like to refer you to the > thread > initiated by Bryan Varner's post proposing the project on the porters-dev > mailing list at [1]. > > cheers, > dalibor topic > > > [1] > http://mail.openjdk.java.net/pipermail/porters-dev/2007-December/000035.html > From Tom.Marble at Sun.COM Thu Jan 3 21:09:18 2008 From: Tom.Marble at Sun.COM (Tom Marble) Date: Thu, 03 Jan 2008 23:09:18 -0600 Subject: Project Proposal: Haiku port of OpenJDK In-Reply-To: <477D29FF.2070605@gmail.com> References: <477D29FF.2070605@gmail.com> Message-ID: <477DBF7E.5070301@sun.com> Dalibor Topic wrote: > I hereby propose, on behalf of the Haiku OpenJDK port developers, > the creation of a new project to create and maintain a port of OpenJDK > to the Haiku operating system. I approve. I believe that encouraging broad adoption of OpenJDK on as many Operating Systems and Architectures as possible is essential to furthering our goals of Java ubiquity. I regret, Bryan (et al), that it has taken so long to get to this point. It has been a long time since that meeting back in June at the IndyJUG where you first mentioned this to Ian Murdock. I'm glad to support this project. Thanks for your patience and for your contributions to OpenJDK. And thank you, Dalibor, for all of your help in getting Porters going! Regards, --Tom From gnu_andrew at member.fsf.org Fri Jan 4 01:25:05 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Fri, 4 Jan 2008 09:25:05 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <55C4C4C6-6E88-482F-98D1-31A0D6399EA9@threerings.net> References: <55C4C4C6-6E88-482F-98D1-31A0D6399EA9@threerings.net> Message-ID: <17c6771e0801040125o69da8abese201576a0cdb7ed8@mail.gmail.com> On 04/01/2008, Landon Fuller wrote: > > > On Jan 3, 2008, at 00:32, Ray Gans wrote: > > - What kind of projects do you think would be valuable to the OpenJDK > community? > > > I'd love to see more community development on porting work -- including > Mac OS X and BSD support (perhaps this is obvious, coming from me). > > Speaking just for Mac OS, it still needs work on multiple fronts (via the > OpenJDK Porting Group and FreeBSD Java Project): > - Merging into OpenJDK (part of the larger BSD porting work). > - Support for hotspot x86_32 16-byte stack alignment (vs. using > -fstack-realign work-around) > - x86_64 16-byte stack alignment fixes > - Integration of PowerPC support? (gbenson's?) > - Native AWT implementation > > I'm not sure if this is something both Sun and the larger community are > interested in =) > > -landonf > > > Speaking as part of the larger community, I'd say I'm definitely interested in such work being done and I think the rate at which porting work has taken place since the OpenJDK was released stands as a vindication of that process (we already have a PPC interpreter, the OpenJDK class library also working with cacao and your own Mac OS work). Concerning the list you've supplied here I'd say that the first (merging the BSD work) is more dependent on Sun (like the bug system example) than such a project would warrant, the second two seem two small to warrant a lengthy project (but correct me if I'm wrong), while the last two sound about right to me. As to a native AWT implementation, I believe there was some success getting the Classpath Gtk+ implementation to work with OpenJDK. Given that Gtk+ now supports Quartz as a backend, making the Classpath implementation work with that backend and then the Gtk+ peers work with OpenJDK would seem like a more manageable target than starting from scratch (and would also benefit other Gtk+ users in general and GNU Classpath... ;) ) -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/fbb8b4ff/attachment.html From gnu_andrew at member.fsf.org Fri Jan 4 01:37:48 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Fri, 4 Jan 2008 09:37:48 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <477D39CA.6030003@jazillian.com> References: <477D39CA.6030003@jazillian.com> Message-ID: <17c6771e0801040137l146003a4x845740dd04464d7e@mail.gmail.com> On 03/01/2008, Andy Tripp wrote: > > Hi Ray, > Here are my thoughts... > snip... > - Proposals will be accepted until March 3, 2008. At this time the > proposals will be judged by a team of people (we're thinking 2 from Sun and > 3 from outside Sun). > > Make this "group of 5" be an OpenJDK group, with all (or most) discussions > on an open mailing list. > Agreed, I meant to mention this as well and it was sort of implied in my answer. Everything about the process should be done in the open, including the development work itself. Note that no money will be available until August and all awards must be > distributed at that time. > > It might be worthwhile to have at least one fairly formal milestone for > each project, so the project developers can be sure they're "on the right > track". > Some sort of mid-term review would seem a good idea to me. Maybe a mentor associated with each project too? Again, the nascent nature of OpenJDK may make this difficult and really depends on the willingness of the Sun OpenJDK people. - Projects can only have limited dependence on Sun > involvement/participation. This is for fairness across all projects. > Likewise, projects cannot require a commitment by Sun for significant > time/effort for success since we cannot guarantee adequate Sun resources > will be available -- for example, a project to build a better bug database > for OpenJDK, while very useful, would require heavy involvement by Sun > personnel to integrate it with Sun's internal bug management systems. > > Each proposal should be required to spell out exactly what it requires > from Sun. Using your bug database project example, it might require a > snapshot of the current Sun bug database as simple comma-separated values. > One snapshot at the start of the project and then another snapshot in August > to cut over to the new database. > - What kind of projects do you think would be valuable to the OpenJDK community? > A "How to Hack the OpenJDK" book. > A better bug database (obviously, you've already thought of that :) ). > Massive improvements to the build system (Kelly O'Hair on steriods). > A remote build system that lets a developer make changes locally and > submit changes to a server and get back executables for his platform, > without him having to know anything about how the build works. > A version of Android based on OpenJDK. > As Ray implied, I think anything to do with build/integration issues with the project would rely too much on Sun to be feasible. The first one sounds a very good idea, and I think that involving non-development projects such as documentation, etc. would be a good idea. There are certainly some areas of the public API documentation that need work as well ;) I think some sort of wiki development would be a good idea for such a book, but again we have an integration/project management issue in setting that up. - What selection criteria should be used to choose the best proposals? > > The group should just have some general criteria statement like "Projects > that advance the goals of the OpenJDK project" and then reference a link > that spells out the goals of OpenJDK. Unfortunately, the only such like I > can find is this: http://www.sun.com/software/opensource/java/faq.jsp#e > So maybe spell out the goals yourselves: "increased adoption and increased > innovation" > Agreed. snip... > - Who you think would make good objective judges for the program and why? > > Inside Sun, someone like you and a Mark Reinhold/Danny Coward type of > person. Outside Sun, you just want to be sure to pick people who want to > advance OpenJDK itself, as opposed to pushing some social agenda or some > OpenJDK alternative. > The Governance board springs to mind as an ideal candidate for supplying the judges, especially given these suggestions (it already has a 2 Sun, 3 external makeup IIRC). snip... - How should abandoned or non-completed projects be handled and what should > constitute a "completed" project? > > I presume accepted project participants will have to sign something simple > that says they will do their best to complete > the project by a certain date. It should also mention that if they can't > complete the project, they must post a notice > on the mailing list. > I think an open monitored development process for the projects would handle what you mention here. This is a FOSS project, so the work should be done in the open. Some of what you're saying seems to suggest that the participants disappear in to a black hole for five months and then give a report back at the end. If the process is open (and also, if each project has a mentor attached), then such problems would be flagged much earlier and either dealt with, if possible, or at least some use can be made of the partial work done later. snip.. Good Luck with it! > Andy > > Seconded. I hope this is a successful scheme and can really help OpenJDK take off. -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/f5937a1c/attachment.html From aph at redhat.com Fri Jan 4 02:22:30 2008 From: aph at redhat.com (Andrew Haley) Date: Fri, 4 Jan 2008 10:22:30 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <477D39CA.6030003@jazillian.com> References: <477D39CA.6030003@jazillian.com> Message-ID: <18302.2278.365414.442029@zebedee.pink> Reformatted as plain text. Andy, please set your mail preferences to multipart/mixed HTML and text. Thanks, Andrew. Hi Ray, Here are my thoughts... Ray Gans wrote: Proposals could be made by groups of individuals, existing F/OSS teams, companies/organizations, Java User Groups, etc. Presumably also individuals, right? - Proposals will be accepted until March 3, 2008. At this time the proposals will be judged by a team of people (we're thinking 2 from Sun and 3 from outside Sun). Make this "group of 5" be an OpenJDK group, with all (or most) discussions on an open mailing list. Note that no money will be available until August and all awards must be distributed at that time. It might be worthwhile to have at least one fairly formal milestone for each project, so the project developers can be sure they're "on the right track". - Projects can only have limited dependence on Sun involvement/participation. This is for fairness across all projects. Likewise, projects cannot require a commitment by Sun for significant time/effort for success since we cannot guarantee adequate Sun resources will be available -- for example, a project to build a better bug database for OpenJDK, while very useful, would require heavy involvement by Sun personnel to integrate it with Sun's internal bug management systems. Each proposal should be required to spell out exactly what it requires from Sun. Using your bug database project example, it might require a snapshot of the current Sun bug database as simple comma-separated values. One snapshot at the start of the project and then another snapshot in August to cut over to the new database. - What kind of projects do you think would be valuable to the OpenJDK community? A "How to Hack the OpenJDK" book. A better bug database (obviously, you've already thought of that :) ). Massive improvements to the build system (Kelly O'Hair on steriods). A remote