From Anthony.Petrov at Sun.COM Wed Apr 1 05:47:04 2009 From: Anthony.Petrov at Sun.COM (Anthony Petrov) Date: Wed, 01 Apr 2009 16:47:04 +0400 Subject: Syndication feed available on cr.openjdk.java.net In-Reply-To: <49D107A9.8090208@sun.com> References: <49D107A9.8090208@sun.com> Message-ID: <49D36248.4080204@sun.com> The feed does not work with the Opera web-browser. I tested Opera 9.64 on both Linux and Windows: when clicking the RSS orange icon in the address bar, the browser presents the .xml content of the feed as a plain text instead of subscribing to the feed. I double checked, other feeds - like from the blogspot.com - work correctly. Firefox 3, however, reads the cr.openjdk feed correctly. I guess there's either a problem with the http-server settings so that it reports an incorrect Content-Type: for the feed, or the generated .xml for that feed is slightly invalid/non-quite-standards-conformant. Could anyone take a look at the issue please? -- best regards, Anthony On 03/30/2009 09:55 PM Tim Bell wrote: > Ongoing updates of postings to cr.openjdk.java.net are now visible via: > > http://cr.openjdk.java.net/feed.atom > > Feel free to subscribe to this stream it you would like to follow it. > > To those publishing webrevs or other materials on cr.openjdk.java.net- > Please refer to the updated example on the main page: > http://cr.openjdk.java.net > for how to customize the title and description fields on your feed > entries. Default values will be used if necessary, so this step is > optional and not required. > > Feedback is welcomed on web-discuss(at)openjdk.java.net. > > Best Regards- > > Tim > > From Tim.Bell at Sun.COM Wed Apr 1 10:16:17 2009 From: Tim.Bell at Sun.COM (Tim Bell) Date: Wed, 01 Apr 2009 10:16:17 -0700 Subject: Syndication feed available on cr.openjdk.java.net In-Reply-To: <49D36248.4080204@sun.com> References: <49D107A9.8090208@sun.com> <49D36248.4080204@sun.com> Message-ID: <49D3A161.8020508@sun.com> Anthony Petrov wrote: > The feed does not work with the Opera web-browser. I tested Opera 9.64 > on both Linux and Windows: when clicking the RSS orange icon in the > address bar, the browser presents the .xml content of the feed as a > plain text instead of subscribing to the feed. I double checked, other > feeds - like from the blogspot.com - work correctly. Firefox 3, however, > reads the cr.openjdk feed correctly. > > I guess there's either a problem with the http-server settings so that > it reports an incorrect Content-Type: for the feed, or the generated > .xml for that feed is slightly invalid/non-quite-standards-conformant. > Could anyone take a look at the issue please? Thank you for reporting the problem - and for providing the diagnosis! This was caused by a configuration change I made in the web server on cr.ojn to tag most files as "text/plain"[1]. This works well for the large variety of file extensions presented in webrevs, but as you reported is wrong for the feed. I guess Opera is checking the mimetype rather than using the data at the top of the feed, which firefox seems to do. I updated the configuration to tag the atom feed to be "text/xml". I tested this using Opera on one of our Solaris/SPARC lab servers[2] and it seems to be OK. Note that I had to exit the Opera browser and restart it before it noticed the change in mime type for that page. Please let me know if it is not working for you. Best regards- Tim Bell [1]http://mail.openjdk.java.net/pipermail/web-discuss/2009-March/000068.html [2]Yes, they build a Solaris/SPARC version, and good for them! From Anthony.Petrov at Sun.COM Wed Apr 1 10:41:12 2009 From: Anthony.Petrov at Sun.COM (Anthony Petrov) Date: Wed, 01 Apr 2009 21:41:12 +0400 Subject: Syndication feed available on cr.openjdk.java.net In-Reply-To: <49D3A161.8020508@sun.com> References: <49D107A9.8090208@sun.com> <49D36248.4080204@sun.com> <49D3A161.8020508@sun.com> Message-ID: <49D3A738.30809@sun.com> On 4/1/2009 9:16 PM Tim Bell wrote: > I updated the configuration to tag the atom feed to be "text/xml". I > tested this using Opera on one of our Solaris/SPARC lab servers[2] and > it seems to be OK. Note that I had to exit the Opera browser and > restart it before it noticed the change in mime type for that page. > > Please let me know if it is not working for you. I've just tried it, and it worked well. Thank you very much! > [2]Yes, they build a Solaris/SPARC version, and good for them! This is actually amazing! I noticed before they had the Solaris/SPARC version on their download page, but I thought it might be as outdated as, for instance, their BeOS, QNX, and OS/2 versions are. But they have the latest Opera 9.64 for Solaris/SPARC. Lovely! -- best regards, Anthony From Weijun.Wang at Sun.COM Mon Apr 13 19:24:08 2009 From: Weijun.Wang at Sun.COM (Weijun Wang) Date: Tue, 14 Apr 2009 10:24:08 +0800 Subject: Syndication feed available on cr.openjdk.java.net In-Reply-To: <49D107A9.8090208@sun.com> References: <49D107A9.8090208@sun.com> Message-ID: <49E3F3C8.5090304@sun.com> Hi Tim I notice a problem with the feed. When the first webrev (webrev.00) is created for a bug id, a new item appears in the feed. However, when webrev.01 is uploaded, nothing happens. Is this a feature or a bug? Thanks Max Tim Bell wrote: > Ongoing updates of postings to cr.openjdk.java.net are now visible via: > > http://cr.openjdk.java.net/feed.atom > > Feel free to subscribe to this stream it you would like to follow