From mark at klomp.org Sat Oct 1 11:40:50 2005 From: mark at klomp.org (Mark Wielaard) Date: Sat, 01 Oct 2005 13:40:50 +0200 Subject: [fedora-java] DevJam reports Message-ID: <1128166851.5800.73.camel@localhost.localdomain> Hi all, The GNU Classpath distro DevJam was a great success. It seems we brought some harmony into the hearts and minds of the different distributions (Ubuntu, SkoleLinux, Debian, Fedora, Suse, Gentoo, OpenEmbedded) that participated. And being able to talk and debug some issues with several of the upstream projects involved (GNU Classpath, kaffe, gcj, Cacao) was definitely inspirational and productive. Here is a list of other summaries and notes of the meeting: - SkoleLinux summaries and pictures: http://skolelinux.de/wiki/FreeJava/Meeting050923 - OpenEmbedded ARM TODO list: http://www.informatik.uni-bremen.de/cgi-bin/cgiwrap/rwagner/pyblosxom.cgi/computers/freejava/gcj-on-arm.html - GCJ maintainer/Fedora impressions by Andrew Haley: http://www.advogato.org/person/aph/diary.html?start=0 - Gentoo DevJam braindump by Petteri R?ty (plus presentation) http://article.gmane.org/gmane.linux.gentoo.java/598 http://dev.gentoo.org/~betelgeuse/show.pdf - DevJam Arrival and Schedule/Discussion notes: http://gnu.wildebeest.org/diary/index.php?p=116 - Debian Project leader notes: http://necrotic.deadbeast.net/~branden/blog/exuberance/Debian/destination_oldenburg.html - LWN article about the meeting that is currently being published for subscribers (please support LWN it is a great magazine): http://lwn.net/Articles/153450/ Next week it will be free for all. (Please send me, or the devjam mailing-list, updates and additions.) On request of several of the participants I have setup a mailing-list so people can keep in touch and coordinate cross-distro/packaging/project things. If you are interested please send an email to devjam-subscribe at developer.classpath.org The mailing-list has a public archive accessible through: http://developer.classpath.org/mailman/listinfo/devjam And if you are interested in participating or helping out with a followup meeting please see the wiki about DevJam++: http://java.debian.net/index.php/DevJam++ Cheers, Mark -- Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html Join the community at http://planet.classpath.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From overholt at redhat.com Fri Oct 7 13:40:50 2005 From: overholt at redhat.com (Andrew Overholt) Date: Fri, 7 Oct 2005 09:40:50 -0400 Subject: [fedora-java] Native code not loaded when it is in the db Message-ID: <20051007134050.GA23465@redhat.com> Hi Andrew, I have natively-compiled all of Eclipse and created .db files and merged them all into the system-wide .db. I have verified that the .db file that is being used by gij (this is a gcc that I have built from 4.0 branch head) contains the .jar,.jar.so combinations that I want to use. However, two of the jars are being loaded as bytecode when I want them to be loaded from their corresponding .so (as evidenced by the following). eclipse -vmargs -verbose: [...] [Loaded (bytecode) org.eclipse.core.resources.ResourcesPlugin from (file:/home/andrew/work/eclipse/binary/eclipse/plugins/org.eclipse.core.resources_3.1.0.jar )] [...] I wrote a little test app (attached) to generate the byte strings from a jar just like they are presented by gcj-dbtool -l (*) so that I could verify that the actual classes I'm interested in are in fact in the .db. This is what I get: java MD5Test /home/andrew/work/eclipse/binary/eclipse/plugins/org.eclipse.core.resources_3.1.0.jar [...] org/eclipse/core/resources/ResourcesPlugin.class -> 617779565aed3728932e63a591ad9d [...] and gcj-dbtool -l `gcj-dbtool -p` [...] [21443] 617779565aed3728932e63a591ad9d -> /home/andrew/work/eclipse/native-3.1reallyclean/org.eclipse.core.resources_3.1.0.jar.so [...] (I have verified that my test program run on another jar that loads classes that are BC-compiled has matching signatures to those in the .db) Interestingly enough, when I try the same jar but with some modifications (I edited some classes in Eclipse, re-generated the jar, and natively-compiled it in the same way I did the above), its classes *are* loaded from the .so. This leads me to believe that something's either wrong with this jar/.jar.so or with the part of gij that intercepts class loading to feed stuff from .jar.sos in the .db. Any ideas? I guess Eclipse could be loading these two jars in unique ways. The second jar is SWT which I know has JNI .sos ripped out the first time it is used so maybe that is causing some weirdness there? Thanks, Andrew * it would be nice if gcj-dbtool could have this functionality but I guess then it'll need to store the original classname ... perhaps I'll write a patch -------------- next part -------------- import java.io.*; import java.nio.channels.*; import java.util.*; import java.util.jar.*; import java.security.MessageDigest; public class MD5Test { public static void main(String[] args) { try { MessageDigest md = MessageDigest.getInstance("MD5"); JarFile jar = new JarFile (args[0]); int count = 0; { Enumeration entries = jar.entries(); while (entries.hasMoreElements()) { JarEntry classfile = (JarEntry)entries.nextElement(); if (classfile.getName().endsWith(".class")) count++; } } Enumeration entries = jar.entries(); while (entries.hasMoreElements()) { JarEntry classfile = (JarEntry)entries.nextElement(); if (classfile.getName().endsWith(".class")) { InputStream str = jar.getInputStream(classfile); int length = (int) classfile.getSize(); if (length == -1) throw new EOFException(); byte[] data = new byte[length]; int pos = 0; while (length - pos > 0) { int len = str.read(data, pos, length - pos); if (len == -1) throw new EOFException("Not enough data reading from: " + classfile.getName()); pos += len; } System.out.println(classfile.getName() + " -> " + bytesToString(md.digest(data))); } } } catch (Exception e) { e.printStackTrace(); } } static String bytesToString(byte[] b) { StringBuffer hexBytes = new StringBuffer(); int length = b.length; for (int i = 0; i < length; ++i) hexBytes.append(Integer.toHexString(b[i] & 0xff)); return hexBytes.toString(); } } From aph at redhat.com Fri Oct 7 13:49:49 2005 From: aph at redhat.com (Andrew Haley) Date: Fri, 7 Oct 2005 14:49:49 +0100 Subject: [fedora-java] Re: Native code not loaded when it is in the db In-Reply-To: <20051007134050.GA23465@redhat.com> References: <20051007134050.GA23465@redhat.com> Message-ID: <17222.31997.60454.493339@zapata.pink> Andrew Overholt writes: > > I have natively-compiled all of Eclipse and created .db files and merged > them all into the system-wide .db. I have verified that the .db file that > is being used by gij (this is a gcc that I have built from 4.0 branch head) > contains the .jar,.jar.so combinations that I want to use. However, two of > the jars are being loaded as bytecode when I want them to be loaded from > their corresponding .so (as evidenced by the following). > > eclipse -vmargs -verbose: > > [...] > [Loaded (bytecode) org.eclipse.core.resources.ResourcesPlugin from (file:/home/andrew/work/eclipse/binary/eclipse/plugins/org.eclipse.core.resources_3.1.0.jar )] > [...] > > I wrote a little test app (attached) to generate the byte strings from a > jar just like they are presented by gcj-dbtool -l (*) so that I could > verify that the actual classes I'm interested in are in fact in the .db. > This is what I get: > > java MD5Test /home/andrew/work/eclipse/binary/eclipse/plugins/org.eclipse.core.resources_3.1.0.jar > [...] > org/eclipse/core/resources/ResourcesPlugin.class -> 617779565aed3728932e63a591ad9d > [...] > > and > > gcj-dbtool -l `gcj-dbtool -p` > [...] > [21443] 617779565aed3728932e63a591ad9d -> /home/andrew/work/eclipse/native-3.1reallyclean/org.eclipse.core.resources_3.1.0.jar.so > [...] > > (I have verified that my test program run on another jar that loads classes > that are BC-compiled has matching signatures to those in the .db) > > Interestingly enough, when I try the same jar but with some modifications > (I edited some classes in Eclipse, re-generated the jar, and > natively-compiled it in the same way I did the above), its classes *are* > loaded from the .so. This leads me to believe that something's either > wrong with this jar/.jar.so or with the part of gij that intercepts class > loading to feed stuff from .jar.sos in the .db. Any ideas? I suspect that there is something wrong with the .so. Perhaps, for example, it has some unresolved symdols. In that case we silently fall back to the .jars. It would be nice if verbose:class indicated that it had tried to load a class from a .so and failed. > I guess Eclipse could be loading these two jars in unique ways. > The second jar is SWT which I know has JNI .sos ripped out the > first time it is used so maybe that is causing some weirdness > there? > > Thanks, > > Andrew > > * it would be nice if gcj-dbtool could have this functionality but I guess > then it'll need to store the original classname ... perhaps I'll write a > patch Something like gcj-dbtool - foo.jar. Yes, that'd be nice. Andrew. From tromey at redhat.com Fri Oct 7 15:30:23 2005 From: tromey at redhat.com (Tom Tromey) Date: 07 Oct 2005 09:30:23 -0600 Subject: [fedora-java] Re: Native code not loaded when it is in the db In-Reply-To: <17222.31997.60454.493339@zapata.pink> References: <20051007134050.GA23465@redhat.com> <17222.31997.60454.493339@zapata.pink> Message-ID: >>>>> "Andrew" == Andrew Haley writes: Andrew> It would be nice if verbose:class indicated that it had tried to load Andrew> a class from a .so and failed. At the point where this is printed, it would be really hard to determine this. I was thinking of adding something like '-Xverbose:libraries' to print all .so activity, including error messages on failed loads. That would be useful in a number of situations. I'll file a PR for this. Tom From aph at redhat.com Fri Oct 7 16:09:52 2005 From: aph at redhat.com (Andrew Haley) Date: Fri, 7 Oct 2005 17:09:52 +0100 Subject: [fedora-java] Re: Native code not loaded when it is in the db In-Reply-To: References: <20051007134050.GA23465@redhat.com> <17222.31997.60454.493339@zapata.pink> Message-ID: <17222.40400.914591.762324@zapata.pink> Tom Tromey writes: > >>>>> "Andrew" == Andrew Haley writes: > > Andrew> It would be nice if verbose:class indicated that it had tried to load > Andrew> a class from a .so and failed. > > At the point where this is printed, it would be really hard to > determine this. Well, we always throw if we fail to load a .so. So all we have to do when -verbose: is turned on is print (" failed + " the Throwable) Andrew. From tromey at redhat.com Fri Oct 7 16:15:19 2005 From: tromey at redhat.com (Tom Tromey) Date: 07 Oct 2005 10:15:19 -0600 Subject: [fedora-java] Re: Native code not loaded when it is in the db In-Reply-To: <17222.40400.914591.762324@zapata.pink> References: <20051007134050.GA23465@redhat.com> <17222.31997.60454.493339@zapata.pink> <17222.40400.914591.762324@zapata.pink> Message-ID: >>>>> "Andrew" == Andrew Haley writes: Andrew> Well, we always throw if we fail to load a .so. So all we have to do Andrew> when -verbose: is turned on is print (" failed + " the Throwable) Yeah, for the .db path. We don't throw on failure in natRuntime.cc. Tom From green at redhat.com Tue Oct 11 14:24:09 2005 From: green at redhat.com (Anthony Green) Date: Tue, 11 Oct 2005 07:24:09 -0700 Subject: [fedora-java] FC5 Release Notes Message-ID: <1129040649.3243.89.camel@localhost.localdomain> I've volunteered to help with the FC5 release notes. According to the schedule, FC5 test1 is only 3 weeks away! To be honest, I'm not quite clear on what kinds of things the release notes should contain. I'll work with the docs team on this, but if anybody has ideas or knows of things that should be included, now would be a good time to mention them. AG From gbenson at redhat.com Tue Oct 11 14:45:22 2005 From: gbenson at redhat.com (Gary Benson) Date: Tue, 11 Oct 2005 15:45:22 +0100 Subject: [fedora-java] FC5 Release Notes In-Reply-To: <1129040649.3243.89.camel@localhost.localdomain> References: <1129040649.3243.89.camel@localhost.localdomain> Message-ID: <20051011144519.GB4914@redhat.com> Anthony Green wrote: > I've volunteered to help with the FC5 release notes. Anthony that is amazing! Thanks! > To be honest, I'm not quite clear on what kinds of things the > release notes should contain. I'll work with the docs team on > this, but if anybody has ideas or knows of things that should > be included, now would be a good time to mention them. I guess just look in old release notes and see what people have said. They've always been a wierd mix of really vague and really specific iirc. Cheers (very much so!), Gary From i.pilcher at comcast.net Tue Oct 11 14:44:53 2005 From: i.pilcher at comcast.net (Ian Pilcher) Date: Tue, 11 Oct 2005 09:44:53 -0500 Subject: [fedora-java] Re: FC5 Release Notes In-Reply-To: <1129040649.3243.89.camel@localhost.localdomain> References: <1129040649.3243.89.camel@localhost.localdomain> Message-ID: Anthony Green wrote: > I've volunteered to help with the FC5 release notes. > > According to the schedule, FC5 test1 is only 3 weeks away! > > To be honest, I'm not quite clear on what kinds of things the release > notes should contain. I'll work with the docs team on this, but if > anybody has ideas or knows of things that should be included, now would > be a good time to mention them. > Fedora Core/JPackage co-existence seems like an obvious candidate. -- ======================================================================== Ian Pilcher i.pilcher at comcast.net ======================================================================== From green at redhat.com Tue Oct 11 15:37:25 2005 From: green at redhat.com (Anthony Green) Date: Tue, 11 Oct 2005 08:37:25 -0700 Subject: [fedora-java] Re: FC5 Release Notes In-Reply-To: References: <1129040649.3243.89.camel@localhost.localdomain> Message-ID: <1129045045.3243.114.camel@localhost.localdomain> On Tue, 2005-10-11 at 09:44 -0500, Ian Pilcher wrote: > > Fedora Core/JPackage co-existence seems like an obvious candidate. Yes, and Sun/gcj co-existence via JPackage and alternatives in particular. FC5 will include JOnAS as well, right? Are there any other new additions in java-land this time 'round? AG From gbenson at redhat.com Tue Oct 11 15:52:54 2005 From: gbenson at redhat.com (Gary Benson) Date: Tue, 11 Oct 2005 16:52:54 +0100 Subject: [fedora-java] Re: FC5 Release Notes In-Reply-To: <1129045045.3243.114.camel@localhost.localdomain> References: <1129040649.3243.89.camel@localhost.localdomain> <1129045045.3243.114.camel@localhost.localdomain> Message-ID: <20051011155251.GC4914@redhat.com> Anthony Green wrote: > On Tue, 2005-10-11 at 09:44 -0500, Ian Pilcher wrote: > > Fedora Core/JPackage co-existence seems like an obvious candidate. > > Yes, and Sun/gcj co-existence via JPackage and alternatives in > particular. I'd rather say "Other runtimes" rather than naming names... > FC5 will include JOnAS as well, right? Are there any other > new additions in java-land this time 'round? Yes (well, probably) and no (not that I know of). Actually, maybe there are some new Eclipse bits. Cheers, Gary From overholt at redhat.com Tue Oct 11 15:58:22 2005 From: overholt at redhat.com (Andrew Overholt) Date: Tue, 11 Oct 2005 11:58:22 -0400 Subject: [fedora-java] Re: FC5 Release Notes In-Reply-To: <20051011155251.GC4914@redhat.com> References: <1129040649.3243.89.camel@localhost.localdomain> <1129045045.3243.114.camel@localhost.localdomain> <20051011155251.GC4914@redhat.com> Message-ID: <20051011155821.GA3960@redhat.com> * Gary Benson [2005-10-11 11:53]: > Anthony Green wrote: > > FC5 will include JOnAS as well, right? Are there any other > > new additions in java-land this time 'round? > > Yes (well, probably) and no (not that I know of). > Actually, maybe there are some new Eclipse bits. I've just gotten back into Fedora stuff in the past few weeks. I'm still unsure as to what we'll have in FC/FE5. Andrew From justin.conover at gmail.com Tue Oct 11 18:32:54 2005 From: justin.conover at gmail.com (Justin Conover) Date: Tue, 11 Oct 2005 13:32:54 -0500 Subject: [fedora-java] a java beagle version? Message-ID: Anyone thought about converting beagle(mono) to java, is it possible with some changes to the code? I saw something months back about the diffs between java/mono and how it is possible to write apps for both without much pain, I'm not a developer and don't really know enough about java atm to say yes/no ;-) Just thought if it is possible, that would be a good project since fedora/rh doesn't want to go down the mono route for what i think are good reasons. For example, didn't Wikipedia mv from GCJ and Lucene to Mono and Dotlucene? I don't know if there is any similarity between DotLucene and Beagle, but if Lucene does some kind of indexing/searching would it be possible to write some kind of Lucene interface to index dir on a desktop? I could just be pulling all of this out of my @$$ and be a bad idea ;) -------------- next part -------------- An HTML attachment was scrubbed... URL: From brion at pobox.com Tue Oct 11 21:34:54 2005 From: brion at pobox.com (Brion Vibber) Date: Tue, 11 Oct 2005 14:34:54 -0700 Subject: [fedora-java] a java beagle version? In-Reply-To: References: Message-ID: <434C2FFE.9090004@pobox.com> Justin Conover wrote: > Anyone thought about converting beagle(mono) to java, is it possible > with some changes to the code? I saw something months back about the > diffs between java/mono and how it is possible to write apps for both > without much pain, I'm not a developer and don't really know enough > about java atm to say yes/no ;-) > > Just thought if it is possible, that would be a good project since > fedora/rh doesn't want to go down the mono route for what i think are > good reasons. > > For example, didn't Wikipedia mv from GCJ and Lucene to Mono and > Dotlucene? I don't know if there is any similarity between DotLucene > and Beagle, but if Lucene does some kind of indexing/searching would it > be possible to write some kind of Lucene interface to index dir on a > desktop? DotLucene is a C# port of Java Lucene; the API is the same (modulo the silly .NET case conventions). Lucene itself is an indexing and retrieval engine, basically a database targetted at text searching. Wikipedia's search engine is a relatively small networked daemon with a bit of wrapper code, so it was easy to move from Java to C# and would be easy to move back. (My biggest problem with GCJ, and the reason I switched that app to C#/Mono, has been memory leaks. Possibly this is from heap fragmentation, I hope to find the time to investigate it someday...) Beagle uses DotLucene too, but there's a lot more code around it, including all the indexing, file monitoring, file format readers, user interface, desktop integration... I'm sure it could be ported to Java, but it seems like it would be a lot of wasted effort just because some folks at Red Hat don't want to ship free software from Novell. ;) -- brion vibber (brion @ pobox.com) -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 253 bytes Desc: OpenPGP digital signature URL: From justin.conover at gmail.com Tue Oct 11 23:49:56 2005 From: justin.conover at gmail.com (Justin Conover) Date: Tue, 11 Oct 2005 18:49:56 -0500 Subject: [fedora-java] a java beagle version? In-Reply-To: <434C2FFE.9090004@pobox.com> References: <434C2FFE.9090004@pobox.com> Message-ID: On 10/11/05, Brion Vibber wrote: > > Justin Conover wrote: > > Anyone thought about converting beagle(mono) to java, is it possible > > with some changes to the code? I saw something months back about the > > diffs between java/mono and how it is possible to write apps for both > > without much pain, I'm not a developer and don't really know enough > > about java atm to say yes/no ;-) > > > > Just thought if it is possible, that would be a good project since > > fedora/rh doesn't want to go down the mono route for what i think are > > good reasons. > > > > For example, didn't Wikipedia mv from GCJ and Lucene to Mono and > > Dotlucene? I don't know if there is any similarity between DotLucene > > and Beagle, but if Lucene does some kind of indexing/searching would it > > be possible to write some kind of Lucene interface to index dir on a > > desktop? > > DotLucene is a C# port of Java Lucene; the API is the same (modulo the > silly .NET case conventions). Lucene itself is an indexing and retrieval > engine, basically a database targetted at text searching. > > Wikipedia's search engine is a relatively small networked daemon with a > bit of wrapper code, so it was easy to move from Java to C# and would be > easy to move back. (My biggest problem with GCJ, and the reason I > switched that app to C#/Mono, has been memory leaks. Possibly this is > from heap fragmentation, I hope to find the time to investigate it > someday...) > > Beagle uses DotLucene too, but there's a lot more code around it, > including all the indexing, file monitoring, file format readers, user > interface, desktop integration... I'm sure it could be ported to Java, > but it seems like it would be a lot of wasted effort just because some > folks at Red Hat don't want to ship free software from Novell. ;) > > -- brion vibber (brion @ pobox.com ) As far as wasted effort, how many of the mono apps are duplicates of software that is already out there? -------------- next part -------------- An HTML attachment was scrubbed... URL: From aph at redhat.com Wed Oct 12 08:49:02 2005 From: aph at redhat.com (Andrew Haley) Date: Wed, 12 Oct 2005 09:49:02 +0100 Subject: [fedora-java] a java beagle version? In-Reply-To: <434C2FFE.9090004@pobox.com> References: <434C2FFE.9090004@pobox.com> Message-ID: <17228.52734.248869.400783@zapata.pink> Brion Vibber writes: > Justin Conover wrote: > > Anyone thought about converting beagle(mono) to java, is it possible > > with some changes to the code? I saw something months back about the > > diffs between java/mono and how it is possible to write apps for both > > without much pain, I'm not a developer and don't really know enough > > about java atm to say yes/no ;-) > > > > Just thought if it is possible, that would be a good project since > > fedora/rh doesn't want to go down the mono route for what i think are > > good reasons. > > > > For example, didn't Wikipedia mv from GCJ and Lucene to Mono and > > Dotlucene? I don't know if there is any similarity between DotLucene > > and Beagle, but if Lucene does some kind of indexing/searching would it > > be possible to write some kind of Lucene interface to index dir on a > > desktop? > > DotLucene is a C# port of Java Lucene; the API is the same (modulo the > silly .NET case conventions). Lucene itself is an indexing and retrieval > engine, basically a database targetted at text searching. > > Wikipedia's search engine is a relatively small networked daemon with a > bit of wrapper code, so it was easy to move from Java to C# and would be > easy to move back. (My biggest problem with GCJ, and the reason I > switched that app to C#/Mono, has been memory leaks. Possibly this is > from heap fragmentation, I hope to find the time to investigate it > someday...) So do I. I would very much like to get to the bottom of this. Andrew. From aph at redhat.com Wed Oct 12 09:00:18 2005 From: aph at redhat.com (Andrew Haley) Date: Wed, 12 Oct 2005 10:00:18 +0100 Subject: [fedora-java] a java beagle version? In-Reply-To: References: <434C2FFE.9090004@pobox.com> Message-ID: <17228.53410.681459.369053@zapata.pink> Justin Conover writes: > On 10/11/05, Brion Vibber wrote: > > > > Beagle uses DotLucene too, but there's a lot more code around it, > > including all the indexing, file monitoring, file format readers, user > > interface, desktop integration... I'm sure it could be ported to Java, > > but it seems like it would be a lot of wasted effort just because some > > folks at Red Hat don't want to ship free software from Novell. ;) > > As far as wasted effort, how many of the mono apps are duplicates of > software that is already out there? Can we please stop this thread now? I'm staring to get a bad feeling about where this might lead. I mean, I like a nice flame war as much as the next man, but, but ... :-) Seriously, let's concentrate on Java development in Fedora. Andrew. From aph at redhat.com Wed Oct 12 10:14:04 2005 From: aph at redhat.com (Andrew Haley) Date: Wed, 12 Oct 2005 11:14:04 +0100 Subject: [fedora-java] Install failure in fedora In-Reply-To: <17228.53410.681459.369053@zapata.pink> References: <434C2FFE.9090004@pobox.com> <17228.53410.681459.369053@zapata.pink> Message-ID: <17228.57836.757096.454408@zapata.pink> yum update gcc ... libgcj-devel-4.0.1-14.i38 100% |=========================| 294 kB 00:01 ---> Package libgcj-devel.i386 0:4.0.1-14 set to be updated --> Running transaction check --> Processing Dependency: libcairo.so.2 for package: libgcj --> Processing Dependency: libpangocairo-1.0.so.0 for package: libgcj --> Processing Dependency: libgcj = 4.0.1-13 for package: libgcj-src --> Restarting Dependency Resolution with new changes. --> Populating transaction set with selected packages. Please wait. ---> Downloading header for libgcj-src to pack into transaction set. libgcj-src-4.0.1-14.i386. 100% |=========================| 25 kB 00:00 ---> Package libgcj-src.i386 0:4.0.1-14 set to be updated --> Running transaction check --> Processing Dependency: libcairo.so.2 for package: libgcj --> Processing Dependency: libpangocairo-1.0.so.0 for package: libgcj --> Finished Dependency Resolution Error: Missing Dependency: libcairo.so.2 is needed by package libgcj Error: Missing Dependency: libpangocairo-1.0.so.0 is needed by package libgcj Looks like yum has no idea where to get libpangocairo-1.0.so.0 from. Andrew. From aph at redhat.com Wed Oct 12 10:27:56 2005 From: aph at redhat.com (Andrew Haley) Date: Wed, 12 Oct 2005 11:27:56 +0100 Subject: [fedora-java] Re: Install failure in fedora In-Reply-To: <17228.57836.757096.454408@zapata.pink> References: <434C2FFE.9090004@pobox.com> <17228.53410.681459.369053@zapata.pink> <17228.57836.757096.454408@zapata.pink> Message-ID: <17228.58668.74029.892282@zapata.pink> Forget that: bad local yum config. :-( Andrew. From tromey at redhat.com Wed Oct 12 13:50:53 2005 From: tromey at redhat.com (Tom Tromey) Date: 12 Oct 2005 07:50:53 -0600 Subject: [fedora-java] a java beagle version? In-Reply-To: References: Message-ID: >>>>> "Justin" == Justin Conover writes: Justin> Anyone thought about converting beagle(mono) to java, is it Justin> possible with some changes to the code? There are a couple projects out there for desktop-search-in-java. One is 'deskcrawler' (uses java-gnome), another is 'regain' (uses swing). Both, unfortunately, are fairly immature. Looks like a fun thing to hack on :-) Justin> For example, didn't Wikipedia mv from GCJ and Lucene to Mono Justin> and Dotlucene? Yeah, we didn't really have the available bandwidth to figure out what was going wrong for gcj in this case :-(. We need some kind of adopt-an-application program :-) Tom From overholt at redhat.com Wed Oct 12 19:43:28 2005 From: overholt at redhat.com (Andrew Overholt) Date: Wed, 12 Oct 2005 15:43:28 -0400 Subject: [fedora-java] Eclipse 3.1.1 Message-ID: <20051012194327.GE21851@redhat.com> Hi, I got Eclipse 3.1.1 built into rawhide last night and I was able to use it to check out GNU Classpath [1] and build everything (using Tom Tromey's builders for the native part and the locale creation, etc.)! This was the first time I've been able to do this with our Eclipse. There is one small wrinkle, though: I forgot to apply one patch to the build [1] so you'll get lots of OutOfMemory errors. I've applied the patch in CVS now and the build should sync up to mirrors tomorrow. If you are absolutely dying to try it today, you can do the following: # First, grab what's in rawhide ... eclipse-jdt can be eclipse-pde, etc. sudo yum --enablerepo=development install eclipse-jdt # Then, grab the patched org.eclipse.jdt.core_3.1.1.jar{,.so{,.db}} wget http://overholt.ca/org.eclipse.jdt.core_3.1.1.jar wget http://overholt.ca/org.eclipse.jdt.core_3.1.1.jar.so wget http://overholt.ca/org.eclipse.jdt.core_3.1.1.jar.db sudo mv /usr/share/eclipse/plugins/org.eclipse.jdt.core_3.1.1.jar{,.bak} sudo mv /usr/lib/gcj/eclipse/org.eclipse.jdt.core_3.1.1.jar.so{,.bak} sudo mv /usr/lib/gcj/eclipse/org.eclipse.jdt.core_3.1.1.jar.db{,.bak} sudo cp org.eclipse.jdt.core_3.1.1.jar /usr/share/eclipse/plugins sudo cp org.eclipse.jdt.core_3.1.1.jar.so /usr/lib/gcj/eclipse sudo cp org.eclipse.jdt.core_3.1.1.jar.db /usr/lib/gcj/eclipse sudo rebuild-gcj-db You should now be able to start up Eclipse. Be careful with workspaces, though, because I read that workspaces that have been "3.1.1-ified" may (will?) not work with older releases. Also, as with most upgrades, if you experience problems, try mv ~/.eclipse{,.bak} (unless you've put plug-ins in there or something you want to keep using). As usual, please report bugs at: https://bugzilla.redhat.com/bugzilla/enter_bug.cgi?product=Fedora%20Core&version=devel&component=eclipse Thanks, Andrew [1] I'm still having issues with varying levels of CVS compression. It seems to be most reliable when checked out with the default level of 0. Yes, this is slow, but I'd rather it actually finished. [2] https://bugs.eclipse.org/bugs/show_bug.cgi?id=111299 patch: http://dev.eclipse.org/viewcvs/index.cgi/org.eclipse.jdt.core/model/org/eclipse/jdt/internal/core/JavaModelCache.java.diff?r2=1.20.4.2&r1=1.20.4.1&diff_format=u From langel at redhat.com Wed Oct 12 20:36:48 2005 From: langel at redhat.com (Lillian Angel) Date: Wed, 12 Oct 2005 16:36:48 -0400 Subject: [fedora-java] Fedora Directory Server Application Message-ID: <1129149408.2953.105.camel@tow.toronto.redhat.com> Hello, I have created a patch for the Netscape/Fedora Directory Server Application. When applied to the console directory, the application builds successfully against Classpath and the JDK. Also, the file console.src.com.netscape.client.comm.Hanger.java should be removed from the cvs repository. It is not needed. The other changes are attached as a patch. Thanks, Lillian Angel -------------- next part -------------- A non-text attachment was scrubbed... Name: Netscape_Patch.diff Type: text/x-patch Size: 4362 bytes Desc: not available URL: From langel at redhat.com Wed Oct 12 20:48:45 2005 From: langel at redhat.com (Lillian Angel) Date: Wed, 12 Oct 2005 16:48:45 -0400 Subject: [fedora-java] Fedora Directory Server Application In-Reply-To: <1129149408.2953.105.camel@tow.toronto.redhat.com> References: <1129149408.2953.105.camel@tow.toronto.redhat.com> Message-ID: <1129150125.2953.110.camel@tow.toronto.redhat.com> > Also, the file console.src.com.netscape.client.comm.Hanger.java should > be removed from the cvs repository. It is not needed. The other changes > are attached as a patch. I should have mentioned why its not needed. Its a problem for free VMs since we don't implement sun.net.www.protocol.http*. I found by removing it, it compiles fine with the JDK and Classpath. Lillian From langel at redhat.com Wed Oct 12 20:57:53 2005 From: langel at redhat.com (Lillian Angel) Date: Wed, 12 Oct 2005 16:57:53 -0400 Subject: [fedora-java] Fedora Directory Server Application In-Reply-To: <1129149408.2953.105.camel@tow.toronto.redhat.com> References: <1129149408.2953.105.camel@tow.toronto.redhat.com> Message-ID: <1129150673.2953.125.camel@tow.toronto.redhat.com> On Wed, 2005-10-12 at 16:36 -0400, Lillian Angel wrote: > Hello, > > I have created a patch for the Netscape/Fedora Directory Server > Application. When applied to the console directory, the application > builds successfully against Classpath and the JDK. > > Also, the file console.src.com.netscape.client.comm.Hanger.java should > be removed from the cvs repository. > My mistake, I typed the filename wrong. Should be: console/src/com/netscape/client/comm/Handler.java Lillian From jdesbonnet at gmail.com Wed Oct 12 14:31:26 2005 From: jdesbonnet at gmail.com (Joe Desbonnet) Date: Wed, 12 Oct 2005 15:31:26 +0100 Subject: [fedora-java] a java beagle version? In-Reply-To: References: Message-ID: <1cef3e950510120731p5aed0958h603b88a6ac30cbe8@mail.gmail.com> As a matter of curiosity why not use the Sun JVM? Wikipedia does not need distribute software, so it should do as a stop gap measure until 100% free Java alternatives are mature enough. I find the Sun JVM a very robust platform for developing applictions. Joe. On 12 Oct 2005 07:50:53 -0600, Tom Tromey wrote: > Justin> For example, didn't Wikipedia mv from GCJ and Lucene to Mono > Justin> and Dotlucene? > > Yeah, we didn't really have the available bandwidth to figure out what was > going wrong for gcj in this case :-(. We need some kind of > adopt-an-application program :-) > > Tom > > -- From overholt at redhat.com Thu Oct 13 18:44:20 2005 From: overholt at redhat.com (Andrew Overholt) Date: Thu, 13 Oct 2005 14:44:20 -0400 Subject: [fedora-java] Eclipse 3.1.1 In-Reply-To: <20051012194327.GE21851@redhat.com> References: <20051012194327.GE21851@redhat.com> Message-ID: <20051013184420.GA31330@redhat.com> * Andrew Overholt [2005-10-12 15:43]: > > I got Eclipse 3.1.1 built into rawhide last night and I was able to use it > to check out GNU Classpath [1] and build everything (using Tom Tromey's > builders for the native part and the locale creation, etc.)