build system that lets a developer make changes locally and submit changes to a server and get back executables for his platform, without him having to know anything about how the build works. A version of Android based on OpenJDK. - What selection criteria should be used to choose the best proposals? The group should just have some general criteria statement like "Projects that advance the goals of the OpenJDK project" and then reference a link that spells out the goals of OpenJDK. Unfortunately, the only such like I can find is this: [1]http://www.sun.com/software/opensource/java/faq.jsp#e So maybe spell out the goals yourselves: "increased adoption and increased innovation" - How many proposals should be accepted? Seven or less sounds good. - Do you think the OpenJDK community at large should have any input into the proposal selection process? Other than to allow anyone to participate in a mailing list discussion, no. - Who you think would make good objective judges for the program and why? Inside Sun, someone like you and a Mark Reinhold/Danny Coward type of person. Outside Sun, you just want to be sure to pick people who want to advance OpenJDK itself, as opposed to pushing some social agenda or some OpenJDK alternative. - What thoughts do you have about how the proposal selection process should be handled? After the proposal submission period ends, publish the proposals, allow a couple weeks of discussion on a mailing list, and then the "group of 5" votes. If greater than 7 submissions, each person may only vote for 7 proposals and the 7 proposals with the most votes win. If 7 or less, A yes/no vote on each. To break down the money awards, after discussion, any of the 5 can submit a proposal for how to split the awards. The first such proposal to get 3 yes votes wins. - Do you think the OpenJDK community at large should have any input into selecting projects that really excel (and be awarded larger prizes)? Again, hold the discussions on open mailing lists and allow anyone to comment. - What criteria should be used to determine the payout for cash awards? Use some vague wording such as "Award amount will be based on potential value to the OpenJDK community". And I would keep it to being based on value, not difficulty. No need for a "degree of difficulty" modifier. - How should abandoned or non-completed projects be handled and what should constitute a "completed" project? I presume accepted project participants will have to sign something simple that says they will do their best to complete the project by a certain date. It should also mention that if they can't complete the project, they must post a notice on the mailing list. - How should awards be handled for project team members who drop out or are added after a proposal is accepted? Ugh. That's a tough one. I guess that this thing that accepted projects would have to sign should designate a single entity (person or corp) that will receive the money, and that Sun will award the money to that entity. Who gets added or deleted or whatever from the project team is then none of Sun's business. You sure don't want to try to micromanage that. So, this "You've been accepted" document might contain: * notice that you've been accepted * expectation that you'll finish by $DATE, that it will be reasonable quality, and it will be substantially what's in the proposal. * promise that Sun will pay the amount of $AMOUNT on date $DATE to $ENTITY upon successful completion * description of any milestones * mention that ultimately, "successful completion" will be determined by this "group of 5". * mention that Sun stays out of issues about who is on the project * mention any help that Sun agrees to (e.g. one bug database dump at start, another at end) * any other required legal mumbo jumbo Please send your thoughts to [2]discuss at openjdk.java.net. Thanks, The OpenJDK team at Sun Good Luck with it! Andy References Visible links 1. http://www.sun.com/software/opensource/java/faq.jsp#e 2. mailto:discuss at openjdk.java.net -- Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK Registered in England and Wales No. 3798903 From ernst.matthias at gmail.com Fri Jan 4 03:02:38 2008 From: ernst.matthias at gmail.com (Matthias Ernst) Date: Fri, 4 Jan 2008 12:02:38 +0100 Subject: Source tarballs In-Reply-To: <477D6553.5000509@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> Message-ID: <22ec15240801040302p7f1469b1ibdb4f0f8103fb1f2@mail.gmail.com> On Jan 3, 2008 11:44 PM, Tom Marble wrote: > Without an OpenJDK upstream tarball this means that the typical > way of getting the OpenJDK source is with "hg fclone" [0]. > Which means that one would need Mercurial [1] with the Forest Extension [2] > (and Mercurial needs python 2.4). This is complicated by the > fact that the Forest Extension (currently, unversioned) is only > supported on Mercurial 0.9.3, 0.9.4, and 0.9.5. While forest does make it easier it is not required. I did not install forest and had good experience with simply cloning the various repositories manually, i.e. hg clone ..../langtools hg clone ..../corba ... If this were copy'n'pasteable on the openjdk homepage it might be good enough? Matthias From gnu_andrew at member.fsf.org Fri Jan 4 04:34:27 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Fri, 4 Jan 2008 12:34:27 +0000 Subject: Source tarballs In-Reply-To: <477D7489.9000205@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> <477D7489.9000205@sun.com> Message-ID: <17c6771e0801040434v3a571da9ued45d5c01bbf34bb@mail.gmail.com> On 03/01/2008, Xiomara Jayasena wrote: > > > Hi Tom, > > Happy New Year! > > Tom Marble wrote: > > I'd like to ask if we (Sun) can reconsider publishing one (1) > source tarball for promoted builds. > > > Yes it is definitely being considered -- as it makes sense, especially if > it is difficult for people to pull down the 7 tar balls that currently exist > now. There are a couple of ways to go about this so that will need to be > sorted out. > > Best regards, > -Xiomara > > > Maybe I'm missing something here, but surely it just takes someone at Sun to pull the b24 changeset using fclone, remove the hg metadata, tarball it and upload it to the site? This would be a lot simpler than pushing the Mercurial+forest extension requirement on to every downstream consumer. -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/e08402b8/attachment.html From dalibor.topic at googlemail.com Fri Jan 4 07:18:49 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Fri, 04 Jan 2008 16:18:49 +0100 Subject: Project Proposal JDK 6 Message-ID: <477E4E59.7010103@kaffe.org> Hi Joe, your proposal seems to have gone lost in the holiday cheer, and I haven't seen any discussion of it on the lists. I think that it's an obvious next step for OpenJDK and would love to see it happen. Do you have a specific group in mind that you'd like to ask to sponsor the project? The build group, also sponsoring the JDK 7 project, would seem to be the right place to go to. cheers, dalibor topic From gnu_andrew at member.fsf.org Fri Jan 4 09:01:03 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Fri, 4 Jan 2008 17:01:03 +0000 Subject: Project Proposal JDK 6 In-Reply-To: <477E4E59.7010103@kaffe.org> References: <477E4E59.7010103@kaffe.org> Message-ID: <17c6771e0801040901w1774ec65m86d48bb8476c5a23@mail.gmail.com> On 04/01/2008, Dalibor Topic wrote: > > Hi Joe, > > your proposal seems to have gone lost in the holiday cheer, and I > haven't seen > any discussion of it on the lists. > > I think that it's an obvious next step for OpenJDK and would love to see > it happen. > > Do you have a specific group in mind that you'd like to ask to sponsor > the project? > The build group, also sponsoring the JDK 7 project, would seem to be the > right place > to go to. > > cheers, > dalibor topic > Hi Joe, Dalibor, I had a similar interest in this project. For one, as I also commented on your blog about this, it seems an ideal opportunity for community involvement with a clear target and fairly low-hanging fruit. I'm assuming that making JDK7 a JDK6 implementation is largely a matter of getting it to pass the JDK6 TCK and not include any superfluous packages/methods/etc. from JDK7. Such cleanup work (rather than fresh development work), obviously with appropriate guidance, would seem to be ideal introductory stuff for a FOSS project, as has been shown by the janitorial tasks on other projects such as Linux and WINE (the latter being my own stepping stone into FOSS development many years ago). However, from your blog, I got the impression that this was instead just something that again would appear from Sun as a complete product. What are your thoughts on this? Cheers, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/ccc8d371/attachment.html From Xiomara.Jayasena at Sun.COM Fri Jan 4 10:45:17 2008 From: Xiomara.Jayasena at Sun.COM (Xiomara Jayasena) Date: Fri, 04 Jan 2008 10:45:17 -0800 Subject: Source tarballs In-Reply-To: <17c6771e0801040434v3a571da9ued45d5c01bbf34bb@mail.gmail.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> <477D7489.9000205@sun.com> <17c6771e0801040434v3a571da9ued45d5c01bbf34bb@mail.gmail.com> Message-ID: <477E7EBD.8080804@sun.com> Andrew John Hughes wrote: > > Yes it is definitely being considered -- as it makes sense, > especially if it is difficult for people to pull down the 7 tar > balls that currently exist now. There are a couple of ways to go > about this so that will need to be sorted out. > > Best regards, > -Xiomara > > > > > Maybe I'm missing something here, but surely it just takes someone at > Sun to pull the b24 changeset using fclone, remove the hg metadata, > tarball it and upload it to the site? > This would be a lot simpler than pushing the Mercurial+forest > extension requirement on to every downstream consumer. Thank you for your input! What you mentioned above is one way to do it. Although I do not think Mercurial is needed to download the zips provided by Mercurial itself. e.g.: http://hg.openjdk.java.net/jdk7/jdk7/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/jdk/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/corba/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/jaxp/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/jaxws/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/hotspot/archive/jdk7-b24.zip http://hg.openjdk.java.net/jdk7/jdk7/langtools/archive/jdk7-b24.zip -Xiomara :-) -- > Andrew :-) > > Help end the Java Trap! > Contribute to GNU Classpath and the OpenJDK > http://www.gnu.org/software/classpath > http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/7f295e6a/attachment.html From robilad at kaffe.org Fri Jan 4 10:51:46 2008 From: robilad at kaffe.org (Dalibor Topic) Date: Fri, 04 Jan 2008 19:51:46 +0100 Subject: Project Proposal JDK 6 In-Reply-To: <477E4E59.7010103@kaffe.org> References: <477E4E59.7010103@kaffe.org> Message-ID: <477E8042.3020802@kaffe.org> Dalibor Topic wrote: > Hi Joe, > > your proposal seems to have gone lost in the holiday cheer, and I > haven't seen > any discussion of it on the lists. > > I think that it's an obvious next step for OpenJDK and would love to > see it happen. > > Do you have a specific group in mind that