it. > > To those publishing webrevs or other materials on cr.openjdk.java.net- > Please refer to the updated example on the main page: > http://cr.openjdk.java.net > for how to customize the title and description fields on your feed > entries. Default values will be used if necessary, so this step is > optional and not required. > > Feedback is welcomed on web-discuss(at)openjdk.java.net. > > Best Regards- > > Tim > From Tim.Bell at Sun.COM Mon Apr 13 19:47:13 2009 From: Tim.Bell at Sun.COM (Tim Bell) Date: Mon, 13 Apr 2009 19:47:13 -0700 Subject: Syndication feed available on cr.openjdk.java.net In-Reply-To: <49E3F3C8.5090304@sun.com> References: <49D107A9.8090208@sun.com> <49E3F3C8.5090304@sun.com> Message-ID: <49E3F931.9070207@sun.com> Hi Max > I notice a problem with the feed. When the first webrev (webrev.00) is > created for a bug id, a new item appears in the feed. However, when > webrev.01 is uploaded, nothing happens. Is this a feature or a bug? That sounds like a bug. We want everyone to be aware of current events... and that includes updates to anything previously put out. I'm checking on it. The update job runs about every 10 minutes, and it should discover anything new or recently modified under the user directories. Tim From Weijun.Wang at Sun.COM Wed Apr 15 04:44:08 2009 From: Weijun.Wang at Sun.COM (Max (Weijun) Wang) Date: Wed, 15 Apr 2009 19:44:08 +0800 Subject: webrev wrapper for mq users Message-ID: #! /bin/bash # # Usage: qr patch_name # # Run this script inside a hg repo with mq. It will create a webrev for # the specified patch (or the tip) and sync it to cr.openjdk.java.net [ "$1" == "--help" ] && { echo Usage: qr [patch]; exit 0; } [ "$1" != "" ] && hg qgoto $1 BUGID=`hg tip --template {desc} | head -n1 | cut -c 1-7` AUTHOR=`hg tip --template {author}` LASTVER=`wget -q -O - http://cr.openjdk.java.net/~$AUTHOR/$BUGID | perl -e ' $m = -1; while () { if (/webrev\.(\d\d)/) { if ($1 > $m) { $m = $1; } } }; print $m; '` VER=`perl -e 'printf "%02d", '$LASTVER'+1'` mkdir -p $BUGID if [ $LASTVER != "-1" ]; then mkdir $BUGID/webrev.$LASTVER wget -q -O $BUGID/webrev.$LASTVER/index.html http://cr.openjdk.java.net/~$AUTHOR/$BUGID/webrev.$LASTVER/index.html perl -p -i -e 's/<\/head>/<\/head>

A newer webrev<\/a> is available<\/h3>/' $BUGID/ webrev.$LASTVER/index.html fi webrev -O -N -r -2 hg tip > $BUGID/.description hg tip --template {desc} | head -n1 > $BUGID/.title mv webrev $BUGID/webrev.$VER mv webrev.zip $BUGID/webrev.$VER.zip echo http://cr.openjdk.java.net/~$AUTHOR/$BUGID/webrev.$VER echo Press Ctrl-C to stop. Enter to continue read rsync -avz $BUGID cr.openjdk.java.net: echo http://cr.openjdk.java.net/~$AUTHOR/$BUGID/webrev.$VER rm -rf $BUGID From Bradford.Wetmore at Sun.COM Thu Apr 30 18:39:41 2009 From: Bradford.Wetmore at Sun.COM (Brad Wetmore) Date: Thu, 30 Apr 2009 18:39:41 -0700 Subject: who should use bugzilla ? In-Reply-To: <20090305044252.60BCAD06B@callebaut.niobe.net> References: <20090305044252.60BCAD06B@callebaut.niobe.net> Message-ID: <49FA52DD.1050002@sun.com> From a discussion in early March. Mark Reinhold wrote: > Until such time as more of our internal tools are externalized, and we > use our Bugzilla for all bugs rather than just for contributions from > non-Sun developers, I see no harm in using Bugzilla to track changes > being proposed by such developers even if they do have push rights into > the appropriate forests. > > If there's general agreement on this then I'll arrange for the warning > header on bugs.openjdk.java.net to be modified accordingly. I agree with the "no harm" sentiment. The original intent was for developers without push rights, but see no reason to restrict it to only those without push rights. Since there was no further discussion, I'll assume this is ok, and have updated the text accordingly. Brad P.S. Mark's full response follows: Mark Reinhold wrote: >> Date: Tue, 03 Mar 2009 13:43:15 -0800 >> From: phil.race at sun.com > >> As you can see from the email below, there's something of >> an apparent hole in the description of who should use bugzilla. >> >> Can we clarify what's supposed to be the process >> >> Specifically if you are external to Sun. but have an openjdk user id >> either because you are a member of an openjdk group (eg 2d, awt), OR >> because you have an openjdk project, then should you in fact use >> bugzilla, since you can't use the Sun internal bugster tool ? > > In the long run, people such as Roman who have push rights into the JDK 6 > and 7 forests -- and in his case have in fact already pushed changes to > at least one of those forests, after appropriate review -- should not > need to use a special process set up primarily for people who haven't yet > earned such rights. > > Having said that, Roman is completely right to point out that he can't > make effective use of bugs.sun.com on his own. There are, moreover, > other internal tools, e.g., the code-review robot, that aren't (yet) > available to external developers. > > Until such time as more of our internal tools are externalized, and we > use our Bugzilla for all bugs rather than just for contributions from > non-Sun developers, I see no harm in using Bugzilla to track changes > being proposed by such developers even if they do have push rights into > the appropriate forests. > > If there's general agreement on this then I'll arrange for the warning > header on bugs.openjdk.java.net to be modified accordingly. > > - Mark