! This was the > first time I've been able to do this with our Eclipse. It appears that the CVS issue [1] is not fixed as I thought it was. I've attached some instructions of how to get to the point I am at for debugging. I've also attached the little scripts that I am using to make the process easier. Andrew [1] https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161483 -------------- next part -------------- A non-text attachment was scrubbed... Name: nativifyjar.sh Type: application/x-sh Size: 580 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: all.sh Type: application/x-sh Size: 516 bytes Desc: not available URL: -------------- next part -------------- How to set things up to figure out WTF is going on with CVS checkouts in Eclipse with libgcj -- Andrew Overholt Last modified: 2005-10-13 . install Eclipse 3.1.1 from rawhide yum --enablerepo=development install eclipse-pde-devel . start it using the Sun (or other proprietary) JVM mkdir -p ~/workspaces eclipse -vm /bin/java -data ~/workspaces/cvs3.1.1 . import the source for the plug-ins we need: Import->External Plug-ins and Fragments . check "Projects with source folders" . next . pick org.eclipse.team.cvs.core, hit "Add ->" . hit "Required Plug-ins ->" . finish . create ant build files for each foreach(org.eclipse.ant.core, org.eclipse.core.resources, org.eclipse.core.runtime, org.eclipse.core.variables, org.eclipse.team.core, org.eclipse.team.cvs.core, org.eclipse.update.configurator): . right-click on plugin.xml . PDE Tools->Create Ant Build File . right-click on build.xml . Run As -> Ant Build... . check "build.update.jar", "build.jars" (should be checked), "clean", "refresh" . change the order to be: clean, build.jars, build.update.jar, refresh . hit Apply and then Run . in another terminal, create a directory to store the native bits of the above jars mkdir -p ~/eclipse/cvsissues/nativebits . get scripts to make nativifying easy (attached) cd ~/eclipse/cvsissues/nativebits wget http://overholt.ca/nativifyjar.sh wget http://overholt.ca/all.sh . make the jar and db files for the above jars point to your modified ones: pushd /usr/lib/gcj/eclipse sudo mv org.eclipse.ant.core_3.1.1.jar.db{,.bak} sudo mv org.eclipse.core.resources_3.1.0.jar.db{,.bak} sudo mv org.eclipse.core.runtime_3.1.1.jar.db{,.bak} sudo mv org.eclipse.core.variables_3.1.0.jar.db{,.bak} sudo mv org.eclipse.team.core_3.1.1.jar.db{,.bak} sudo mv org.eclipse.team.cvs.core_3.1.1.jar.db{,.bak} sudo mv org.eclipse.update.configurator_3.1.0.jar.db{,.bak} sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.ant.core_3.1.1.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.core.resources_3.1.0.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.core.runtime_3.1.1.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.core.variables_3.1.0.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.team.core_3.1.1.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.team.cvs.core_3.1.1.jar.db sudo ln -s ~/eclipse/cvsissues/nativebits/org.eclipse.update.configurator_3.1.0.jar.db popd pushd /usr/share/eclipse/plugins sudo mv org.eclipse.ant.core_3.1.1.jar /tmp sudo mv org.eclipse.core.resources_3.1.0.jar /tmp sudo mv org.eclipse.core.runtime_3.1.1.jar /tmp sudo mv org.eclipse.core.variables_3.1.0.jar /tmp sudo mv org.eclipse.team.core_3.1.1.jar /tmp sudo mv org.eclipse.team.cvs.core_3.1.1.jar /tmp sudo mv org.eclipse.update.configurator_3.1.0.jar /tmp sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.ant.core/org.eclipse.ant.core_3.1.1.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.core.resources/org.eclipse.core.resources_3.1.0.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.core.runtime/org.eclipse.core.runtime_3.1.1.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.core.variables/org.eclipse.core.variables_3.1.0.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.team.core/org.eclipse.team.core_3.1.1.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.team.cvs.core/org.eclipse.team.cvs.core_3.1.1.jar sudo ln -s ~/workspaces/cvs3.1.1/org.eclipse.update.configurator/org.eclipse.update.configurator_3.1.0.jar popd . make changes in Eclipse as you see fit, re-run the Ant build for the encompassing plug-in each time (right-click on build.xml -> Run As -> Ant Build), re-run the native part for either all jars (all.sh) or just the one you changed: either ./all.sh sudo rebuild-gcj-db OR one or more of: ./nativifyjar.sh org.eclipse.ant.core 3.1.1 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.core.resources 3.1.0 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.core.runtime 3.1.1 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.core.variables 3.1.0 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.team.core 3.1.1 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.team.cvs.core 3.1.1 ~/workspaces/cvs3.1.1 ./nativifyjar.sh org.eclipse.update.configurator 3.1.0 ~/workspaces/cvs3.1.1 followed by: sudo rebuild-gcj-db . enable tracing if you want by editing /.options and then tacking a -debug ~/workspaces/cvs3.1.1//.options onto the end of the eclipse launch command . run native eclipse: rm -rf ~/workspaces/testCVS eclipse -data ~/workspaces/testCVS [-debug ~/workspaces/cvs3.1.1/.options] . NOTE: you should also be able to run it from within Eclipse but I can't figure out how to get the classes in /bin (which are theoretically the same that are in the generated jars) to be mapped to their corresponding .jar.so (in Eclipse, set up a run profile that runs an Eclipse application with gij (you can set up an installed JRE for java-gcj-compat pretty easily by using /usr/lib/jvm as the location in the Installed JREs dialog)) From greenrd at greenrd.org Fri Oct 14 16:49:30 2005 From: greenrd at greenrd.org (Robin Green) Date: Fri, 14 Oct 2005 16:49:30 +0000 Subject: [fedora-java] Re: Native code not loaded when it is in the db References: <20051007134050.GA23465@redhat.com> Message-ID: On Fri, 07 Oct 2005 09:40:50 -0400, Andrew Overholt wrote: > I have natively-compiled all of Eclipse and created .db files and merged > them all into the system-wide .db. I have verified that the .db file that > is being used by gij (this is a gcc that I have built from 4.0 branch head) > contains the .jar,.jar.so combinations that I want to use. However, two of > the jars are being loaded as bytecode when I want them to be loaded from > their corresponding .so (as evidenced by the following). Is this bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163969 ? This was "fixed" ages ago - except it wasn't actually fixed for a long while, because rawhide eclipse was not updated for about 2 months. -- Robin From aph at redhat.com Mon Oct 17 16:17:24 2005 From: aph at redhat.com (Andrew Haley) Date: Mon, 17 Oct 2005 17:17:24 +0100 Subject: [fedora-java] JOnAS 4.3.3 on gcj: test results Message-ID: <17235.52884.243814.17824@zapata.pink> I fixed a few testsuite and gcj bugs and re-ran the conformance tests on jonas-4.3.3-1jpp_13fc. We're up to 94.39%. Results here: http://people.redhat.com/~aph/current-jonas-on-gcj-conformance-test-results.html Not that you won't be able to do this until the bug fixes for the testsuite go in to rawhide. The main cause for concern seems to be org.objectweb.jonas.jtests.clients.jms, almost all of which fail. I'm not sure what is going on there. Andrew. From aph at redhat.com Tue Oct 18 10:34:48 2005 From: aph at redhat.com (Andrew Haley) Date: Tue, 18 Oct 2005 11:34:48 +0100 Subject: [fedora-java] Re: JOnAS 4.3.3 on gcj: test results In-Reply-To: <17235.52884.243814.17824@zapata.pink> References: <17235.52884.243814.17824@zapata.pink> Message-ID: <17236.53192.819634.914549@zapata.pink> http://people.redhat.com/~aph/current-jonas-on-gcj-conformance-test-results.html Archit Shah pointed out that all the JMS tests were failing due to the class org.objectweb.joram.client.jms.admin.ObjectFactory not being installed. Fixing that brings our total up to 96.79%, which as far as I'm aware is as good as we're likely to get with 4.3.x jonas. As far as I'm aware this is the first time that anyone has run a full Java app server on a system using entirely free software, from the operating system kernel all the way to the top of the stack. Thanks and congratulations to everyone who has been involved. There's still plenty to do, in particular reducing memory consumption, but this is a significant milestone. We'll respin gcj and JOnAS in Fedora in a week or two, and after that I'd like to set up regular testing of JOnAS on gcj to make sure we don't break anything. Andrew. From david at zarb.org Tue Oct 18 10:44:49 2005 From: david at zarb.org (David Walluck) Date: Tue, 18 Oct 2005 06:44:49 -0400 Subject: [fedora-java] Re: JOnAS 4.3.3 on gcj: test results In-Reply-To: <17236.53192.819634.914549@zapata.pink> References: <17235.52884.243814.17824@zapata.pink> <17236.53192.819634.914549@zapata.pink> Message-ID: <4354D221.6020708@zarb.