you'd like to ask to sponsor > the project? > The build group, also sponsoring the JDK 7 project, would seem to be > the right place > to go to. While I'm talking about the build group, I should probably poke Mark and Kelly about this. Mark, Kelly, what do you think? cheers, dalibor topic From gnu_andrew at member.fsf.org Fri Jan 4 10:52:34 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Fri, 4 Jan 2008 18:52:34 +0000 Subject: Source tarballs In-Reply-To: <477E7EBD.8080804@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> <477D7489.9000205@sun.com> <17c6771e0801040434v3a571da9ued45d5c01bbf34bb@mail.gmail.com> <477E7EBD.8080804@sun.com> Message-ID: <17c6771e0801041052p32730fa7o2ca1f5e830f6906e@mail.gmail.com> On 04/01/2008, Xiomara Jayasena wrote: > > Andrew John Hughes wrote: > > Yes it is definitely being considered -- as it makes sense, especially if > > it is difficult for people to pull down the 7 tar balls that currently exist > > now. There are a couple of ways to go about this so that will need to be > > sorted out. > > > > Best regards, > > -Xiomara > > > > > > > > Maybe I'm missing something here, but surely it just takes someone at Sun > to pull the b24 changeset using fclone, remove the hg metadata, tarball it > and upload it to the site? > This would be a lot simpler than pushing the Mercurial+forest extension > requirement on to every downstream consumer. > > > Thank you for your input! > > What you mentioned above is one way to do it. Although I do not think > Mercurial is needed to download the zips provided by Mercurial itself. > > e.g.: > http://hg.openjdk.java.net/jdk7/jdk7/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jdk/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/corba/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jaxp/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jaxws/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/hotspot/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/langtools/archive/jdk7-b24.zip > > > -Xiomara :-) > > > -- > > Andrew :-) > > Help end the Java Trap! > Contribute to GNU Classpath and the OpenJDK > http://www.gnu.org/software/classpath > http://openjdk.java.net > > > Yes, I'm aware of the zips which is probably a slightly cleaner way of doing the tarball than deleting the Mercurial metadata manually. I actually find it quite astonishing that a tarball of a release was just dropped and we are even having to debate this. It was probably tricker to roll a tarball from the internal Sun repositories so why it's not being done just because the source repository is now available is strange. I don't know of any other FOSS project that doesn't have bother available, at least for releases. Thanks, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/3378b953/attachment-0001.html From Xiomara.Jayasena at Sun.COM Fri Jan 4 11:23:05 2008 From: Xiomara.Jayasena at Sun.COM (Xiomara Jayasena) Date: Fri, 04 Jan 2008 11:23:05 -0800 Subject: Source tarballs In-Reply-To: <17c6771e0801041052p32730fa7o2ca1f5e830f6906e@mail.gmail.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> <477D7489.9000205@sun.com> <17c6771e0801040434v3a571da9ued45d5c01bbf34bb@mail.gmail.com> <477E7EBD.8080804@sun.com> <17c6771e0801041052p32730fa7o2ca1f5e830f6906e@mail.gmail.com> Message-ID: <477E8799.2050703@sun.com> Andrew John Hughes wrote: > > > On 04/01/2008, *Xiomara Jayasena* > wrote: > > Andrew John Hughes wrote: >> >> Yes it is definitely being considered -- as it makes sense, >> especially if it is difficult for people to pull down the 7 >> tar balls that currently exist now. There are a couple of >> ways to go about this so that will need to be sorted out. >> >> Best regards, >> -Xiomara >> >> >> >> >> Maybe I'm missing something here, but surely it just takes >> someone at Sun to pull the b24 changeset using fclone, remove the >> hg metadata, tarball it and upload it to the site? >> This would be a lot simpler than pushing the Mercurial+forest >> extension requirement on to every downstream consumer. > > Thank you for your input! > > What you mentioned above is one way to do it. Although I do not > think Mercurial is needed to download the zips provided by > Mercurial itself. > > e.g.: > http://hg.openjdk.java.net/jdk7/jdk7/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jdk/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/corba/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jaxp/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/jaxws/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/hotspot/archive/jdk7-b24.zip > http://hg.openjdk.java.net/jdk7/jdk7/langtools/archive/jdk7-b24.zip > > > > -Xiomara :-) > > > -- >> Andrew :-) >> >> Help end the Java Trap! >> Contribute to GNU Classpath and the OpenJDK >> http://www.gnu.org/software/classpath >> http://openjdk.java.net > > > Yes, I'm aware of the zips which is probably a slightly cleaner way of > doing the tarball than deleting the Mercurial metadata manually. > I actually find it quite astonishing that a tarball of a release was > just dropped and we are even having to debate this. It was probably > tricker > to roll a tarball from the internal Sun repositories so why it's not > being done just because the source repository is now available is strange. > I don't know of any other FOSS project that doesn't have bother > available, at least for releases. There are a few options to get the source now ;-) unlike before. Many things have changed that have gotten us to the point of where we are -- that said nothing is final and we are flexible :-) -Xiomara > > Thanks, > -- > Andrew :-) > > Help end the Java Trap! > Contribute to GNU Classpath and the OpenJDK > http://www.gnu.org/software/classpath > http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/d6ee9ce8/attachment.html From robilad at kaffe.org Fri Jan 4 11:26:27 2008 From: robilad at kaffe.org (Dalibor Topic) Date: Fri, 04 Jan 2008 20:26:27 +0100 Subject: Source tarballs In-Reply-To: <477D6553.5000509@sun.com> References: <475C15F7.7060808@nefkom.net> <475D7D5D.6010602@Sun.COM> <475DCC6D.7020404@nefkom.net> <475DD72F.5030701@sun.com> <475DD81E.3040806@sun.com> <477D6553.5000509@sun.com> Message-ID: <477E8863.3090102@kaffe.org> Tom Marble wrote: > Andreas Sterbenz wrote: > >> Xiomara Jayasena wrote: >> >>> We really didn't see the need, hence we decided to get rid of them. >>> It seems anyone working in JDK 7 may need to become familiar with hg >>> -- that said I appreciate your input. Here at Sun we will no longer >>> be using the tarred source and expect engineers to do clones and that >>> is the same expectation for developers outside of Sun. If many >>> people think this is crucial then the decision can be re-evaluated. >>> >> It is still possible to get source tarballs - by going to >> http://hg.openjdk.java.net/ . Just click on the zip/gz/bz2 link next to >> the desired repository to get the tip (e.g. >> http://hg.openjdk.java.net/jdk7/jdk7/archive/tip.tar.bz2) or navigate to >> the changeset or tag you like and follow the download link (e.g. >> http://hg.openjdk.java.net/jdk7/jdk7/rev/0a5c5386a678). >> >> The only thing it doesn't do is understand forests, so you have to >> download the source for each of the seven repositories in the forest >> separately (/if/ you really want them all). >> > > I'd like to ask if we (Sun) can reconsider publishing one (1) > source tarball for promoted builds. > > This question has come up in the past in the context of many > Free Software distro build daemons which proscribe live Internet > access during binary builds. Usually there is a "source package" > which is uploaded to the build daemon which specifies any build > time dependencies (e.g. specific compilers, header file packages) > and includes the upstream tarball(s). > > This question came up again, today, on IRC in the context of > building OpenJDK for stable distros. Developers stated that it > is very desirable for non-root users to be able to build OpenJDK > on stable distros by making only one simple request to their > system administrator for a list of build dependency packages. > While the user may have Internet access she cannot specify > packages which were not part of the stable release. That makes sense, and usually having some easy way to get the source code for particular snapshots without requiring the installation of a VCS is a nice thing for redistributors, build daemons, etc. The gcc project, for example, publishes weekly snapshots, which are then made available from GNU mirrors like ftp://ftp.mirrorservice.org/sites/sourceware.org/pub/gcc/snapshots/ as regular tar.bz2 archives. cheers, dalibor topic From openjdk at jazillian.com Fri Jan 4 12:34:51 2008 From: openjdk at jazillian.com (Andy Tripp) Date: Fri, 04 Jan 2008 15:34:51 -0500 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <17c6771e0801040137l146003a4x845740dd04464d7e@mail.gmail.com> References: <477D39CA.6030003@jazillian.com> <17c6771e0801040137l146003a4x845740dd04464d7e@mail.gmail.com> Message-ID: <477E986B.70001@jazillian.com> Andrew John Hughes wrote: > Agreed, I meant to mention this as well and it was sort of implied in > my answer. Everything about the process should be done in the open, > including the development work itself. I agree, though it may be difficult for Sun to enforce any specific "openness". Maybe Sun could just add an item to their "Your project has been accepted" letter asking that you keep it as open as possible. > > There are certainly some areas of the public API documentation that > need work as well ;) Which makes me think of another project idea: a project to clean up all the cruft in the bug database. Last I checked, there were over 25,000 bugs, and going through them at random seemed to indicate that most are simple API documentation fixes and clarifications, non-bugs, and other junk. I would bet that most of the active bugs are not being worked and never will be. I'd love to see some iron-fisted person go through and get it down to around 5,000 "real" bugs. But then again, maybe I'm the only one who's bothered by the volume. And I suppose this would require too much work inside Sun also. > > I think an open monitored development process for the projects would > handle what you mention here. This is a FOSS project, so the work > should be done in the open. Some of what you're saying seems to > suggest that the participants disappear in to a black hole for five > months and then give a report back at the end. If the process is open > (and also, if each project has a mentor attached), then such problems > would be flagged much earlier and either dealt with, if possible, or > at least some use can be made of the partial work done later. Yes, it's tough to break out of my closed-source mindset :) On the other hand, I'm thinking that the progress of writing a book or choosing a better bug database is a bit abstract, and not well captured by an SVN tree. From Joe.Darcy at Sun.COM Fri Jan 4 14:08:17 2008 From: Joe.Darcy at Sun.COM (Joseph D. Darcy) Date: Fri, 04 Jan 2008 14:08:17 -0800 Subject: Project Proposal JDK 6 In-Reply-To: <17c6771e0801040901w1774ec65m86d48bb8476c5a23@mail.gmail.com> References: <477E4E59.7010103@kaffe.org> <17c6771e0801040901w1774ec65m86d48bb8476c5a23@mail.gmail.com> Message-ID: <477EAE51.8010704@sun.com> Hello Andrew and Dalibor. Andrew John Hughes wrote: > > > On 04/01/2008, *Dalibor Topic* > wrote: > > Hi Joe, > > your proposal seems to have gone lost in the holiday cheer, and I > haven't seen > any discussion of it on the lists. > > I think that it's an obvious next step for OpenJDK and would love to > see > it happen. > > Do you have a specific group in mind that you'd like to ask to sponsor > the project? > The build group, also sponsoring the JDK 7 project, would seem to be the > right place > to go to. > > cheers, > dalibor topic > > > Hi Joe, Dalibor, > > I had a similar interest in this project. For one, as I also commented > on your blog about this, it seems an ideal opportunity for community > involvement with a clear target and fairly low-hanging fruit. I'm > assuming that making JDK7 a JDK6 implementation is largely a matter of > getting it to pass the JDK6 TCK and not include any superfluous > packages/methods/etc. from JDK7. Such cleanup work (rather than fresh > development work), obviously with appropriate guidance, would seem to be > ideal introductory stuff for a FOSS project, as has been shown by the > janitorial tasks on other projects such as Linux and WINE (the latter > being my own stepping stone into FOSS development many years ago). > However, from your blog, I got the impression that this was instead just > something that again would appear from Sun as a complete product. What > are your thoughts on this? > > Cheers, For the last few weeks I've been on vacation taking in some of that holiday cheer myself and will be officially back to work on Monday :-) As discussed in previous email from Kelly, the Mercurial procedures for JDK repositories aren't fully worked out yet so in the meantime we've been working internally on addressing the "antibugs" of JDK 7 changes inappropriate for a Java SE 6 implementation back in teamware workspaces inside Sun. That effort is already fairly far along, but we'll certainly welcome external community involvement in working on remaining aspects of the project. More later in 2008, regards, -Joe From tony.p.lee at gmail.com Fri Jan 4 17:03:55 2008 From: tony.p.lee at gmail.com (Tony Lee) Date: Fri, 4 Jan 2008 17:03:55 -0800 Subject: Building openjdk for ARM Linux. Message-ID: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> Hello, I am new to jdk but very comfortable with compiling opensource packages. I try to build jdk7 on ARM Linux platform with Fedora Core 6. The build failed with the following message: >From the msg, it looks like I need to install older version of jdk before I compile for jdk7, is this correct? Has anyone succeed in compile the openjdk 7 for ARM linux platform? --------------------Error msg -------------------------- -bash-3.1# make /bin/sh: /NOT-SET/devtools/share/ant/latest/bin/ant: No such file or directory /bin/sh: /NOT-SET/devtools/share/findbugs/latest/bin/findbugs: No such file or directory ../make/common/shared/Sanity-Settings.gmk:121: WARNING: ANT_VER should not be empty [Sanity-Settings.gmk] ../make/common/shared/Sanity-Settings.gmk:122: WARNING: FINDBUGS_VER should not be empty [Sanity-Settings.gmk] /bin/sh: line 0: [: /bin/sh:: integer expression expected /bin/sh: line 0: [: /bin/sh:: integer expression expected linux armv5tejl 1.7.0-internal build started: 08-01-04 16:09 make[1]: Entering directory `/share/openjdk/compiler/jdk-jdk7-b24/make/tools/freetypecheck' freetypecheck.c: In function ???main???: freetypecheck.c:45: warning: comparison is always false due to limited range of data type freetypecheck.c:54: warning: comparison is always false due to limited range of data type make[1]: Leaving directory `/share/openjdk/compiler/jdk-jdk7-b24/make/tools/freetypecheck' Bootstrap Settings: BOOTDIR = /NO_BOOTDIR ALT_BOOTDIR = BOOT_VER = /bin/sh: /NO_BOOTDIR/bin/java: No such file or directory [requires at least 1.5] OUTPUTDIR = ./../build/linux-armv5tejl ALT_OUTPUTDIR = ABS_OUTPUTDIR = /share/openjdk/compiler/jdk-jdk7-b24/build/linux-armv5tejl -- -Tony Having fun with FPGA, gige, mpeg, snmp, ppc, Linux, ARM. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080104/1ed98b71/attachment.html From Bradford.Wetmore at Sun.COM Fri Jan 4 19:10:58 2008 From: Bradford.Wetmore at Sun.COM (Brad Wetmore) Date: Fri, 04 Jan 2008 19:10:58 -0800 Subject: Building openjdk for ARM Linux. In-Reply-To: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> Message-ID: <477EF542.4030303@sun.com> Tony Lee wrote: > >From the msg, it looks like I need to install older version of jdk > before I compile for jdk7, is this correct? Yes. You need a "bootstrap" JDK for various parts of the build. Please see: http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html specifically: http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html#bootjdk http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html#linux Hope this helps. Brad From aph at redhat.com Sat Jan 5 02:53:58 2008 From: aph at redhat.com (Andrew Haley) Date: Sat, 5 Jan 2008 10:53:58 +0000 Subject: Building openjdk for ARM Linux. In-Reply-To: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> Message-ID: <18303.25030.93337.109497@zebedee.pink> Tony Lee writes: > Hello, > > I am new to jdk but very comfortable with compiling opensource packages. > > I try to build jdk7 on ARM Linux platform with Fedora Core 6. > > The build failed with the following message: > > > >From the msg, it looks like I need to install older version of jdk before I > compile for jdk7, is this correct? > > Has anyone succeed in compile the openjdk 7 for ARM linux platform? A great deal of OpenJDK is written in assembly language. Unless you've written an ARM version of all this it isn't going to work. Andrew. -- Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK Registered in England and Wales No. 3798903 From pieter.libin at mybiodata.eu Sat Jan 5 05:37:08 2008 From: pieter.libin at mybiodata.eu (Pieter Libin) Date: Sat, 5 Jan 2008 14:37:08 +0100 Subject: Building openjdk for ARM Linux. In-Reply-To: <18303.25030.93337.109497@zebedee.pink> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> <18303.25030.93337.109497@zebedee.pink> Message-ID: <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> I thought there also was a c++ interpreter available, maybe it's a good idea to start with this instead of writing all the ARM assembly. Pieter On Jan 5, 2008 11:53 AM, Andrew Haley wrote: > Tony Lee writes: > > Hello, > > > > I am new to jdk but very comfortable with compiling opensource packages. > > > > I try to build jdk7 on ARM Linux platform with Fedora Core 6. > > > > The build failed with the following message: > > > > > > >From the msg, it looks like I need to install older version of jdk before I > > compile for jdk7, is this correct? > > > > Has anyone succeed in compile the openjdk 7 for ARM linux platform? > > A great deal of OpenJDK is written in assembly language. Unless > you've written an ARM version of all this it isn't going to work. > > Andrew. > > -- > Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK > Registered in England and Wales No. 3798903 > From David.Herron at Sun.COM Sat Jan 5 08:18:48 2008 From: David.Herron at Sun.COM (David Herron) Date: Sat, 05 Jan 2008 08:18:48 -0800 Subject: Building openjdk for ARM Linux. In-Reply-To: <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> <18303.25030.93337.109497@zebedee.pink> <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> Message-ID: <477FADE8.7080904@sun.com> Gary Benson has been making some interesting postings along these lines. http://gbenson.livejournal.com/ - David Herron Pieter Libin wrote: > I thought there also was a c++ interpreter available, > maybe it's a good idea to start with this instead of writing all the > ARM assembly. > > Pieter > > On Jan 5, 2008 11:53 AM, Andrew Haley wrote: > >> Tony Lee writes: >> > Hello, >> > >> > I am new to jdk but very comfortable with compiling opensource packages. >> > >> > I try to build jdk7 on ARM Linux platform with Fedora Core 6. >> > >> > The build failed with the following message: >> > >> > >> > >From the msg, it looks like I need to install older version of jdk before I >> > compile for jdk7, is this correct? >> > >> > Has anyone succeed in compile the openjdk 7 for ARM linux platform? >> >> A great deal of OpenJDK is written in assembly language. Unless >> you've written an ARM version of all this it isn't going to work. >> >> Andrew. >> >> -- >> Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK >> Registered in England and Wales No. 3798903 >> >> From tony.p.lee at gmail.com Sat Jan 5 14:29:05 2008 From: tony.p.lee at gmail.com (Tony Lee) Date: Sat, 5 Jan 2008 14:29:05 -0800 Subject: Building openjdk for ARM Linux. In-Reply-To: <477EF542.4030303@sun.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> <477EF542.4030303@sun.com> Message-ID: <470b63970801051429gfc469a4mddad21c049227a5b@mail.gmail.com> On Jan 4, 2008 7:10 PM, Brad Wetmore wrote: > Yes. You need a "bootstrap" JDK for various parts of the build. > > Please see: > > http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html > > specifically: > > > http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html#bootjdk > http://hg.openjdk.java.net/jdk7/jdk7/raw-file/tip/README-builds.html#linux > > Hope this helps. > > Thanks Brad. I guess the answer is no for now. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080105/583b9b37/attachment.html From gnu_andrew at member.fsf.org Sat Jan 5 16:56:43 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Sun, 6 Jan 2008 00:56:43 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <477E986B.70001@jazillian.com> References: <477D39CA.6030003@jazillian.com> <17c6771e0801040137l146003a4x845740dd04464d7e@mail.gmail.com> <477E986B.70001@jazillian.com> Message-ID: <17c6771e0801051656v5bd33351ubb4f559e61962b3f@mail.gmail.com> On 04/01/2008, Andy Tripp wrote: > > Andrew John Hughes wrote: > > Agreed, I meant to mention this as well and it was sort of implied in > > my answer. Everything about the process should be done in the open, > > including the development work itself. > I agree, though it may be difficult for Sun to enforce any specific > "openness". Maybe Sun could just add an > item to their "Your project has been accepted" letter asking that you > keep it as open as possible. Well, in running the program, I believe Sun have some power over things. At the very minimum, it should be required that all work is uploaded somewhere at the end of the time period. This is what happens with Google's Summer of Code -- code.google.com has a project which hosts all the code created during this summer's event. > > > There are certainly some areas of the public API documentation that > > need work as well ;) > Which makes me think of another project idea: a project to clean up all > the cruft in the bug database. Last I checked, there > were over 25,000 bugs, and going through them at random seemed to > indicate that most are simple API documentation fixes and > clarifications, non-bugs, and other junk. I would bet that most of the > active bugs are not being