Justin,we ran into a few build failures after your GWC VERSION.txt changes (IRC log attached).
I committed 8e51ce269eaab07cfa04b700e2dcb909133675ed to master: "Fixed broken ant groupId, which appears to have changed from ant to org.apache.ant for 1.8.x. This was causing build failures for anyone without ant:ant:jar:1.8.2 in their local repo. Also changed ant and ant-launcher version to 1.8.4."
Apologies in advance for the cross-module commit, and doubly so if I have b0rked anything.
On the plus side, I can confirm that the GWC version info now appears in VERSION.txt, so your changes work. Thanks!
Kind regards, -- Ben Caradoc-Davies <[email protected]> Software Engineer CSIRO Earth Science and Resource Engineering Australian Resources Research Centre
[2012-07-31 15:13:03] === 8 unknown connection(s) [2012-07-31 15:13:03] =-= User mode for bencaradocdavies is now +i [2012-07-31 15:13:04] *NickServ* This nickname is registered. Please choose a different nickname, or identify via /msg NickServ identify <password>. [2012-07-31 15:13:08] -->| YOU (bencaradocdavies) have joined #geoserver [2012-07-31 15:13:08] =-= Topic for #geoserver is ``All GeoServer chat, all the time (http://geoserver.org)'' [2012-07-31 15:13:08] =-= Topic for #geoserver was set by aaime on Thu 17 Dec 2009 01:38:02 WST [2012-07-31 15:13:09] *NickServ* You are now identified for bencaradocdavies. [2012-07-31 15:16:03] <bencaradocdavies> jgarnett, that gwc change broke the build for anyone without ant:ant:jar:1.8.2 in their local repo as it is gone from maven central [2012-07-31 15:16:53] <jgarnett> darn [2012-07-31 15:17:06] <jgarnett> thinking [2012-07-31 15:17:31] <jgarnett> I can revert it from here, cannot think of an easy fix [2012-07-31 15:17:48] <jgarnett> (unless you know of an a version that exists in maven central) [2012-07-31 15:18:23] <jgarnett> it is such a pain to review patches - that build took me 42:20 according to maven [2012-07-31 15:18:45] <bencaradocdavies> It is a dependency of org.apache.maven.plugins:maven-antrun-plugin:1.7 [2012-07-31 15:19:00] <bencaradocdavies> The sneaky thing to do jgarnett is to deploy to osgeo maven repo. [2012-07-31 15:19:04] <bencaradocdavies> :-) [2012-07-31 15:19:10] <bencaradocdavies> If you have that ant. [2012-07-31 15:19:29] <jgarnett> I apparently do .. but why would I have it and you do not? [2012-07-31 15:19:53] <jgarnett> http://mvnrepository.com/artifact/org.apache.ant/ant lists 1.8.2+ as available [2012-07-31 15:19:54] <sigq> Title: Maven Repository: org.apache.ant ant (at mvnrepository.com) [2012-07-31 15:20:24] <bencaradocdavies> But not in central, which is in the poms, jgarnett. [2012-07-31 15:20:53] <bencaradocdavies> I'm just trying to figure out why it built for jdeolive. [2012-07-31 15:21:05] <bencaradocdavies> Perhaps another project? [2012-07-31 15:21:36] <jgarnett> it builds for me as well [2012-07-31 15:22:39] <bencaradocdavies> Both my local and Jenkins have very fresh local repos, jgarnett. [2012-07-31 15:22:51] <jgarnett> I am not doubting you ben [2012-07-31 15:23:13] <bencaradocdavies> I am not suggesting that you are doubting me [2012-07-31 15:23:18] <bencaradocdavies> just postulating a cause [2012-07-31 15:23:37] <jgarnett> yep, perhaps it was removed from central or something - does central have any versions of ant? [2012-07-31 15:23:44] <bencaradocdavies> that is that your build passes because at some time in the past you got that version [2012-07-31 15:23:51] <bencaradocdavies> Yes 1.7.0 [2012-07-31 15:24:31] <jgarnett> So should i upload 1.8.2 to our repository then? [2012-07-31 15:24:32] <bencaradocdavies> aha [2012-07-31 15:24:39] <bencaradocdavies> there is a 1.8.4 [2012-07-31 15:24:46] <bencaradocdavies> but with a different groupid [2012-07-31 15:24:53] <jgarnett> or back out the change [2012-07-31 15:24:56] <bencaradocdavies> I think maven ant plugin has a bad dependency [2012-07-31 15:25:04] <jgarnett> ⦠thinking - was the dependency not on a 1.8.2 ? [2012-07-31 15:25:28] <jgarnett> (we can make a dependency on 1.8.2 with correct grouped, and