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Andrew Haley wrote: > As far as I'm aware this is the first time that anyone has run a full > Java app server on a system using entirely free software, from the > operating system kernel all the way to the top of the stack. Thanks This isn't entirely true since there are a bunch of pre-built jars left in the packages. Even if it's because ow uses patched versions of these jars, it makes sense to package these patched versions separately. Otherwise, how can you be sure they will even build with free software stack? - -- Sincerely, David Walluck -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDVNIharJDwJ6gwowRAgVzAJ0Z4/IV0o/+heTizlky4tKVpImUaQCePTMr MRcgIHZe55PSjwMCVqcUfqA= =9OgB -----END PGP SIGNATURE----- From aph at redhat.com Tue Oct 18 10:56:32 2005 From: aph at redhat.com (Andrew Haley) Date: Tue, 18 Oct 2005 11:56:32 +0100 Subject: [fedora-java] Re: JOnAS 4.3.3 on gcj: test results In-Reply-To: <4354D221.6020708@zarb.org> References: <17235.52884.243814.17824@zapata.pink> <17236.53192.819634.914549@zapata.pink> <4354D221.6020708@zarb.org> Message-ID: <17236.54496.746309.507008@zapata.pink> David Walluck writes: > > Andrew Haley wrote: > > > As far as I'm aware this is the first time that anyone has run a full > > Java app server on a system using entirely free software, from the > > operating system kernel all the way to the top of the stack. Thanks > > This isn't entirely true since there are a bunch of pre-built jars left > in the packages. I thought these had been purged. If you are aware of any unfree software in Fedora, let me know. Andrew. From charlescurley at charlescurley.com Tue Oct 18 14:37:56 2005 From: charlescurley at charlescurley.com (Charles Curley) Date: Tue, 18 Oct 2005 08:37:56 -0600 Subject: [fedora-java] tomcat not expanding macros Message-ID: <20051018143756.GE26151@charlescurley.com> I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has sent me a war package, foo.war, to install & document. I can deploy it via tomcat's manager. One problem I hit is that the web.xml file contains a macro, used several places, e.g: -------------------------------------------------- ${catalina.home}/webapps/foo -------------------------------------------------- It apears it is not being expanded, so that the log fle complains like so: -------------------------------------------------- ERROR: License Manager Error: ${catalina.home}/webapps/FOO/WEB-INF/lib/foo.jar (No such file or directory) ERROR: License Manager Error: ${catalina.home}/webapps/FOO/WEB-INF/lib/foo.jar (No such file or directory) -------------------------------------------------- When I globally search and replace the correct expansion of the macro into web.xml, I get other problems but not that one, leading me to think that the problem is lack of macro expansion. I don't see anything in the tomcat docs that suggests that macro expansion can be turned on or off. -- Charles Curley /"\ ASCII Ribbon Campaign Looking for fine software \ / Respect for open standards and/or writing? X No HTML/RTF in email http://www.charlescurley.com / \ No M$ Word docs in email Key fingerprint = CE5C 6645 A45A 64E4 94C0 809C FFF6 4C48 4ECD DFDB -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From gbenson at redhat.com Wed Oct 19 08:49:38 2005 From: gbenson at redhat.com (Gary Benson) Date: Wed, 19 Oct 2005 09:49:38 +0100 Subject: [fedora-java] tomcat not expanding macros In-Reply-To: <20051018143756.GE26151@charlescurley.com> References: <20051018143756.GE26151@charlescurley.com> Message-ID: <20051019084935.GA8020@redhat.com> Charles Curley wrote: > I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has sent > me a war package, foo.war, to install & document. I can deploy it > via tomcat's manager. > > One problem I hit is that the web.xml file contains a macro, used > several places, e.g: > > -------------------------------------------------- > ${catalina.home}/webapps/foo > -------------------------------------------------- [snip] > > I don't see anything in the tomcat docs that suggests that macro > expansion can be turned on or off. Does Tomcat have macro expansion? That looks like an Ant macro to me... Cheers, Gary From charlescurley at charlescurley.com Wed Oct 19 13:27:18 2005 From: charlescurley at charlescurley.com (Charles Curley) Date: Wed, 19 Oct 2005 07:27:18 -0600 Subject: [fedora-java] tomcat not expanding macros In-Reply-To: <20051019084935.GA8020@redhat.com> References: <20051018143756.GE26151@charlescurley.com> <20051019084935.GA8020@redhat.com> Message-ID: <20051019132718.GA16675@charlescurley.com> On Wed, Oct 19, 2005 at 09:49:38AM +0100, Gary Benson wrote: > Charles Curley wrote: > > I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has sent > > me a war package, foo.war, to install & document. I can deploy it > > via tomcat's manager. > > > > One problem I hit is that the web.xml file contains a macro, used > > several places, e.g: > > > > -------------------------------------------------- > > ${catalina.home}/webapps/foo > > -------------------------------------------------- > [snip] > > > > I don't see anything in the tomcat docs that suggests that macro > > expansion can be turned on or off. > > Does Tomcat have macro expansion? That looks like an Ant macro to me... Yes, it does look like an Ant macro. Maybe they share the same macro expansion code. I installed a Sun JVM and Tomcat from Apache's web site on another machine, and have had no problems with this. I moved the Tomcat tree to another location, and it continued to work. -- Charles Curley /"\ ASCII Ribbon Campaign Looking for fine software \ / Respect for open standards and/or writing? X No HTML/RTF in email http://www.charlescurley.com / \ No M$ Word docs in email Key fingerprint = CE5C 6645 A45A 64E4 94C0 809C FFF6 4C48 4ECD DFDB -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From aph at redhat.com Wed Oct 19 13:36:55 2005 From: aph at redhat.com (Andrew Haley) Date: Wed, 19 Oct 2005 14:36:55 +0100 Subject: [fedora-java] tomcat not expanding macros In-Reply-To: <20051019132718.GA16675@charlescurley.com> References: <20051018143756.GE26151@charlescurley.com> <20051019084935.GA8020@redhat.com> <20051019132718.GA16675@charlescurley.com> Message-ID: <17238.19447.958722.414488@zapata.pink> Charles Curley writes: > On Wed, Oct 19, 2005 at 09:49:38AM +0100, Gary Benson wrote: > > Charles Curley wrote: > > > I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has sent > > > me a war package, foo.war, to install & document. I can deploy it > > > via tomcat's manager. > > > > > > One problem I hit is that the web.xml file contains a macro, used > > > several places, e.g: > > > > > > -------------------------------------------------- > > > ${catalina.home}/webapps/foo > > > -------------------------------------------------- > > [snip] > > > > > > I don't see anything in the tomcat docs that suggests that macro > > > expansion can be turned on or off. > > > > Does Tomcat have macro expansion? That looks like an Ant macro to me... > > Yes, it does look like an Ant macro. Maybe they share the same macro > expansion code. > > I installed a Sun JVM and Tomcat from Apache's web site on another > machine, and have had no problems with this. I moved the Tomcat tree > to another location, and it continued to work. Where is the macro ${catalina.home} defined? How does this macro propagate to the context in which Tomcat reads web.xml? Is there any way to prepare a test case for this stuff? Andrew. From green at redhat.com Wed Oct 19 17:53:44 2005 From: green at redhat.com (Anthony Green) Date: Wed, 19 Oct 2005 10:53:44 -0700 Subject: [fedora-java] batik [Fwd: Re: com.sun.image.*] In-Reply-To: <433B8965.5040802@pobox.com> References: <433B8965.5040802@pobox.com> Message-ID: <1129744425.2969.14.camel@localhost.localdomain> On Wed, 2005-09-28 at 23:27 -0700, Brion Vibber wrote: > Anthony Green wrote: > > If anybody is interested in running batik on our free stack, here's some > > info on what needs doing... > > Has anybody been working on this? We'd be interested in using Batik to > rasterize SVG content for Wikipedia, and as we have a bit of a free > software fetish it would be nice to make sure it can run under GCJ. > > (Currently we're using librsvg, which is not totally satisfactory as > there are some things it doesn't render too well.) > > If nobody's working on it, I might pick it up sometime in the next while > depending on how hard it looks. :) Brion - have you looked into this at all? The Fedora docs people would like to use FOP to help convert DocBook XML into various outputs using a fully free stack. Unfortunately FOP depends on Batik.... AG From brion at pobox.com Wed Oct 19 18:31:57 2005 From: brion at pobox.com (Brion Vibber) Date: Wed, 19 Oct 2005 11:31:57 -0700 Subject: [fedora-java] batik [Fwd: Re: com.sun.image.*] In-Reply-To: <1129744425.2969.14.camel@localhost.localdomain> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> Message-ID: <4356911D.3040205@pobox.com> Anthony Green wrote: > On Wed, 2005-09-28 at 23:27 -0700, Brion Vibber wrote: >>Anthony Green wrote: >>>If anybody is interested in running batik on our free stack, here's some >>>info on what needs doing... >> >>Has anybody been working on this? We'd be interested in using Batik to >>rasterize SVG content for Wikipedia, and as we have a bit of a free >>software fetish it would be nice to make sure it can run under GCJ. >> >>(Currently we're using librsvg, which is not totally satisfactory as >>there are some things it doesn't render too well.) >> >>If nobody's working on it, I might pick it up sometime in the next while >>depending on how hard it looks. :) > > Brion - have you looked into this at all? I've looked a little bit at the files but haven't had a chance to try actually working on it. If someone else is itching to work on it, PLEASE feel free to start! > The Fedora docs people would like to use FOP to help convert DocBook XML > into various outputs using a fully free stack. Unfortunately FOP > depends on Batik.... -- brion vibber (brion @ pobox.com) -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 253 bytes Desc: OpenPGP digital signature URL: From green at redhat.com Wed Oct 19 23:11:28 2005 From: green at redhat.com (Anthony Green) Date: Wed, 19 Oct 2005 16:11:28 -0700 Subject: Fedora docs processing (Was: [fedora-java] batik [Fwd: Re: com.sun.image.*]) In-Reply-To: <4356911D.3040205@pobox.com> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> <4356911D.3040205@pobox.com> Message-ID: <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> On Wed, 2005-10-19 at 11:31 -0700, Brion Vibber wrote: > I've looked a little bit at the files but haven't had a chance to try > actually working on it. If someone else is itching to work on it, PLEASE > feel free to start! Ok, thanks for the update Brion. I'm wondering if the simplest solution for the Fedora Docs team is for us to strip out Batik's JPEG handling. Perhaps it's not required for their work-flow. Karsten - can you provide some simple docs and show us how you would process them using a proprietary Java solution? We should be able to determine from this whether or not we can ignore the problem classes (as a near term solution). Thanks, AG From gbenson at redhat.com Thu Oct 20 09:49:45 2005 From: gbenson at redhat.com (Gary Benson) Date: Thu, 20 Oct 2005 10:49:45 +0100 Subject: [fedora-java] tomcat not expanding macros In-Reply-To: <20051019132718.GA16675@charlescurley.com> References: <20051018143756.GE26151@charlescurley.com> <20051019084935.GA8020@redhat.com> <20051019132718.GA16675@charlescurley.com> Message-ID: <20051020094942.GD5122@redhat.com> Charles Curley wrote: > On Wed, Oct 19, 2005 at 09:49:38AM +0100, Gary Benson wrote: > > Charles Curley wrote: > > > I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has > > > sent me a war package, foo.war, to install & document. I can > > > deploy it via tomcat's manager. > > > > > > One problem I hit is that the web.xml file contains a macro, > > > used several places, e.g: > > > > > > -------------------------------------------------- > > > ${catalina.home}/webapps/foo > > > -------------------------------------------------- > > [snip] > > > > > > I