worked and never will be. > I'd love to see some iron-fisted person go through and get it down to > around 5,000 "real" bugs. But then again, maybe I'm > the only one who's bothered by the volume. And I suppose this would > require too much work inside Sun also. This would work fine if OpenJDK already had a usable bug database like other FOSS projects (e.g. I could see such a thing being done fairly easily with GNU Classpath, which has the same issue but on a much smaller scale). Given that bug reports currently require all sorts of internal Sun approval, this just wouldn't be practical. That said, I agree it's an issue and again a nice piece of low-hanging fruit when things get going properly. > > > I think an open monitored development process for the projects would > > handle what you mention here. This is a FOSS project, so the work > > should be done in the open. Some of what you're saying seems to > > suggest that the participants disappear in to a black hole for five > > months and then give a report back at the end. If the process is open > > (and also, if each project has a mentor attached), then such problems > > would be flagged much earlier and either dealt with, if possible, or > > at least some use can be made of the partial work done later. > Yes, it's tough to break out of my closed-source mindset :) On the other > hand, I'm thinking that the progress of writing a book > or choosing a better bug database is a bit abstract, and not well > captured by an SVN tree. > > Well, yes, a design process like choosing and implementing a new bug database is more a case for a wiki I think, but again, this was given as an impractical project in the first e-mail anyway! Certainly a book could be developed collaboratively using a version control system. Most FOSS project documentation is developed along with the source code base in this manner, and I can think of a few examples of books done this way (e.g. the one for SVN itself IIRC). They can be used for more than just source code. Thanks, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080106/bb0cbb45/attachment.html From arnd-hendrik.mathias at nefkom.net Sat Jan 5 17:54:48 2008 From: arnd-hendrik.mathias at nefkom.net (Arnd-Hendrik Mathias) Date: Sun, 06 Jan 2008 02:54:48 +0100 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: References: Message-ID: <478034E8.5000207@nefkom.net> Hi everybody, Ray Gans wrote: > - What kind of projects do you think would be valuable to the OpenJDK > community? Personally, I'd love to see the OpenJDK build environment get rid of the dependencies stuff like - pre-installed bootstrap JDK - binary plugs - ANT (again depending on a pre-installed JDK) ... (...sure, the rest of the "usual" make/build/... tools can remain... ;?) Aiming at an OpenJDK capable to be grabbed as tarball and be built from scratch on a new system. Best regards Arnd-Hendrik -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080106/08731925/attachment.html From gnu_andrew at member.fsf.org Sat Jan 5 18:49:04 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Sun, 6 Jan 2008 02:49:04 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <478034E8.5000207@nefkom.net> References: <478034E8.5000207@nefkom.net> Message-ID: <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> On 06/01/2008, Arnd-Hendrik Mathias wrote: > > Hi everybody, > > Ray Gans wrote: > > - What kind of projects do you think would be valuable to the OpenJDK > community? > > Personally, I'd love to see the OpenJDK build environment get rid of the > dependencies stuff like > - pre-installed bootstrap JDK > - binary plugs > - ANT (again depending on a pre-installed JDK) > ... > (...sure, the rest of the "usual" make/build/... tools can remain... ;?) > Aiming at an OpenJDK capable to be grabbed as tarball and be built from > scratch on a new system. > Best regards > > Arnd-Hendrik > Have you looked at IcedTea? (http://icedtea.classpath.org) It at least gets rid of the binary plugs issue and goes a long way to making the build process usable. Hopefully Sun will integrate these changes at some point. The issue of having a Java VM+class library already will remain a problem because there are no non-Java compilers any more (for 1.5 on). It's pretty much Sun's or the Eclipse compiler, both of which are written in Java, so you need a Java VM + class library to run them. I'd love to see Ant disappear as a requirement, but Sun are moving in the opposite direction (using the Ant build for langtools, etc. was introduced in b21). If it could be built using the autotools like most FOSS projects, that would be ideal. -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080106/0dad9744/attachment.html From arnd-hendrik.mathias at nefkom.net Sun Jan 6 15:47:16 2008 From: arnd-hendrik.mathias at nefkom.net (Arnd-Hendrik Mathias) Date: Mon, 07 Jan 2008 00:47:16 +0100 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> Message-ID: <47816884.8040604@nefkom.net> Hi Andrew, Andrew John Hughes wrote: > Have you looked at IcedTea? (http://icedtea.classpath.org) Thanks for the hint. It looks cool. > It at least gets rid of the binary plugs issue and goes a long way to > making the build process usable. Hopefully Sun will integrate these > changes at some point. I'd welcome this to happen. > The issue of having a Java VM+class library already will remain a > problem because there are no non-Java compilers any more (for 1.5 > on). It's pretty much Sun's or the Eclipse compiler, both of which > are written in Java, so you need a Java VM + class library to run them. How about some concept like using some java2native compilers or some java enabled gcc to build the necessary set of bootstrap JDK components in a pre-build-step and then using these for building the real OpenJDK. > I'd love to see Ant disappear as a requirement, but Sun are moving in > the opposite direction (using the Ant build for langtools, etc. was > introduced in b21). Sounds not so good. > If it could be built using the autotools like most FOSS projects, that > would be ideal. Yes! Yes! Yes!... ;?) By the way...as another "proposal" (already asked a number of times by different people): Are there currently any plans or activities to create an OpenJDK SeaMonkey(Mozilla)-plugin? Arnd-Hendrik From gnu_andrew at member.fsf.org Sun Jan 6 16:43:55 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Mon, 7 Jan 2008 00:43:55 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <47816884.8040604@nefkom.net> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> Message-ID: <17c6771e0801061643w320d8032l86796d779773152@mail.gmail.com> snip... > > The issue of having a Java VM+class library already will remain a > > problem because there are no non-Java compilers any more (for 1.5 > > on). It's pretty much Sun's or the Eclipse compiler, both of which > > are written in Java, so you need a Java VM + class library to run them. > How about some concept like using some java2native compilers or some > java enabled gcc to build the necessary set of bootstrap JDK components > in a pre-build-step and then using these for building the real OpenJDK. This is what I've been looking at with IcePick (mentioned on the IcedTea website). This can take just the langtools repository from OpenJDK and build it using an autotools setup to give you a copy of javac, javap, javadoc and apt. One feature I'm going to add there is being able to do a native compile of the tools as well, which would remove the remaining requirement of a VM+class library. snip... > > By the way...as another "proposal" (already asked a number of times by > different people): Are there currently any plans or activities to create > an OpenJDK SeaMonkey(Mozilla)-plugin? > gcjwebplugin is in IcedTea. > Arnd-Hendrik > Thanks, -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net From gbenson at redhat.com Mon Jan 7 00:43:36 2008 From: gbenson at redhat.com (Gary Benson) Date: Mon, 7 Jan 2008 08:43:36 +0000 Subject: Building openjdk for ARM Linux. In-Reply-To: <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> <18303.25030.93337.109497@zebedee.pink> <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> Message-ID: <20080107084335.GA3799@redhat.com> Pieter Libin wrote: > I thought there also was a c++ interpreter available, > maybe it's a good idea to start with this instead of writing all > the ARM assembly. The C++ interpreter still requires a fair amount of assembly, and platform specific code in general. Check out the ports directory in IcedTea: there's a bare minimum ppc port in there which is is essentially what you need to write for any new processor. Cheers, Gary From aph at redhat.com Mon Jan 7 02:03:19 2008 From: aph at redhat.com (Andrew Haley) Date: Mon, 7 Jan 2008 10:03:19 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <47816884.8040604@nefkom.net> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> Message-ID: <18305.63719.125363.896219@zebedee.pink> Arnd-Hendrik Mathias writes: > How about some concept like using some java2native compilers or > some java enabled gcc to build the necessary set of bootstrap JDK > components in a pre-build-step and then using these for building > the real OpenJDK. That's how IcedTea works. Andrew. -- Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK Registered in England and Wales No. 3798903 From gnu_andrew at member.fsf.org Mon Jan 7 02:06:24 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Mon, 7 Jan 2008 10:06:24 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <18305.63719.125363.896219@zebedee.pink> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> <18305.63719.125363.896219@zebedee.pink> Message-ID: <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> On 07/01/2008, Andrew Haley wrote: > > Arnd-Hendrik Mathias writes: > > > How about some concept like using some java2native compilers or > > some java enabled gcc to build the necessary set of bootstrap JDK > > components in a pre-build-step and then using these for building > > the real OpenJDK. > > That's how IcedTea works. > > Andrew. > > -- > Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, > Berkshire, SL4 1TE, UK > Registered in England and Wales No. 3798903 > It doesn't do anything native, AFAIK -- it just creates a pseudo-bootstrap JDK. -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080107/0e6e59f2/attachment.html From aph at redhat.com Mon Jan 7 03:45:59 2008 From: aph at redhat.com (Andrew Haley) Date: Mon, 7 Jan 2008 11:45:59 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> <18305.63719.125363.896219@zebedee.pink> <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> Message-ID: <18306.4343.714808.312952@zebedee.pink> Andrew John Hughes writes: > On 07/01/2008, Andrew Haley wrote: > > > > Arnd-Hendrik Mathias writes: > > > > > How about some concept like using some java2native compilers or > > > some java enabled gcc to build the necessary set of bootstrap JDK > > > components in a pre-build-step and then using these for building > > > the real OpenJDK. > > > > That's how IcedTea works. > > It doesn't do anything native, AFAIK -- it just creates a pseudo-bootstrap > JDK. What does "it doesn't do anything native, AFAIK" mean? I can't tell