then mark the other one as "provided" [2012-07-31 15:25:34] <jgarnett> (as a workaround) [2012-07-31 15:25:45] <bencaradocdavies> maven-antrun-plugin:1.7 depends on ant:ant:jar:1.8.2 [2012-07-31 15:25:57] <bencaradocdavies> not or.apache.ant:ant:jar:1.8.2 [2012-07-31 15:26:13] <jgarnett> That makes sense, they have started getting better about group id use these days [2012-07-31 15:26:31] <bencaradocdavies> wondering if we can try a different antrun plugin [2012-07-31 15:27:40] <bencaradocdavies> jgarnett, hand on, it is our problem. [2012-07-31 15:27:44] <bencaradocdavies> ^hang on [2012-07-31 15:28:02] <bencaradocdavies> The patch from jdeolive adds and explicit ant dependency [2012-07-31 15:28:38] <bencaradocdavies> jgarnett, the top-level pom change refers directly to this nonexistent artifact [2012-07-31 15:29:48] <jgarnett> so if we change that dependency we can give it a go [2012-07-31 15:30:53] <jgarnett> well we still need to change the group id in the top level version, and then elsewhere where it is used [2012-07-31 15:31:17] <jgarnett> <dependency> [2012-07-31 15:31:18] <jgarnett> <groupId>org.apache.ant</groupId> [2012-07-31 15:31:19] <jgarnett> <artifactId>ant</artifactId> [2012-07-31 15:31:20] <jgarnett> <version>${ant.version}</version> [2012-07-31 15:31:21] <jgarnett> </dependency> [2012-07-31 15:32:41] <bencaradocdavies> yes, that seems to fix it. [2012-07-31 15:32:58] <bencaradocdavies> there are a bunch of other places the wrong groupId is used [2012-07-31 15:33:09] <bencaradocdavies> and there are other versions too [2012-07-31 15:33:18] <jgarnett> great ... [2012-07-31 15:34:56] -->| ahocevar ([email protected]) has joined #geoserver [2012-07-31 15:35:15] <jgarnett> trying to search on this end but i expect you are ahead of me [2012-07-31 15:36:21] -->| matthijsln ([email protected]) has joined #geoserver [2012-07-31 15:36:46] <bencaradocdavies> find . -name pom.xml -exec grep -H -B2 -A2 "artifactId>ant<" {} \; [2012-07-31 15:36:51] <bencaradocdavies> Yep [2012-07-31 15:36:58] <jgarnett> bencaradocdavies most of the time it looks like maven-antrun-plugin is used; without saying what version of ant [2012-07-31 15:37:14] <bencaradocdavies> there is a dep set at the toplevel [2012-07-31 15:37:38] <jgarnett> ah okay [2012-07-31 15:40:02] <jgarnett> so I have swapped those over [2012-07-31 15:40:03] <bencaradocdavies> lots of them are groupId fixes [2012-07-31 15:40:09] <jgarnett> and can start a fresh build [2012-07-31 15:40:10] <bencaradocdavies> hacking now [2012-07-31 15:40:15] <jgarnett> I only found 3 to fix [2012-07-31 15:40:30] <jgarnett> is that "a lot" ? [2012-07-31 15:40:59] <jgarnett> searching for "groupId>ant<" now to see if there is a lot [2012-07-31 15:41:16] <bencaradocdavies> five [2012-07-31 15:42:25] <jgarnett> I got an [2012-07-31 15:42:26] <jgarnett> <dependency> [2012-07-31 15:42:27] <jgarnett> <groupId>ant</groupId> [2012-07-31 15:42:27] <bencaradocdavies> jgarnett, shall we bump ant to 1.8.4 while we are at it? Are you feeling luck> :-) [2012-07-31 15:42:28] <jgarnett> <artifactId>ant-optional</artifactId> [2012-07-31 15:42:29] <jgarnett> <version>1.5.3-1</version> [2012-07-31 15:42:29] <jgarnett> </dependency> [2012-07-31 15:42:35] <bencaradocdavies> ouch [2012-07-31 15:42:51] <bencaradocdavies> I found an ant-launcher in wps-core [2012-07-31 15:42:52] <jgarnett> sure; although I think we should consider sharing a "diff" so we test the same thing [2012-07-31 15:43:15] <jgarnett> http://mvnrepository.com/artifact/ant/ant-optional [2012-07-31 15:43:16] <sigq> Title: Maven Repository: ant ant-optional (at mvnrepository.com) [2012-07-31 15:43:24] <jgarnett> 1.5.3-1 is the "latest" [2012-07-31 15:43:38] <jgarnett> yeah I found the wps one as well [2012-07-31 15:44:14] <bencaradocdavies> For which groupId: http://search.maven.org/#search|ga|1|ant-optional [2012-07-31 15:44:15] <sigq> Title: The Central Repository Search Engine (at search.maven.org) [2012-07-31 15:45:06] <jgarnett> fair point ben [2012-07-31 15:46:04] <jgarnett> 1.5.3-1 is later then 1.5.4 [2012-07-31 