don't see anything in the tomcat docs that suggests that macro > > > expansion can be turned on or off. > > > > Does Tomcat have macro expansion? That looks like an Ant macro to > > me... > > Yes, it does look like an Ant macro. Maybe they share the same macro > expansion code. Or maybe the macro was supposed to be expanded when foo.war was built, by the Ant that built it, however... > I installed a Sun JVM and Tomcat from Apache's web site on another > machine, and have had no problems with this. I moved the Tomcat tree > to another location, and it continued to work. ...if this is the case then I guess it _is_ supposed to be expanded by Tomcat. It seems fishy though, as any webapp that did such a thing would automatically become container-specific. Is it actually part of the Servlet specification or is it a Tomcat-specific feature do you know? Cheers, Gary From charlescurley at charlescurley.com Thu Oct 20 12:23:04 2005 From: charlescurley at charlescurley.com (Charles Curley) Date: Thu, 20 Oct 2005 06:23:04 -0600 Subject: [fedora-java] tomcat not expanding macros In-Reply-To: <20051020094942.GD5122@redhat.com> References: <20051018143756.GE26151@charlescurley.com> <20051019084935.GA8020@redhat.com> <20051019132718.GA16675@charlescurley.com> <20051020094942.GD5122@redhat.com> Message-ID: <20051020122304.GK16675@charlescurley.com> On Thu, Oct 20, 2005 at 10:49:45AM +0100, Gary Benson wrote: > Charles Curley wrote: > > On Wed, Oct 19, 2005 at 09:49:38AM +0100, Gary Benson wrote: > > > Charles Curley wrote: > > > > I have tomcat5-5.0.30-5jpp_6fc on FC4 as updated. My client has > > > > sent me a war package, foo.war, to install & document. I can > > > > deploy it via tomcat's manager. > > > > > > > > One problem I hit is that the web.xml file contains a macro, > > > > used several places, e.g: > > > > > > > > -------------------------------------------------- > > > > ${catalina.home}/webapps/foo > > > > -------------------------------------------------- > > > [snip] > > > > > > > > I don't see anything in the tomcat docs that suggests that macro > > > > expansion can be turned on or off. > > > > > > Does Tomcat have macro expansion? That looks like an Ant macro to > > > me... > > > > Yes, it does look like an Ant macro. Maybe they share the same macro > > expansion code. > > Or maybe the macro was supposed to be expanded when foo.war was built, > by the Ant that built it, however... I doubt that. The build box might have a different ${catalina.home}, or no Tomcat at all. > > > I installed a Sun JVM and Tomcat from Apache's web site on another > > machine, and have had no problems with this. I moved the Tomcat tree > > to another location, and it continued to work. > > ...if this is the case then I guess it _is_ supposed to be expanded by > Tomcat. It seems fishy though, as any webapp that did such a thing > would automatically become container-specific. Is it actually part of > the Servlet specification or is it a Tomcat-specific feature do you > know? I searched the spec on "macro", and "$" and found nothing that indicated that it was required by the spec. However, SRV.1.5 Relationship to Java 2 Platform, Enterprise Edition The Java Servlet API v.2.4 is a required API of the Java 2 Platform, Enterprise Edition, v.1.41. Servlet containers and servlets deployed into them must meet additional requirements, described in the J2EE specification, for executing in a J2EE environment. (Java(tm) Servlet Specification Version 2.4P) I have not checked the J2EE spec. I can ask my client, but they're in a product release crunch, so I shouldn't right now. > > Cheers, > Gary > > -- > fedora-devel-java-list mailing list > fedora-devel-java-list at redhat.com > https://www.redhat.com/mailman/listinfo/fedora-devel-java-list -- Charles Curley /"\ ASCII Ribbon Campaign Looking for fine software \ / Respect for open standards and/or writing? X No HTML/RTF in email http://www.charlescurley.com / \ No M$ Word docs in email Key fingerprint = CE5C 6645 A45A 64E4 94C0 809C FFF6 4C48 4ECD DFDB -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From green at redhat.com Fri Oct 21 23:28:00 2005 From: green at redhat.com (Anthony Green) Date: Fri, 21 Oct 2005 16:28:00 -0700 Subject: Fedora docs processing (Was: [fedora-java] batik [Fwd: Re: com.sun.image.*]) In-Reply-To: <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> <4356911D.3040205@pobox.com> <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> Message-ID: <1129937281.2988.75.camel@dhcp-172-16-25-224.sfbay.redhat.com> On Wed, 2005-10-19 at 16:11 -0700, Anthony Green wrote: > I'm wondering if the simplest solution for the Fedora Docs team is for > us to strip out Batik's JPEG handling. Perhaps it's not required for > their work-flow. It turns out that it's impossible to simply strip out the JPEG code because some of the proprietary Sun JPEG APIs are used for more general non-JPEG tasks. "Liberating" Batik will take some real effort (I'm looking for a cheap solution right now - since this is something we want in the FC5 timeframe). So, the next option is to look at using FOP without Batik. If this is even possible I suppose this means no SVG processing in FOP. Perhaps the doc generation work-flow could be tweaked to pre-render the SVG into some format FOP can handle. I'll poke at this a little. AG From david at zarb.org Sat Oct 22 01:40:09 2005 From: david at zarb.org (David Walluck) Date: Fri, 21 Oct 2005 21:40:09 -0400 Subject: [fedora-java] Re: JOnAS 4.3.3 on gcj: test results In-Reply-To: <17236.54496.746309.507008@zapata.pink> References: <17235.52884.243814.17824@zapata.pink> <17236.53192.819634.914549@zapata.pink> <4354D221.6020708@zarb.org> <17236.54496.746309.507008@zapata.pink> Message-ID: <43599879.9020001@zarb.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 - -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Andrew Haley wrote: > I thought these had been purged. If you are aware of any unfree > software in Fedora, let me know. There are some bootstrap issues with jonas (which I am sure you're aware of): * jonas needs jorm * jorm needs medor * medor needs medor-expression and jorm * medor-expression needs jorm You probably fixed this by at first including a binary-only (non-free-compiled) jar. I was able to find some minimal subset of the classes needed and include them into the initial medor-expression and jorm packages. From here one can rebuild without these bootstrap classes. If anyone else is interested, specifically what I did was: * build medor-expression with a small subset of jorm * build jorm with some small subset of medor (I tried it the other way, and realized after much pain that this wasn't the way to go) * rebuild medor-expression * build medor * rebuild jorm * build jonas (*) (*) Now, there are a couple jars left in the jonas package. Specifically, I found externals/ishmael.jar externals/tools/xbean-apache-1.0-DEV.jar I guess that the xmlbeans 1.0.4 rpm from jpackage does not work? Whatever this version of xbean is, it looks to have come from between 1.0.3 and 1.0.4. I would be surprised if neither of these worked. I didn't try yet because I assumed there was a reason for keeping this. But, the outlook for xbean looks good, assuming we can find out what version it is. But I also checked the xmlbeans cvn and I am not sure you can get date snapshots. As for ishmael, I am not sure yet how to build this and it looks to contain some non-free j2ee classes. The outlook for ishmael is not so good. I couldn't even tell how to build it from my initial glance at it (if it needs jonas, then that's another big problem). - - -- Sincerely, David Walluck - -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDWZRsarJDwJ6gwowRAtxIAJ4w9W0NuivbsT8zokoOEB13OUoPdwCfV8ak yRgske4DQH9UeYSmHzDZy+4= =Y+et - -----END PGP SIGNATURE----- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDWZh4arJDwJ6gwowRAolBAJ9BgPN8TowKJxcW/6eVjPAXrE+FmwCffYmx YNiGXiOe/GkOzkJRm4AnXuk= =1PZC -----END PGP SIGNATURE----- From green at redhat.com Sat Oct 22 02:26:46 2005 From: green at redhat.com (Anthony Green) Date: Fri, 21 Oct 2005 19:26:46 -0700 Subject: Fedora docs processing (Was: [fedora-java] batik [Fwd: Re: com.sun.image.*]) In-Reply-To: <1129937281.2988.75.camel@dhcp-172-16-25-224.sfbay.redhat.com> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> <4356911D.3040205@pobox.com> <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> <1129937281.2988.75.camel@dhcp-172-16-25-224.sfbay.redhat.com> Message-ID: <1129948006.2985.1.camel@localhost.localdomain> On Fri, 2005-10-21 at 16:28 -0700, Anthony Green wrote: > So, the next option is to look at using FOP without Batik. If this is > even possible I suppose this means no SVG processing in FOP. This isn't possible. Batik is fundamental to getting fop to work. We're stuck until somebody tackles Batik. AG From vadimn at redhat.com Sat Oct 22 15:07:47 2005 From: vadimn at redhat.com (Vadim Nasardinov) Date: Sat, 22 Oct 2005 11:07:47 -0400 Subject: bootstrap issues in Jonas (was: Re: [fedora-java] Re: JOnAS 4.3.3 on gcj: test results) In-Reply-To: <43599879.9020001@zarb.org> References: <17235.52884.243814.17824@zapata.pink> <17236.54496.746309.507008@zapata.pink> <43599879.9020001@zarb.org> Message-ID: <200510221107.47791.vadimn@redhat.com> On Friday 21 October 2005 21:40, David Walluck wrote: > There are some bootstrap issues with jonas (which I am sure you're aware > of): > > * jonas needs jorm > * jorm needs medor > * medor needs medor-expression and jorm > * medor-expression needs jorm I think this has been addressed: - http://mail-archive.objectweb.org/speedo/2005-07/msg00012.html - http://cvs.forge.objectweb.org/cgi-bin/viewcvs.cgi/medor/expression/ - http://cvs.forge.objectweb.org/cgi-bin/viewcvs.cgi/jorm/adapter/ (But Fedora's version of Jonas predates those changes.) From green at redhat.com Sat Oct 22 19:09:13 2005 From: green at redhat.com (Anthony Green) Date: Sat, 22 Oct 2005 12:09:13 -0700 Subject: Fedora docs processing (Was: [fedora-java] batik [Fwd: Re: com.sun.image.