if you agree with me or not. Andrew. -- Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK Registered in England and Wales No. 3798903 From gnu_andrew at member.fsf.org Mon Jan 7 05:08:30 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Mon, 7 Jan 2008 13:08:30 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <18306.4343.714808.312952@zebedee.pink> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> <18305.63719.125363.896219@zebedee.pink> <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> <18306.4343.714808.312952@zebedee.pink> Message-ID: <17c6771e0801070508s52a9544fy38efccc395dfaf6f@mail.gmail.com> On 07/01/2008, Andrew Haley wrote: > > Andrew John Hughes writes: > > On 07/01/2008, Andrew Haley wrote: > > > > > > Arnd-Hendrik Mathias writes: > > > > > > > How about some concept like using some java2native compilers or > > > > some java enabled gcc to build the necessary set of bootstrap JDK > > > > components in a pre-build-step and then using these for building > > > > the real OpenJDK. > > > > > > That's how IcedTea works. > > > > It doesn't do anything native, AFAIK -- it just creates a > pseudo-bootstrap > > JDK. > > What does "it doesn't do anything native, AFAIK" mean? I can't tell > if you agree with me or not. > > Andrew. > > -- > Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, > Berkshire, SL4 1TE, UK > Registered in England and Wales No. 3798903 > I interpreted the question (perhaps wrongly) as looking for a solution that wouldn't require a Java VM and class library i.e. that there would be a preliminary step that created a native toolchain of javac, etc. which could then be used to bootstrap the OpenJDK in full. IcedTea only does half of this; it creates the pseudo-bootstrap environment using ecj and a Java VM, but this is not native. I'm not sure how much of an issue that is, as you'd still need libgcj to run the native binaries. So I suppose I half agree with what you said... :D -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080107/4ec84e9b/attachment.html From aph at redhat.com Mon Jan 7 05:46:09 2008 From: aph at redhat.com (Andrew Haley) Date: Mon, 7 Jan 2008 13:46:09 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <17c6771e0801070508s52a9544fy38efccc395dfaf6f@mail.gmail.com> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> <18305.63719.125363.896219@zebedee.pink> <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> <18306.4343.714808.312952@zebedee.pink> <17c6771e0801070508s52a9544fy38efccc395dfaf6f@mail.gmail.com> Message-ID: <18306.11553.202792.517850@zebedee.pink> Andrew John Hughes writes: > On 07/01/2008, Andrew Haley wrote: > > > > Andrew John Hughes writes: > > > On 07/01/2008, Andrew Haley wrote: > > > > > > > > Arnd-Hendrik Mathias writes: > > > > > > > > > How about some concept like using some java2native compilers or > > > > > some java enabled gcc to build the necessary set of bootstrap JDK > > > > > components in a pre-build-step and then using these for building > > > > > the real OpenJDK. > > > > > > > > That's how IcedTea works. > > > > > > It doesn't do anything native, AFAIK -- it just creates a > > pseudo-bootstrap > > > JDK. > > > > What does "it doesn't do anything native, AFAIK" mean? I can't tell > > if you agree with me or not. > > I interpreted the question (perhaps wrongly) as looking for a > solution that wouldn't require a Java VM and class library > i.e. that there would be a preliminary step that created a native > toolchain of javac, etc. which could then be used to bootstrap the > OpenJDK in full. IcedTea only does half of this; it creates the > pseudo-bootstrap environment using ecj and a Java VM, but this is > not native. I still don't know what you mean by this. The OP said "use some java enabled gcc to build the necessary set of bootstrap JDK components in a pre-build-step." That's exactly what we do with gcj! > I'm not sure how much of an issue that is, as you'd still need > libgcj to run the native binaries. None at all, I wouldn't have thought. Andrew. -- Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK Registered in England and Wales No. 3798903 From gnu_andrew at member.fsf.org Mon Jan 7 06:51:25 2008 From: gnu_andrew at member.fsf.org (Andrew John Hughes) Date: Mon, 7 Jan 2008 14:51:25 +0000 Subject: Feedback request: OpenJDK Community Innovator's Challenge Grants In-Reply-To: <18306.11553.202792.517850@zebedee.pink> References: <478034E8.5000207@nefkom.net> <17c6771e0801051849n4dd8cc50kdc8553da54b57c85@mail.gmail.com> <47816884.8040604@nefkom.net> <18305.63719.125363.896219@zebedee.pink> <17c6771e0801070206k4b10d300h5544b40c67679c1@mail.gmail.com> <18306.4343.714808.312952@zebedee.pink> <17c6771e0801070508s52a9544fy38efccc395dfaf6f@mail.gmail.com> <18306.11553.202792.517850@zebedee.pink> Message-ID: <17c6771e0801070651m645b5692m316a9ca630547a7a@mail.gmail.com> On 07/01/2008, Andrew Haley wrote: > > Andrew John Hughes writes: > > On 07/01/2008, Andrew Haley wrote: > > > > > > Andrew John Hughes writes: > > > > On 07/01/2008, Andrew Haley wrote: > > > > > > > > > > Arnd-Hendrik Mathias writes: > > > > > > > > > > > How about some concept like using some java2native compilers or > > > > > > some java enabled gcc to build the necessary set of bootstrap > JDK > > > > > > components in a pre-build-step and then using these for building > > > > > > the real OpenJDK. > > > > > > > > > > That's how IcedTea works. > > > > > > > > It doesn't do anything native, AFAIK -- it just creates a > > > pseudo-bootstrap > > > > JDK. > > > > > > What does "it doesn't do anything native, AFAIK" mean? I can't tell > > > if you agree with me or not. > > > > I interpreted the question (perhaps wrongly) as looking for a > > solution that wouldn't require a Java VM and class library > > i.e. that there would be a preliminary step that created a native > > toolchain of javac, etc. which could then be used to bootstrap the > > OpenJDK in full. IcedTea only does half of this; it creates the > > pseudo-bootstrap environment using ecj and a Java VM, but this is > > not native. > > I still don't know what you mean by this. The OP said "use some java > enabled gcc to build the necessary set of bootstrap JDK components in > a pre-build-step." That's exactly what we do with gcj! > > > I'm not sure how much of an issue that is, as you'd still need > > libgcj to run the native binaries. > > None at all, I wouldn't have thought. > > Andrew. > > -- > Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, > Berkshire, SL4 1TE, UK > Registered in England and Wales No. 3798903 > I guess it was the mention of 'java2native compilers' that confused me -- I thought he specifically wanted a native solution, and we were discussing the bootstrap issue of every 1.5 compiler being written in Java. Yes, IcedTea works with gcj but equally what it does could be achieved with any of the Classpath VMs. -- Andrew :-) Help end the Java Trap! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.openjdk.java.net/pipermail/discuss/attachments/20080107/20641e5e/attachment-0001.html From Steve.Goldman at Sun.COM Mon Jan 7 06:59:40 2008 From: Steve.Goldman at Sun.COM (steve goldman) Date: Mon, 07 Jan 2008 09:59:40 -0500 Subject: Building openjdk for ARM Linux. In-Reply-To: <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> References: <470b63970801041703t24e9e121y39681700a6b5265f@mail.gmail.com> <18303.25030.93337.109497@zebedee.pink> <50be7b020801050537l621861a5nd398372679fc29f0@mail.gmail.com> Message-ID: <47823E5C.2050505@sun.com> Pieter Libin wrote: > I thought there also was a c++ interpreter available, > maybe it's a good idea to start with this instead of writing all the > ARM assembly. > There is still plenty of assembly code (and an assembler) that you need to write even if you use the c++ interpreter. -- Steve From Kelly.Ohair at Sun.COM Wed Jan 9 11:58:30 2008 From: Kelly.Ohair at Sun.COM (Kelly O'Hair) Date: Wed, 09 Jan 2008 11:58:30 -0800 Subject: JDK 7 build 24 is available at the openjdk.java.net website In-Reply-To: <476E835E.1020308@nefkom.net> References: <47563FCC.4080907@sun.com> <475D7AB3.8060000@redhat.com> <475DDA7F.1080302@sun.com> <47606D05.2040709@nefkom.net> <476AE241.1050801@sun.com> <476AF2A7.1000009@nefkom.net> <476B1835.6010408@sun.com> <476E835E.1020308@nefkom.net> Message-ID: <47852766.3000404@sun.com> Adding build-dev to the CC list, seems like these are build issues... Arnd-Hendrik Mathias wrote: > Hi Andreas, > > Andreas Sterbenz wrote: >> See below output of a rebuild of make/javax/crypto after a previous >> build of a fresh clone of the Mercurial jdk repository. If you see >> something different, that must be related to your build environment, >> e.g. version of make used, out of date binary plugs, etc. > Since my output looked quite different, it was not really comparable > with the correct output you posted. After further examination of the > make-/build environment, I found out that the make environment assumes > some tools to reside in /usr/bin per default. I found that this can be > overwritten by the ALT_USRBIN_PATH variable. However, if tools reside in > different directories (/bin /usr/bin ...) setting > > ALT_USRBIN_PATH= > > and having the PATH variable set to have all these tools available > directly may be the method the most likely to work. In my special case > "head" was missing, residing in /bin on my system. At some point in time (a long long time ago?), it was decided that the jdk Makefiles should always run utilities with a full pathname and not rely on the PATH setting. I kind of understand what the issue was, PATH ordering issues can be extremely hard to diagnose. But I myself find it a maintenance nightmare because of what you have run into. I don't know if this will change, and there is some risk of relying on PATH especially on Windows, but just so you know, this is a known issue and debate. I tend to think that we should rely on PATH more, especially when it relates to the basic utility commands like 'cd', 'sed', etc. I hate having to use '$(CD)' in the makefiles instead of just 'cd'. :^( Maybe a PATH sanity check on the utilities used somehow, devise a simple use case for each utility used, and verify the utility from PATH is acceptable... Ah... but I know what you will say, just use autotools... well... does autotools work on Windows with VS2003? :^( > > After changing this setting jdk7 build process ran a bit longer, until > it ran into another problem: > Linking of a mass of object files from the jdk/make/sun/splashscreen > directory results in a number of missing symbols from the png library: > > /opt/jdk/openjdk-1.7.0_b24/tmp/sun/sun.awt/splashscreen/obj64/png.o: In > function `png_init_mmx_flags': > png.c:(.text+0xbc): undefined reference to `png_mmx_support' > /opt/jdk/openjdk-1.7.0_b24/tmp/sun/sun.awt/splashscreen/obj64/pngpread.o: > In function `png_push_process_row': > pngpread.c:(.text+0x89e): undefined reference to `png_read_filter_row' > pngpread.c:(.text+0x964): undefined reference to `png_do_read_interlace' > /opt/jdk/openjdk-1.7.0_b24/tmp/sun/sun.awt/splashscreen/obj64/pngpread.o: > In function `png_progressive_combine_row': > pngpread.c:(.text+0x141): undefined reference to `png_combine_row' > /opt/jdk/openjdk-1.7.0_b24/tmp/sun/sun.awt/splashscreen/obj64/pngread.o: > In function `png_read_row': > pngread.c:(.text+0xb64): undefined reference to `png_combine_row' > pngread.c:(.text+0xb79): undefined reference to `png_combine_row' > pngread.c:(.text+0xc20): undefined reference to `png_read_filter_row' > pngread.c:(.text+0xc6a): undefined reference to `png_combine_row' > pngread.c:(.text+0xc92): undefined reference to `png_combine_row' > pngread.c:(.text+0xc9f): undefined reference to `png_do_read_interlace' > pngread.c:(.text+0xd10): undefined reference to `png_combine_row' > collect2: ld returned 1 exit status > > Extending the OTHER_LDLIBS variable in the > jdk/make/sun/splashscreen/Makefile by > > -lpng > > helps to remove the undefined reference to `png_mmx_support'. The other > undefined references remain. > After taking a closer look at the libpng.so by e.g. > > nm /usr/lib/libpng.so | grep png_read_filter_row > > I found out that these symbols exist in the text segment, but only as > internal symbols: > > 0000000000007de0 t png_read_filter_row > > How come they can link in your environment at all? Have you built your > libpng in a special way and is there a workaround to replace these > library calls by OpenJDK-internal implementations or something likewise? Sounds like the above is a known png issue... > > Anyway, I still got one other question: What is the background of all of > those overwrite variables in the jdk/make/common/Defs-linux.gmk? > Omitting the overwrite directive would ease redefining some environment > definitions like OPENWIN_HOME or OPENWIN_LIB from command line, for > which no ALT_... variables exist. This is the same question I have. These have been handed down over time and I'm not exactly sure why the override (you did mean override, right?) was used so much. I consider the use of override bad style, figuring out how make variables get defined is hard enough without complicating things with an override setting. The 'override VAR += XXX' might make sense, but even then you can get the same results by introducing another variable, and it might be more obvious what's happening that way, but we don't use the += with override. (For those of you unfamiliar with GNU make 'override', see http://www.gnu.org/software/make/manual/make.html#Override-Directive) I do know that the name TMPDIR was an issue, but we changed that to TEMPDIR a long time ago and the override on TEMPDIR is probably not necessary anymore. (TMPDIR was treated special in Linux or GNU make at some point??? but it's a name that should be avoided, or at least avoided as your own variable name :^). But I'm a little nervous about all these Linux variables that use override: --- override ALT_CODESET_KEY = _NL_CTYPE_CODESET_NAME override AWT_RUNPATH = override HAVE_ALTZONE = false override HAVE_FILIOH = false override HAVE_GETHRTIME = false override HAVE_GETHRVTIME = false override HAVE_SIGIGNORE = true override LEX_LIBRARY = -lfl ifeq ($(STATIC_CXX),true) override LIBCXX = -Wl,-Bstatic -lstdc++ -lgcc -Wl,-Bdynamic else override LIBCXX = -lstdc++ endif override LIBPOSIX4 = override LIBSOCKET = override LIBTHREAD = override MOOT_PRIORITIES = true override NO_INTERRUPTIBLE_IO = true override OPENWIN_HOME = /usr/X11R6 ifeq ($(ARCH), amd64) override OPENWIN_LIB = $(OPENWIN_HOME)/lib64 else override OPENWIN_LIB = $(OPENWIN_HOME)/lib endif override OTHER_M4FLAGS = -D__GLIBC__ -DGNU_ASSEMBLER override SUN_CMM_SUBDIR = override THREADS_FLAG = native override USE_GNU_M4 = true override USING_GNU_TAR = true override WRITE_LIBVERSION = false # USE_EXECNAME forces the launcher to look up argv[0] on $PATH, and put the # resulting resolved absolute name of the executable in the environment # variable EXECNAME. That executable name is then used that to locate the # installation area. override USE_EXECNAME = true --- Some I recognize, but every one of these variables would need to be investigated to see why the override was necessary and if we can safely remove it. Why not allow someone to set these on the command line if they wanted? I'm certainly in favor of removing unnecessary override's. -kto > Best regards > > Arnd-Hendrik From Debra.Scott at Sun.COM Wed Jan 9 16:25:43 2008 From: Debra.Scott at Sun.COM (Debra Scott) Date: Wed, 09 Jan 2008 18:25:43 -0600 Subject: JDK documentation Message-ID: <47856607.4040202@sun.com> Hi All, I'd like to get people's thoughts on the documentation for the JDK platform. * How many people feel it is important for the the docs to be open sourced so the community can contribute to them? * How many people have seen at least one occasion where they would have added something to the documentation, or made a correction, if they had been able? (a large or small percentage?) * Given that we need documentation in a structured format that allows for content sharing and multiple delivery vehicles, does anyone have any recommendations for a system that is both easily editable, like a Wiki, and easily repurposable, like XML-structured docs? * If so, do people know of conversion tools, and/or would they be willing to help with conversions? * If not, would they be willing to participate in the development of such a system? -Thanks for your feedback, Debbie -- ========================================================================== Debra Scott 39 Golden Wheat Ln Manager of Java SE Documentation Wrightstown, WI 54180 Information Products Group Phone: 877-219-2362 x51743 Global Product Development & Operations, Sun Microsystems, Inc ========================================================================== From dalibor.topic at googlemail.com Wed Jan 9 16:29:01 2008 From: dalibor.topic at googlemail.com (Dalibor Topic) Date: Thu, 10 Jan 2008 01:29:01 +0100 Subject: JDK 7 build 24 is available at the openjdk.java.net website In-Reply-To: <47852766.3000404@sun.com> References: <47563FCC.4080907@sun.com> <475D7AB3.8060000@redhat.com> <475DDA7F.1080302@sun.com> <47606D05.2040709@nefkom.net> <476AE241.1050801@sun.com> <476AF2A7.1000009@nefkom.net> <476B1835.6010408@sun.com> <476E835E.1020308@nefkom.net> <47852766.3000404@sun.com> Message-ID: <478566CD.5060600@kaffe.org> Kelly O'Hair wrote: > Maybe a PATH sanity check on the utilities used somehow, devise a simple > use case for each utility used, and verify the utility from PATH is > acceptable... Ah... but I know what you will say, just use autotools... > well... does autotools work on Windows with VS2003? :^( Last time around I had to use msvc 2003 from autotools, I used http://cccl.sourceforge.net/ to make it work, and it worked for me. cheers, dalibor topic From scolebourne at joda.org Wed Jan 9 16:56:20 2008 From: scolebourne at joda.org (Stephen Colebourne) Date: Thu, 10 Jan 2008 00:56:20 +0000 Subject: JDK documentation In-Reply-To: <47856607.4040202@sun.com> References: <47856607.4040202@sun.com> Message-ID: <47856D34.40301@joda.org> Debra Scott wrote: > I'd like to get people's thoughts on the documentation > for the JDK platform. > > * How many people feel it is important for the the docs to be open > sourced so the community can contribute to them? Are you referring to javadoc, or other docs? > * How many people have seen at least one occasion > where they would have added something to the > documentation, or made a correction, if they > had been able? (a large or small percentage?) I have had plenty of simple javadoc fixes I've seen and might have made (like documenting what happens with null). But I suspect that is not what you are driving at. Stephen From amitksaha at netbeans.org Wed Jan 9 21:42:27 2008 From: amitksaha at netbeans.org (Amit Kumar Saha) Date: Thu, 10 Jan 2008 11:12:27 +0530 Subject: JDK documentation In-Reply-To: <547db2260801092141y57e0135fga204ea1706365144@mail.gmail.com> References: <47856607.4040202@sun.com> <547db2260801092141y57e0135fga204ea1706365144@mail.gmail.com> Message-ID: <547db2260801092142m4e7c6c23jf9410e66d2821af2@mail.gmail.com> On 1/10/08, Debra Scott wrote: > Hi All, > > I'd like to get people's thoughts on the documentation > for the JDK platform. > > * How many people feel it is important for the the docs to be open > sourced so the community can contribute to them? How about something which is modeled on the "NetBeans Community Docs" (http://wiki.netbeans.org/wiki/view/CommunityDocs)? > > > * How many people have seen at least one occasion > where they would have added something to the > documentation, or made a correction, if they > had been able? (a large or small percentage?) > > * Given that we need documentation in a structured > format that allows for content sharing and multiple > delivery vehicles, does anyone have any recommendations > for a system that is both easily editable, like a > Wiki, and easily repurposable, like XML-structured > docs? How about using ODT? Just my thoughts, opinions! Thanks, Amit -- Amit Kumar Saha *NetBeans Community Docs Coordinator* Writer, Programmer, Researcher http://amitsaha.in.googlepages.com http://amitksaha.blogspot.com From mthornton at optrak.co.uk Thu Jan 10 01:04:35 2008 From: mthornton at optrak.co.uk (Mark Thornton) Date: Thu, 10 Jan 2008 09:04:35 +0000 Subject: JDK documentation In-Reply-To: <47856607.4040202@sun.com> References: <47856607.4040202@sun.com> Message-ID: <4785DFA3.2070809@optrak.co.uk> Debra Scott wrote: > Hi All, > > I'd like to get people's thoughts on the documentation > for the JDK platform. > > * How many people feel it is important for the the docs to be open > sourced so the community can contribute to them? I think this would be very valuable. > > > * How many people have seen at least one occasion > where they would have added something to the > documentation, or made a correction, if they > had been able? (a large or small percentage?) There are numerous changes many small, some more substantial that I might have made. As it is some exist as comments to bugs on the bug parade (and have been referenced by the evaluation). A particular flaw in the current documentation is the poor (often non-existent) documenting of platform specific behaviour. Some people seem very clear that this information should go somewhere else, without doing anything to ensure that a suitable other place is created and populated. Mark Thornton From david.gilbert at object-refinery.com Thu Jan 10 02:00:08 2008 From: david.gilbert at object-refinery.com (David Gilbert) Date: Thu, 10 Jan 2008 10:00:08 +0000 Subject: JDK documentation In-Reply-To: <47856607.4040202@sun.com> References: <47856607.4040202@sun.com> Message-ID: <4785ECA8.1030201@object-refinery.com> Hi, Debra Scott wrote: > Hi All, > > I'd like to get people's thoughts on the documentation > for the JDK platform. > > * How many people feel it is important for the the docs to be open > sourced so the community can contribute to them? Very important. > > * How many people have seen at least one occasion > where they would have added something to the > documentation, or made a correction, if they > had been able? (a large or small percentage?) In the API docs, I've often spotted corner cases that aren't documented. In the past, I've documented these in the GNU Classpath API. I should probably start submitting