15:46:23] <jgarnett> Jan 2007, but perhaps that is due to a migration of structure or something [2012-07-31 15:46:39] <jgarnett> ben I am tempted to build here to see if it goes [2012-07-31 15:47:15] <jgarnett> I can issue a pull request from there or something; at least enough to start discussion [2012-07-31 15:47:20] <jgarnett> (sorry for the trouble) [2012-07-31 15:48:11] <bencaradocdavies> OK, I think we leave that one alone as it is separate from org.apache.ant [2012-07-31 15:48:27] <jgarnett> yep [2012-07-31 15:48:29] <bencaradocdavies> [sorry, Simon C was in my office] [2012-07-31 15:49:29] <bencaradocdavies> ant-contrib is safe and latest [2012-07-31 15:54:21] <bencaradocdavies> jgarnett, I am testing a skipTests build with this: http://pastebin.com/raw.php?i=NPJdWZ13 [2012-07-31 15:59:05] <bencaradocdavies> jgarnett, install passed, checking assembly:attached (as I changed the release pom) [2012-07-31 16:00:27] <bencaradocdavies> jgarnett, release worked as well [2012-07-31 16:00:55] <bencaradocdavies> and the VERSION.txt changes are confirmed to work [2012-07-31 16:01:10] <jgarnett> I am trying out http://pastebin.com/3h9VASfG [2012-07-31 16:01:15] <sigq> Title: geoserver fix for ant groupId - Pastebin.com (at pastebin.com) [2012-07-31 16:01:33] <bencaradocdavies> aaime, you might want to have a look as well [2012-07-31 16:01:43] <aaime> uh= [2012-07-31 16:02:07] <aaime> in a meeting, what's up? [2012-07-31 16:02:12] <bencaradocdavies> TL;DR: VERSION.txt fix broke ant dependencies [2012-07-31 16:02:39] <bencaradocdavies> Those with ant:ant:jar:1.8.2 in their local repos can build geoserver [2012-07-31 16:02:59] <bencaradocdavies> otherwise, tough luck as the groupId is supposed to be or.apache.ant [2012-07-31 16:03:18] <aaime> seems like an easy fix? [2012-07-31 16:03:23] <jgarnett> checking diffs - ben you jumped to 1.8.4, and found "ant-launcher" to correct [2012-07-31 16:03:35] <bencaradocdavies> org.apache.ant:ant:jar:1.8.2 is the only one in central [2012-07-31 16:03:53] <bencaradocdavies> Yes, 1.8.4 is latest, and ant-launcher appear to be in sync with ant [2012-07-31 16:03:53] <jgarnett> so your fix looks better than mine +1 [2012-07-31 16:04:39] <bencaradocdavies> Thanks, your search was better than mine. Dropping the last < gave me: find . -name pom.xml -exec grep -H -B2 -A2 "artifactId>ant" {} \; [2012-07-31 16:04:51] <bencaradocdavies> And I say the ant-contrib and so on. [2012-07-31 16:04:56] <bencaradocdavies> ^saw [2012-07-31 16:05:23] <jgarnett> so good team work, if your full build works I say push it up [2012-07-31 16:05:28] <bencaradocdavies> So, i am ready to commit. I have not run tests. But we have jenkins and hudson for that. :-) [2012-07-31 16:06:16] <jgarnett> okie dokie [2012-07-31 16:06:22] <jgarnett> let me know and I will restart my run on this side [2012-07-31 16:06:25] <bencaradocdavies> OK, jgarnett. I will run a full build and then commit if it passes. [2012-07-31 16:06:34] <bencaradocdavies> Or shall I just commit it as it is? [2012-07-31 16:06:49] <bencaradocdavies> The ant stuff is mainly packaging, not core tests, AFAIK. [2012-07-31 16:07:03] <jgarnett> I always try to run a full build; but that also keeps me from participating often [2012-07-31 16:07:20] <bencaradocdavies> By the time I have crafted a pithy log message, the tests will be done. ;-) [2012-07-31 16:07:45] <jgarnett> you must have a fast machine [2012-07-31 16:08:48] <jgarnett> okay I have stopped by build, and await your changes bencaradocdavies [2012-07-31 16:09:14] <bencaradocdavies> No, I am just a slow typist and agonising grammarian. [2012-07-31 16:11:17] <jgarnett> see I could never spell words like "agonising grammarian" so I am faster [2012-07-31 16:23:28] <bencaradocdavies> jgarnett, got tired waiting for the tests so I committed in 8e51ce269eaab07cfa04b700e2dcb909133675ed to master. [2012-07-31 16:25:31] <jgarnett> fetching now
------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________ Geoserver-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/geoserver-devel