*]) In-Reply-To: <1129948006.2985.1.camel@localhost.localdomain> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> <4356911D.3040205@pobox.com> <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> <1129937281.2988.75.camel@dhcp-172-16-25-224.sfbay.redhat.com> <1129948006.2985.1.camel@localhost.localdomain> Message-ID: <1130008153.2985.17.camel@localhost.localdomain> On Fri, 2005-10-21 at 19:26 -0700, Anthony Green wrote: > On Fri, 2005-10-21 at 16:28 -0700, Anthony Green wrote: > > So, the next option is to look at using FOP without Batik. If this is > > even possible I suppose this means no SVG processing in FOP. > > This isn't possible. Batik is fundamental to getting fop to work. > > We're stuck until somebody tackles Batik. Ok, so I tackled Batik and have an SRPM that builds now. This basically boiled down to removing JPEG and TIFF image handling, but - in theory - PNG support should work. The changes weren't nearly as bad as I first thought. Other than removing classes from the build, I only had to add two patches. But now we're stuck on GNU Classpath/libgcj not implementing some core AWT functionality in java.awt.BasicStroke. There was already a bug filed for this: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22938 gnu.java.awt.peer.gtk.GdkTextLayout.getCharacterLevel(int) is also a problem. Hopefully somebody (fitzsim?) can comment on the status of these classes. This goal of rendering Fedora docs on a free stack is beginning to look like an FC6 goal. AG From green at redhat.com Sat Oct 22 19:43:17 2005 From: green at redhat.com (Anthony Green) Date: Sat, 22 Oct 2005 12:43:17 -0700 Subject: [fedora-java] ecj problem and patch In-Reply-To: <1124988273.3983.87.camel@tortoise.toronto.redhat.com> References: <1124845178.3057.74.camel@localhost.localdomain> <1124988273.3983.87.camel@tortoise.toronto.redhat.com> Message-ID: <1130010197.2985.21.camel@localhost.localdomain> On Thu, 2005-08-25 at 12:44 -0400, Thomas Fitzsimmons wrote: > On Thu, 2005-08-25 at 07:28 -0600, Tom Tromey wrote: > > >>>>> "Anthony" == Anthony Green writes: > > > > Anthony> I produced the following patch, which needs to be applied > > Anthony> last in our set of Eclipse patches. > > > > Want to file this in Eclipse bugzilla? > > If you do, let me know the PR number as I'd like to track it. > > I'd like to test this before we send the patch. I'm not convinced it's > entirely correct. But we probably should file the bug itself in Eclipse > bugzilla. I've done this now: https://bugs.eclipse.org/bugs/show_bug.cgi?id=113458 I needed to patch Eclipse in order to build Batik. That bugzilla entry includes a patch against Eclipse 3.1.1. AG From green at redhat.com Sat Oct 22 23:15:05 2005 From: green at redhat.com (Anthony Green) Date: Sat, 22 Oct 2005 16:15:05 -0700 Subject: [fedora-java] Release Notes Message-ID: <1130022905.2985.48.camel@localhost.localdomain> The current "Java" content for the release notes (from FC4) are pretty slim. I'll quote them below with my comments and questions... > Fedora Core 4 users are advised not to use the Java RPM provided by > Sun. It contains Provides that conflict with names used in packages > provided as part of Fedora Core 4. Because of this, Sun Java might > disappear from an installed system during package upgrade operations. I assume this is still true. Does anybody have a pointer to the details? > Fedora Core 4 users should use either the RPM available from > http://www.jpackage.org/, or manually install the Sun Java tarball > into `/opt/`. I think we should discourage the /opt solution, since it doesn't integrate with out alternatives-based solution. Also, I thought there were problems with using the binary RPM from JPackage, and users _have_ to rebuild from the SRPM. Does anybody know about this? > Sun Java 1.5+ is recommended for stability purposes. Is this really true? What about IBM or BEA? Should we even be making specific recommendations here? Perhaps this is best avoided. Thanks, AG From david at zarb.org Sat Oct 22 23:51:43 2005 From: david at zarb.org (David Walluck) Date: Sat, 22 Oct 2005 19:51:43 -0400 Subject: [fedora-java] Release Notes In-Reply-To: <1130022905.2985.48.camel@localhost.localdomain> References: <1130022905.2985.48.camel@localhost.localdomain> Message-ID: <435AD08F.2010401@zarb.org> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Anthony Green wrote: >>Fedora Core 4 users are advised not to use the Java RPM provided by >>Sun. It contains Provides that conflict with names used in packages >>provided as part of Fedora Core 4. Because of this, Sun Java might >>disappear from an installed system during package upgrade operations. > > > I assume this is still true. Does anybody have a pointer to the > details? This shouldn't be the case if the java-1.4.2-sun-compat jpackage rpm works (this is supposed to be an jpackage-compatible add-on to the Sun rpm). Also, even if this is a problem, we should at least discuss if there are any workarounds. > I think we should discourage the /opt solution, since it doesn't > integrate with out alternatives-based solution. I agree. It also subverts the packaging system entirely. If a commercial rpm is desired, it is best to build the jpackage rpms. Is there any other way? > Also, I thought there were problems with using the binary RPM from > JPackage, and users _have_ to rebuild from the SRPM. Does anybody > know about this? I don't know, aren't these normally built on FC or something ``close enough''? So the questions remain: does the sun-compat rpm work and do the normal rpms work? (I wish I could answer but I've been out of the loop lately). > Is this really true? What about IBM or BEA? Should we even be > making specific recommendations here? Perhaps this is best > avoided. FC, being free, shouldn't be recommending non-free solutions (IMO). Yes, users will be interested in integration with commercial jvm's (not necessarily Sun either), but not only is this statement suggesting the use of non-free software, but it is somehow implying that it's supported by FC, when FC really only accepts bug reports for gcj. - -- Sincerely, David Walluck -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) iD8DBQFDWtCParJDwJ6gwowRAjJSAJ9Ifw4e4PPo7GoqaqII73D3HuZBEACfV1nc ac1ViAKeYHAyI2g0bofuVgY= =TUd7 -----END PGP SIGNATURE----- From green at redhat.com Sun Oct 23 17:48:05 2005 From: green at redhat.com (Anthony Green) Date: Sun, 23 Oct 2005 10:48:05 -0700 Subject: [fedora-java] ecj problem and patch In-Reply-To: <1130010197.2985.21.camel@localhost.localdomain> References: <1124845178.3057.74.camel@localhost.localdomain> <1124988273.3983.87.camel@tortoise.toronto.redhat.com> <1130010197.2985.21.camel@localhost.localdomain> Message-ID: <1130089685.2985.76.camel@localhost.localdomain> On Sat, 2005-10-22 at 12:43 -0700, Anthony Green wrote: > I've done this now: https://bugs.eclipse.org/bugs/show_bug.cgi?id=113458 > > I needed to patch Eclipse in order to build Batik. That bugzilla entry > includes a patch against Eclipse 3.1.1. Ok, apparently this is wrong. The JDK 1.5 documentation seems to clear up once and for all the search order for classes. So, it seems like that way to deal with this is to force the problem package onto the endorsed dir path. Is there even any way to specify this to the eclipse compiler from the ant command line? AG From gbenson at redhat.com Mon Oct 24 09:46:27 2005 From: gbenson at redhat.com (Gary Benson) Date: Mon, 24 Oct 2005 10:46:27 +0100 Subject: [fedora-java] Release Notes In-Reply-To: <1130022905.2985.48.camel@localhost.localdomain> References: <1130022905.2985.48.camel@localhost.localdomain> Message-ID: <20051024094626.GC12313@redhat.com> Anthony Green wrote: > The current "Java" content for the release notes (from FC4) are > pretty slim. I'll quote them below with my comments and > questions... > > > Fedora Core 4 users are advised not to use the Java RPM provided > > by Sun. It contains Provides that conflict with names used in > > packages provided as part of Fedora Core 4. Because of this, Sun > > Java might disappear from an installed system during package > > upgrade operations. > > I assume this is still true. Does anybody have a pointer to the > details? I'd drop the "Because of this..." bit, since there are many other bad things that would be tedious to list. > > Fedora Core 4 users should use either the RPM available from > > http://www.jpackage.org/, or manually install the Sun Java tarball > > into `/opt/`. > > I think we should discourage the /opt solution, since it doesn't > integrate with out alternatives-based solution. I wouldn't mention it at all. > Also, I thought there were problems with using the binary RPM from > JPackage, and users _have_ to rebuild from the SRPM. Does anybody > know about this? Users have to fetch the proprietary code themselves, since JPackage can't distribute it. > > Sun Java 1.5+ is recommended for stability purposes. > > Is this really true? What about IBM or BEA? Should we even be > making specific recommendations here? Perhaps this is best avoided. We shouldn't be recommending proprietary stuff IMO. Cheers, Gary From fitzsim at redhat.com Mon Oct 24 13:47:32 2005 From: fitzsim at redhat.com (Thomas Fitzsimmons) Date: Mon, 24 Oct 2005 09:47:32 -0400 Subject: Fedora docs processing (Was: [fedora-java] batik [Fwd: Re: com.sun.image.*]) In-Reply-To: <1130008153.2985.17.camel@localhost.localdomain> References: <433B8965.5040802@pobox.com> <1129744425.2969.14.camel@localhost.localdomain> <4356911D.3040205@pobox.com> <1129763488.2843.19.camel@dhcp-172-16-25-224.sfbay.redhat.com> <1129937281.2988.75.camel@dhcp-172-16-25-224.sfbay.redhat.com> <1129948006.2985.1.camel@localhost.localdomain> <1130008153.2985.17.camel@localhost.localdomain> Message-ID: <1130161652.12761.13.camel@tortoise.toronto.redhat.com> On Sat, 2005-10-22 at 12:09 -0700, Anthony Green wrote: > On Fri, 2005-10-21 at 19:26 -0700, Anthony Green wrote: > > On Fri, 2005-10-21 at 16:28 -0700, Anthony Green wrote: > > > So, the next option is to look at using FOP without Batik. If this is > > > even possible I suppose this means no SVG processing in FOP. > > > > This isn't possible. Batik is fundamental to getting fop to work. > > > > We're stuck until somebody tackles Batik. > > Ok, so I tackled Batik and have an SRPM that builds now. This basically > boiled down to removing JPEG and TIFF image handling, but - in theory - > PNG support should work. The changes weren't nearly as bad as I first > thought. Other than removing classes from the build, I only had to add > two patches. > > But now we're stuck on GNU Classpath/libgcj not implementing some core > AWT functionality in java.awt.BasicStroke. > > There was already a bug filed for this: > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=22938 > > gnu.java.awt.peer.gtk.GdkTextLayout.getCharacterLevel(int) is also a > problem. > > Hopefully somebody (fitzsim?) can comment on the status of these > classes. Yeah, there are two large chunks of work still to do in the GTK peers; port GdkGraphics2D to Cairo 1.0 and implement java.awt.font using Pango and Freetype. I'm hoping to finish the first task before FC5, but the second one will likely have to wait. Tom > > This goal of rendering Fedora docs on a free stack is beginning to look > like an FC6 goal. > > AG > > From che666 at gmail.com Mon Oct 24 13:57:20 2005 From: che666 at gmail.com (Rudolf Kastl) Date: Mon, 24 Oct 2005 15:57:20 +0200 Subject: [fedora-java] Release Notes In-Reply-To: <20051024094626.GC12313@redhat.com> References: <1130022905.2985.48.camel@localhost.localdomain> <20051024094626.GC12313@redhat.com> Message-ID: 2005/10/24, Gary Benson : > Anthony Green wrote: > > The current "Java" content for the release notes (from FC4) are > > pretty slim. I'll quote them below