patches to OpenJDK for these, but I wonder if that's very efficient because the changes are small and the process is not so lightweight as we had for GNU Classpath (that's not a criticism, it's just an observation, because I understand the need for greater rigor in the OpenJDK process). > * Given that we need documentation in a structured > format that allows for content sharing and multiple > delivery vehicles, does anyone have any recommendations > for a system that is both easily editable, like a > Wiki, and easily repurposable, like XML-structured > docs? A few years ago, the guys at JavaLobby started an excellent initiative called JDocs. They loaded up the Javadocs for the JDK and invited the community to annotate them with useful pointers, sample code etc. Sun (or its lawyers) unfortunately squashed the initiative and it more or less died (JDocs is still there, but the project seemed to lose momentum after the Java SE APIs were removed). Maybe you could get in touch with Rick Ross and Matthew Schmidt at Javalobby and ask them about it...an perhaps inject some new life into JDocs. Regards, Dave Gilbert http://www.jfree.org/ From Debra.Scott at Sun.COM Thu Jan 10 02:53:57 2008 From: Debra.Scott at Sun.COM (Debra Scott) Date: Thu, 10 Jan 2008 04:53:57 -0600 Subject: JDK documentation In-Reply-To: <47856D34.40301@joda.org> References: <47856607.4040202@sun.com> <47856D34.40301@joda.org> Message-ID: <4785F945.1080604@sun.com> As javadoc is already open source, I'm not talking about those docs (which are fixed through the source code files) I'm talking about the JDK guide documentation that is part of the JDK documentation download bundle -Debbie Stephen Colebourne wrote: > Debra Scott wrote: > >> I'd like to get people's thoughts on the documentation >> for the JDK platform. >> >> * How many people feel it is important for the the docs to be open >> sourced so the community can contribute to them? > > > Are you referring to javadoc, or other docs? > >> * How many people have seen at least one occasion >> where they would have added something to the >> documentation, or made a correction, if they >> had been able? (a large or small percentage?) > > > I have had plenty of simple javadoc fixes I've seen and might have made > (like documenting what happens with null). But I suspect that is not > what you are driving at. > > Stephen -- ========================================================================== Debra Scott 39 Golden Wheat Ln Manager of Java SE Documentation Wrightstown, WI 54180 Information Products Group Phone: 877-219-2362 x51743 Global Product Development & Operations, Sun Microsystems, Inc ========================================================================== From David.Holmes at Sun.COM Thu Jan 10 03:15:45 2008 From: David.Holmes at Sun.COM (David Holmes - Sun Microsystems) Date: Thu, 10 Jan 2008 21:15:45 +1000 Subject: JDK documentation In-Reply-To: <4785F945.1080604@sun.com> References: <47856607.4040202@sun.com> <47856D34.40301@joda.org> <4785F945.1080604@sun.com> Message-ID: <4785FE61.2050904@sun.com> Debra, Thanks for clarifying that it was the guides you meant - I totally agree they should be opened up and expanded/improved by the community. But I disagree with what you say about the javadocs. While the javadoc are part of the source files, they form the specification for the platform API's and as far as I am aware the specification for the Java platform is not open-sourced. So any "fixes" to the javadocs would not, I believe, be acceptable through OpenJDK contributions, unless done as part of a JSR. Hopefully Mark, or someone else in the know, could clarify this. I know I've been frustrated over the years by the apparent inability to get anything but the most trivial typos fixed in the docs, except during major releases. It would be nice if that could change but I'm not aware that it has at this stage. Cheers, David Holmes Debra Scott said the following on 10/01/08 08:53 PM: > As javadoc is already open source, I'm not talking about those docs > (which are fixed through the source code files) > > I'm talking about the JDK guide documentation that is part > of the JDK documentation download bundle > > -Debbie > > Stephen Colebourne wrote: >> Debra Scott wrote: >> >>> I'd like to get people's thoughts on the documentation >>> for the JDK platform. >>> >>> * How many people feel it is important for the the docs to be open >>> sourced so the community can contribute to them? >> >> >> Are you referring to javadoc, or other docs? >> >>> * How many people have seen at least one occasion >>> where they would have added something to the >>> documentation, or made a correction, if they >>> had been able? (a large or small percentage?) >> >> >> I have had plenty of simple javadoc fixes I've seen and might have >> made (like documenting what happens with null). But I suspect that is >> not what you are driving at. >> >> Stephen > From Debra.Scott at Sun.COM Thu Jan 10 03:26:42 2008 From: Debra.Scott at Sun.COM (Debra Scott) Date: Thu, 10 Jan 2008 05:26:42 -0600 Subject: JDK documentation In-Reply-To: <4785ECA8.1030201@object-refinery.com> References: <47856607.4040202@sun.com> <4785ECA8.1030201@object-refinery.com> Message-ID: <478600F2.8000400@sun.com> All, I apologize for not being clear on my original post. The JDK documentation, that provided in the zip file we refer to as the JDK documentation bundle, includes: * API docs (javadocs, which are already open source as they are derived from the .java source files) -- I'm not referring to these as you already have access to changing these as part of source code contributions/fixes * guides documentation (for example, see http://java.sun.com/javase/6/docs/technotes/guides/) * tools (for example, see http://java.sun.com/javase/6/docs/technotes/tools) Do most of you only use the API docs and are these the docs you think of? If so, maybe it isn't worth the effort to open source the rest of our documentation? -Debbie David Gilbert wrote: > Hi, > > Debra Scott wrote: > >> Hi All, >> >> I'd like to get people's thoughts on the documentation >> for the JDK platform. >> >> * How many people feel it is important for the the docs to be open >> sourced so the community can contribute to them? > > Very important. > >> >> * How many people have seen at least one occasion >> where they would have added something to the >> documentation, or made a correction, if they >> had been able? (a large or small percentage?) > > In the API docs, I've often spotted corner cases that aren't > documented. In the past, I've documented these in the GNU Classpath > API. I should probably start submitting patches to OpenJDK for these, > but I wonder if that's very efficient because the changes are small and > the process is not so lightweight as we had for GNU Classpath (that's > not a criticism, it's just an observation, because I understand the need > for greater rigor in the OpenJDK process). > >> * Given that we need documentation in a structured >> format that allows for content sharing and multiple >> delivery vehicles, does anyone have any recommendations >> for a system that is both easily editable, like a >> Wiki, and easily repurposable, like XML-structured >> docs? > > A few years ago, the guys at JavaLobby started an excellent initiative > called JDocs. They loaded up the Javadocs for the JDK and invited the > community to annotate them with useful pointers, sample code etc. Sun > (or its lawyers) unfortunately squashed the initiative and it more or > less died (JDocs is still there, but the project seemed to lose momentum > after the Java SE APIs were removed). Maybe you could get in touch with > Rick Ross and Matthew Schmidt at Javalobby and ask them about it...an > perhaps inject some new life into JDocs. > > Regards, > > Dave Gilbert > http://www.jfree.org/ -- ========================================================================== Debra Scott 39 Golden Wheat Ln Manager of Java SE Documentation Wrightstown, WI 54180 Information Products Group Phone: 877-219-2362 x51743 Global Product Development & Operations, Sun Microsystems, Inc ========================================================================== From david.gilbert at object-refinery.com Thu Jan 10 03:29:07 2008 From: david.gilbert at object-refinery.com (David Gilbert) Date: Thu, 10 Jan 2008 11:29:07 +0000 Subject: JDK documentation In-Reply-To: <4785FE61.2050904@sun.com> References: <47856607.4040202@sun.com> <47856D34.40301@joda.org> <4785F945.1080604@sun.com> <4785FE61.2050904@sun.com> Message-ID: <47860183.7090603@object-refinery.com> David Holmes - Sun Microsystems wrote: > > But I disagree with what you say about the javadocs. While the javadoc > are part of the source files, they form the specification for the > platform API's and as far as I am aware the specification for the Java > platform is not open-sourced. So any "fixes" to the javadocs would > not, I believe, be acceptable through OpenJDK contributions, unless > done as part of a JSR. Hopefully Mark, or someone else in the know, > could clarify this. > > I know I've been frustrated over the years by the apparent inability > to get anything but the most trivial typos fixed in the docs, except > during major releases. It would be nice if that could change but I'm > not aware that it has at this stage. I really think there is a need for two versions of the Javadocs, one that is a specification (that is, what we have now) and another that is "developer documentation". The latter is, I believe, what the Javalobby guys were/are trying to do with JDocs.com. Promoting their existing effort seems worth exploring, in my opinion, especially since it can exist independently of the source code. Regards, Dave Gilbert http://www.jfree.org/ From mthornton at optrak.co.uk Thu Jan 10 03:34:55 2008 From: mthornton at optrak.co.uk (Mark Thornton) Date: Thu, 10 Jan 2008 11:34:55 +0000 Subject: JDK documentation In-Reply-To: <47860183.7090603@object-refinery.com> References: <47856607.4040202@sun.com> <47856D34.40301@joda.org> <4785F945.1080604@sun.com> <4785FE61.2050904@sun.com> <47860183.7090603@object-refinery.com> Message-ID: <478602DF.1010700@optrak.co.uk> David Gilbert wrote: > David Holmes - Sun Microsystems wrote: >> >> But I disagree with what you say about the javadocs. While the >> javadoc are part of the source files, they form the specification for >> the platform API's and as far as I am aware the specification for the >> Java platform is not open-sourced. So any "fixes" to the javadocs >> would not, I believe, be acceptable through OpenJDK contributions, >> unless done as part of a JSR. Hopefully Mark, or someone else in the >> know, could clarify this. >> >> I know I've been frustrated over the years by the apparent inability >> to get anything but the most trivial typos fixed in the docs, except >> during major releases. It would be nice if that could change but I'm >> not aware that it has at this stage. > I really think there is a need for two versions of the Javadocs, one > that is a specification (that is, what we have now) and another that > is "developer documentation". The latter is, I believe, what the > Javalobby guys were/are trying t