with my comments and > > questions... > > > > > Fedora Core 4 users are advised not to use the Java RPM provided > > > by Sun. It contains Provides that conflict with names used in > > > packages provided as part of Fedora Core 4. Because of this, Sun > > > Java might disappear from an installed system during package > > > upgrade operations. > > > > I assume this is still true. Does anybody have a pointer to the > > details? > > I'd drop the "Because of this..." bit, since there are many other bad > things that would be tedious to list. especially the past shows that... extremly dangerous bugs in the web browser plugin that have been fixed 6 months after their initial discovery... > > > > Fedora Core 4 users should use either the RPM available from > > > http://www.jpackage.org/, or manually install the Sun Java tarball > > > into `/opt/`. > > > > I think we should discourage the /opt solution, since it doesn't > > integrate with out alternatives-based solution. > > I wouldn't mention it at all. Thats probably better in the FAQ anyways. because its an "after installation issue". > > > Also, I thought there were problems with using the binary RPM from > > JPackage, and users _have_ to rebuild from the SRPM. Does anybody > > know about this? > > Users have to fetch the proprietary code themselves, since JPackage > can't distribute it. end users dont know how to build rpms... also currently the package can only be built as root because else rpath freaks out (atleast that was the case the last time i tried to build it in a user environment) > > > > Sun Java 1.5+ is recommended for stability purposes. > > > > Is this really true? What about IBM or BEA? Should we even be > > making specific recommendations here? Perhaps this is best avoided. > > We shouldn't be recommending proprietary stuff IMO. what most end users really need is a working and secure gcj web plugin. end users mainly install java in my experience to get web applets working in their browser. mindmanager and azureus are more features going towards "more beautiful living". > > Cheers, > Gary > > -- > fedora-devel-java-list mailing list > fedora-devel-java-list at redhat.com > https://www.redhat.com/mailman/listinfo/fedora-devel-java-list > From overholt at redhat.com Mon Oct 24 22:00:08 2005 From: overholt at redhat.com (Andrew Overholt) Date: Mon, 24 Oct 2005 18:00:08 -0400 Subject: [fedora-java] Re: Fedora Core 4 Test Update: eclipse-3.1.1-1jpp_1fc.FC4.1 In-Reply-To: <200510242117.j9OLHjFE028021@devserv.devel.redhat.com> References: <200510242117.j9OLHjFE028021@devserv.devel.redhat.com> Message-ID: <20051024220006.GA3246@redhat.com> * Andrew Overholt [2005-10-24 17:18]: > --------------------------------------------------------------------- > Fedora Test Update Notification > FEDORA-2005-1017 > 2005-10-24 > --------------------------------------------------------------------- > > Product : Fedora Core 4 > Name : eclipse > Version : 3.1.1 > Release : 1jpp_1fc.FC4.1 > Summary : An open, extensible IDE > Description : > The Eclipse Platform is designed for building integrated development > environments (IDEs) that can be used to create applications as diverse > as web sites, embedded Java(tm) programs, C++ programs, and Enterprise > JavaBeans(tm). > > --------------------------------------------------------------------- > Update Information: > > This is a release of Eclipse 3.1.1 natively-compiled for > FC4. I had been holding off on releasing this, hoping that > we could fix > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161483 > first, but that looks unlikely at the moment. Since that > bug does not affect all users (and there are some > workarounds for those cases), I'm going ahead with this release. > > Please test and comment on the bugs listed. Sorry, these are the potentially fixed bugs: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=159564 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161518 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161635 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162383 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162443 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163079 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=164095 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165204 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165330 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167548 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=168040 Thanks, Andrew From overholt at redhat.com Mon Oct 24 22:00:08 2005 From: overholt at redhat.com (Andrew Overholt) Date: Mon, 24 Oct 2005 18:00:08 -0400 Subject: [fedora-java] Re: Fedora Core 4 Test Update: eclipse-3.1.1-1jpp_1fc.FC4.1 In-Reply-To: <200510242117.j9OLHjFE028021@devserv.devel.redhat.com> References: <200510242117.j9OLHjFE028021@devserv.devel.redhat.com> Message-ID: <20051024220006.GA3246@redhat.com> * Andrew Overholt [2005-10-24 17:18]: > --------------------------------------------------------------------- > Fedora Test Update Notification > FEDORA-2005-1017 > 2005-10-24 > --------------------------------------------------------------------- > > Product : Fedora Core 4 > Name : eclipse > Version : 3.1.1 > Release : 1jpp_1fc.FC4.1 > Summary : An open, extensible IDE > Description : > The Eclipse Platform is designed for building integrated development > environments (IDEs) that can be used to create applications as diverse > as web sites, embedded Java(tm) programs, C++ programs, and Enterprise > JavaBeans(tm). > > --------------------------------------------------------------------- > Update Information: > > This is a release of Eclipse 3.1.1 natively-compiled for > FC4. I had been holding off on releasing this, hoping that > we could fix > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161483 > first, but that looks unlikely at the moment. Since that > bug does not affect all users (and there are some > workarounds for those cases), I'm going ahead with this release. > > Please test and comment on the bugs listed. Sorry, these are the potentially fixed bugs: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=159564 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161518 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161635 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162383 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162443 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=163079 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=164095 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165204 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=165330 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167548 https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=168040 Thanks, Andrew -- fedora-test-list mailing list fedora-test-list at redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list From overholt at redhat.com Mon Oct 24 22:27:55 2005 From: overholt at redhat.com (Andrew Overholt) Date: Mon, 24 Oct 2005 18:27:55 -0400 Subject: [fedora-java] Non-native Eclipse 3.1 builds - now what? In-Reply-To: References: Message-ID: <20051024222755.GA1619@redhat.com> * Ian Pilcher [2005-09-22 13:07]: > > Note BTW, that changing the package VERSION to 3.1.0_fc is pretty nasty. > Can this be moved to the RELEASE? I fixed this with 3.1.1. Thanks for noticing my stupidity. Andrew From i.pilcher at comcast.net Tue Oct 25 14:09:05 2005 From: i.pilcher at comcast.net (Ian Pilcher) Date: Tue, 25 Oct 2005 09:09:05 -0500 Subject: [fedora-java] Latest OO.o requires java-1.4.2-gcj-compat? Message-ID: The latest openoffice.org update has brought back this requirement. Why? -- ======================================================================== Ian Pilcher i.pilcher at comcast.net ======================================================================== From i.pilcher at comcast.net Tue Oct 25 14:41:33 2005 From: i.pilcher at comcast.net (Ian Pilcher) Date: Tue, 25 Oct 2005 09:41:33 -0500 Subject: [fedora-java] Re: Non-native Eclipse 3.1 builds - now what? In-Reply-To: <20051024222755.GA1619@redhat.com> References: <20051024222755.GA1619@redhat.com> Message-ID: Andrew Overholt wrote: > * Ian Pilcher [2005-09-22 13:07]: > >>Note BTW, that changing the package VERSION to 3.1.0_fc is pretty nasty. >>Can this be moved to the RELEASE? > > > I fixed this with 3.1.1. Thanks for noticing my stupidity. > > Andrew > Thanks. You forgot, however, to change non-gcj builds from noarch to architecture specific. I've bugzilla'ed this as #171711. -- ======================================================================== Ian Pilcher i.pilcher at comcast.net ======================================================================== From overholt at redhat.com Fri Oct 28 19:06:22 2005 From: overholt at redhat.com (Andrew Overholt) Date: Fri, 28 Oct 2005 15:06:22 -0400 Subject: [fedora-java] Re: Non-native Eclipse 3.1 builds - now what? In-Reply-To: References: <20051024222755.GA1619@redhat.com> Message-ID: <20051028190622.GD13018@redhat.com> * Ian Pilcher [2005-10-25 10:47]: > Andrew Overholt wrote: > > * Ian Pilcher [2005-09-22 13:07]: > > > >>Note BTW, that changing the package VERSION to 3.1.0_fc is pretty nasty. > >>Can this be moved to the RELEASE? > > > > > > I fixed this with 3.1.1. Thanks for noticing my stupidity. > > > > Andrew > > > > Thanks. You forgot, however, to change non-gcj builds from noarch to > architecture specific. I've bugzilla'ed this as #171711. Thanks again. Fixed. Andrew From overholt at redhat.com Fri Oct 28 19:06:58 2005 From: overholt at redhat.com (Andrew Overholt) Date: Fri, 28 Oct 2005 15:06:58 -0400 Subject: [fedora-java] Latest OO.o requires java-1.4.2-gcj-compat? In-Reply-To: References: Message-ID: <20051028190658.GE13018@redhat.com> * Ian Pilcher [2005-10-25 10:18]: > The latest openoffice.org update has brought back this requirement. > > Why? I'm not sure. Caolan is the best person to ask. Perhaps bugzilla it? Andrew From mark at klomp.org Sat Oct 29 16:20:52 2005 From: mark at klomp.org (Mark Wielaard) Date: Sat, 29 Oct 2005 18:20:52 +0200 Subject: [fedora-java] ecj problem and patch In-Reply-To: <1130089685.2985.76.camel@localhost.localdomain> References: <1124845178.3057.74.camel@localhost.localdomain> <1124988273.3983.87.camel@tortoise.toronto.redhat.com> <1130010197.2985.21.camel@localhost.localdomain> <1130089685.2985.76.camel@localhost.localdomain> Message-ID: <1130602853.11930.55.camel@localhost.localdomain> Hi Anthony, On Sun, 2005-10-23 at 10:48 -0700, Anthony Green wrote: > Ok, apparently this is wrong. The JDK 1.5 documentation seems to > clear up once and for all the search order for classes. I couldn't quickly find this documentation. Where did you find it? Thanks, Mark -- Escape the Java Trap with GNU Classpath! http://www.gnu.org/philosophy/java-trap.html Join the community at http://planet.classpath.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: