/builds/jakarta-ant?
Sorry about that... should be fixed now.
- Sam Ruby
Test of new mailing lists address
- Sam Ruby
Stefan Bodewig wrote:
On 28 Jan 2003, <[EMAIL PROTECTED]> wrote:
Change distribution name
Sam may need to know to adapt his nightly publishing script. dist-ant
will now result in apache-ant* instead of jakarta-ant*.
Fixed. Should http://cvs.apache.org/builds/jakarta-ant/ be renamed?
- Sa
Conor MacNeill wrote:
I've changed the AntClassLoader2 code to use a JarEntry now. Can you
rerun and let me know if that fixes the MX4J build
BUILD SUCCESSFUL.
For xml-rpc too.
Thanks!
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-ma
Fails today with a class cast exception:
http://cvs.apache.org/builds/gump/2003-01-26/mx4j.html
Yet it worked yesterday:
http://cvs.apache.org/builds/gump/2003-01-25/mx4j.html
And there were no changes to mx4j:
http://cvs.apache.org/builds/gump/2003-01-26/cvs_mx4j.html
- Sam Ruby
--
To unsubscribe
FYI.
In related news, jakarta-gump's cvs is open to all ASF committers.
*Every* ASF committer has karma.
http://jakarta.apache.org/gump/
http://cvs.apache.org/viewcvs/jakarta-gump/
http://jakarta.apache.org/site/mail2.html#Gump
Original Message
Subject: gump added to Gmane
Date:
http://cvs.apache.org/viewcvs.cgi/jakarta-ant/src/main/org/apache/tools/ant/taskdefs/Classloader.java.diff?r1=1.1&r2=1.2&diff_format=h
causes
http://cvs.apache.org/builds/gump/2002-12-25/xml-apis.html
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For addition
Conor MacNeill wrote:
Sam Ruby wrote:
I'm particularly interested in seeing that the separate instances of
Ant inherit the build.sysclasspath property. ;-)
AKA the build.sam.ruby property :-)
AKA "the dirty nasty hack".
http://jakarta.apache.org/gump/why.html#2000
;-)
-
Conor MacNeill wrote:
Sam Ruby wrote:
I would like to specify the version property for xdoclet builds within
gump, but apparently the value I specify is overridden. Any
suggestions would be appreciated...
Wow, their build is complex. I did a run with the version property
override and it
I would like to specify the version property for xdoclet builds within
gump, but apparently the value I specify is overridden. Any suggestions
would be appreciated...
- - -
http://cvs.apache.org/builds/gump/2002-12-18/module_xdoclet.html
- - -
http://cvs.sourceforge.net/cgi-bin
Are there plans to fix this?
Original Message
Subject: Re: [Xdoclet-devel] [GUMP] Build Failure - xdoclet
Date: Sun, 15 Dec 2002 14:52:23 +1100
From: Conor MacNeill <[EMAIL PROTECTED]>
Reply-To: "Alexandria Developers List"
CC: Aslak Hellesøy <[EMAIL PROTECTED]>,
xdoclet-devel@li
erride setChecked(boolean) in
org.apache.tools.ant.types.DataType
[javac] 2 errors
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
in from time to time anyway).
- Sam Ruby, Ant Committer, Board member, in that order.
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
apache.org/site/decisions.html
http://jakarta.apache.org/site/pmc/01-01-17-meeting-minutes.html
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
Sam Ruby wrote:
Yet another build that continues to fail due to this change.
http://jakarta.apache.org/builds/gump/2002-07-18/xml-axis.html
The behavior is like the nested element is not processed on
line 98 of
http://cvs.apache.org/viewcvs.cgi/xml-axis/java/test/wsdl/Wsdl2javaTestSuite.xml
Yet another build that continues to fail due to this change.
http://jakarta.apache.org/builds/gump/2002-07-18/xml-axis.html
The behavior is like the nested element is not processed on
line 98 of
http://cvs.apache.org/viewcvs.cgi/xml-axis/java/test/wsdl/Wsdl2javaTestSuite.xml?annotate=1.113
bode
FYI: this commit caused a number of Gump builds to fail. If I back this
change out, they succeed again.
See http://jakarta.apache.org/builds/gump/2002-07-17/ for details.
bodewig 2002/07/16 06:06:46
Modified:.WHATSNEW build.xml
src/etc/testcases/taskdefs ant.xml
Steve Loughran wrote:
>
> That leaves the timeout: Sam, what should we do about that?
My twenty minute per step limit is admittedly arbitrary, and perhaps I
could be talked into changing this.
An obvious workaround is to split this into multiple steps.
- Sam Ruby
--
To unsubscribe,
Steve Loughran wrote:
>
> I think maybe GUMP is using 1.6alpha, not 1.5beta; correct Sam?
It is based on CVS HEAD, whatever that might be at the moment.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
It's BaaAAAaack!
http://gump.covalent.net/log/index.html
- Sam Ruby
-- Forwarded by Sam Ruby/Raleigh/IBM on 04/24/2002
05:57 PM ---
Sam Ruby/Raleigh/[EMAIL PROTECTED] on 04/20/2002 10:13:11 PM
Please respond to "Ant Developers Lis
Apparently, RegexpMatcher is now required to build jakarta-ant, but it is
in optional.jar. The error message produced when this class can not be
found is somewhat misleading:
No JAXP compliant XML parser found. Please visit http://xml.apache.org
for a suitable parser
- Sam Ruby
--
To
Steve Loughran wrote:
>
> Sam Ruby was actually in the audience, which is something I only
> found out afterwards. Good thing I didnt say anything bad about the
> Gump.
Nah... just don't dis Ant. ;-)
Not everybody here remembers or realizes this, but for an extended peri
Costin Manolache wrote:
>
> or any of the things Sam got with gump and jdk1.4
Still do. http://nagoya.apache.org/~rubys/gump/java14/index.html
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
w of ignores those deprecation
warnings, which sets a bad precident as once you get into the habit of
ignoring some warnings you miss the really important ones. Finally, the
code that supports the attribute is pretty much self contained.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EM
build tools
of them all: javac?
Or are you like most of us and when you install a new patch level of the
JDK you pretty much do it globally across all the projects that you deal
with?
I find the line is different for everyone, but over the past few years I
have noticed a trend whereby most pe
is no known project using that item, and it was marked deprecated,
and the removal is announced here and nobody can identify an instance where
it is being used, then proceed.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
o know is that "86" first appeared as "kitchen
slang" meaning "out of that item" in the 1930's, and fairly quickly came
to mean "stop serving that customer" as well. Eventually, "86" spread to
general usage, where it came to mean simply
Stefan Bodewig wrote:
>
> +1 - if you explain the "86" to a poor German who doesn't get it ;-)
To "86" something means to remove it from the menu. While the origin of
the term is unclear, many attribute it to restaurants where it became known
as "the item on
ant to get rid of and nobody we care
about () uses it...
Oh, and watch the next interim gump run, just in case. ;-)
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
Magesh: I must say that I am impressed by how much you are keeping on top
of this...
;-)
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
...
- Sam Ruby
-- Forwarded by Sam Ruby/Raleigh/IBM on 01/28/2002
08:24 AM ---
Craig McClanahan <[EMAIL PROTECTED]> on 01/28/2002 07:41:39 AM
Please respond to "Tomcat Developers List"
To:tomcat-dev@jakarta.apache.org
cc:
Subject:
Peter Donald wrote:
>
> >http://jakarta.apache.org/builds/gump/2002-01-27/
>
> Isn't yellow a pretty color ? ;)
Errr, um... not in this case, no. ;-)
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
TH=$CLASSPATH:/opt/jaxp-1.1/jaxp.jar
export CLASSPATH=$CLASSPATH:/opt/jaxp-1.1/crimson.jar
eval "./bootstrap.sh $OUT 2>&1"
test $? -ge 1 && export STATUS="FAILED"
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
Eric Hatcher wrote:
>
> Why does the Status column say SUCCESS for bootstrap-ant when BUILD FAILED?
Because the bootstrap.sh shell script returns a successful return code. It
would be nice if this were fixed... (hint, hint) ;-)
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL
Stephane Bailliez wrote:
>
> oops, fixing it
FYI: look what happens when such errors are present:
http://jakarta.apache.org/builds/gump/2002-01-27/
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
asking before and not hearing back an answer...
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
know and I will be
glad to set it up.
My guess is that a proposal with a cleaner design and a fair measure of
backwards compatibility would be a slam dunk.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
manner outlined above. In
other words, instead of a "which codebase do we base v2 on", pick a single
code base and decide whether or not the community feels that it is ready.
Repeat until done. Note that the existence of an alternative which is
actively being developed is likely to hav
But I like it:
http://www.letsbuyit.com/images/flagpageant.gif
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
, that is. ;-)
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
Stefan Bodewig wrote:
>
> That would be a non-trivial piece of code ... If you have one, I
> think Sam could install it.
I wouldn't dare! I've see too much code eaten by well-intentioned post
processing tools gone astray.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:
Peter Donald wrote:
>
> Is this something that is
> automativally done on a nightly basis or something (or could be made to do
> so?).
I certainly am interested in helping in any way I can to make something
like this happen.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[
rs/latest/xml-commons/which.jar
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
dience, and causes
non-developers to flee to the -user list. ;-)
I can't help but wonder if the real root cause is that the people
interested in following myrmidon is distinct from the constituency for ant.
If so, a separate cvs and -cvs mailing list for that code base may be more
to the point.
cant performance improvement.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
Stefan Bodewig wrote:
>
> I guess my response to your commit message hasn't made it to
> fop-dev yet.
I've manually approved two more messages. Anyone on the fop-dev team want
to volunteer to become moderator? My guess is that the current moderator
has become backlogged for some reason.
> There
of the build.xml, I get one message:
Trying to override old definition of task manifest
Then everything proceeds fine - apparently the xml-fop version of the task
is used.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
satisifies all requirements?
http://cvs.apache.org/viewcvs/jakarta-ant/src/main/org/apache/tools/ant/taskdefs/Manifest.java
http://cvs.apache.org/viewcvs/xml-fop/src/org/apache/fop/tools/anttasks/Manifest.java
- Sam Ruby
-- Forwarded by Sam Ruby/Raleigh/IBM on 12/13/2001
09:00 AM
.
I'll make the fixes to stylebook.
> So, no matter what we do (if we do something), it is going to cause
> red and yellow lines in Gump.
Hopefully, that will get some attention. ;-)
My plan would be to *not* specify -legacy for any Apache project in any
gump run. I may weaken that posi
Ant 1.5 appears.
I guess that depends on how you count releases. I'd recommend a more
gradual approach be taken initially. There were plenty of things
deprecated in 1.3 - use those to see how the user community reacts to a
more aggressive approach to deprecation than has been taken la
ne public release. Taking the jar
attribute name example, if making it a legacy option were done at the same
time, the I would be routinely setting the flag, and thereby eliminating
any incremental benefit that there would exist over and above the current
deprecation policy.
- Sam Ruby
--
To uns
major new release, I can make Gump do experimental runs on
a daily basis. I do that today with Xerces2 and JDK 1.4.
Yes, it is like herding cats, but Gump does effectively represent
coordination... results are posted publically, nag messages are sent,
and patches are made.
- Sam Ruby
Curt Arnold wrote:
>
> There do not appear to be any nightly builds for the last several
> nights.
It looks like the distribution directory was restructured. I need to
update my scripts.
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional c
the change to Ant that broke Cornerstone's build:
http://cvs.apache.org/viewcvs/jakarta-ant/src/main/org/apache/tools/ant/taskdefs/Ant.java.diff?r1=1.28&r2=1.29
- Sam Ruby
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
lean so that the next
call to build jakarta ant would rebuild everything.
I'll take a look into using ANT_OPTS, but I may choose to add support for
$* (sh) or %1 %2 %3 %4 %5 %6 %7 %8 %9 (bat) instead.
- Sam Ruby
[java] #http://xml.apache.org/xalan-j/faq.html
>
> [junit] javax/xml/parsers/ParserConfigurationException
> [junit] java.lang.NoClassDefFoundError:
> javax/xml/parsers/ParserConfigurationException
Sorry about the false alarm. Sometimes the build failure reports are my fault
too. ;-)
- Sam Ruby
Peter Donald wrote:
>
> It is already possible now ! ;)
Care to enlighten the xml-cocoon2 folks? What I would love to see is the
following file go away:
http://cvs.apache.org/viewcvs.cgi/xml-cocoon2/bin/anttasks/
- Sam Ruby
gn.
The root problem is (was?) that tasks objects are created and then executed
one by one. It is hard to create a strongly typed task object when the
task has not been defined just yet.
- Sam Ruby
7;t know if it can be addressed at the same time, but as more projects
define tasks for their own use, one thing to consider is adding the ability
to define tasks which are built during the build process. Perhaps an
antlib task could allow nested javac and jar elements. The javac tasks
would simply be executed in order. The jar tasks would not only execute
but the results would be added to the path.
- Sam Ruby
tform eol when not specified) ;)
I'm not sure I understand this requirement... I don't see anything
resembling this in any task except fixcrlf. IMHO, one is not likely to be
modifying a .cvspass file intended for another platform using this task.
- Sam Ruby
In my
case, I potentially will be updating multiple passwords at once, making it
worthwhile for me to create a hashtable of entries in order to avoid
repetitive scans.
- Sam Ruby
() > now) {
> task.log("Warning: "+files[i]+" modified in the future.",
>Project.MSG_WARN);
http://nagoya.apache.org/~rubys/gump/dist-ant.html
Approximately 9 hours left until the "official" Gump run. ;-)
- Sam Ruby
sks are too
dang hard to write (IMHO), so people don't. This is something that should
be corrected (IMHO) in future versions of Ant, not propagated.
- Sam Ruby
Erik Hatcher wrote:
>
> JavaScript does not itself have a way to print that I know of - it relies on
> provided objects to do so ('alert' in web browsers to pop up a dialog box).
> I believe (and please correct me if I'm wrong) that to do a
> System.out.println from JavaScript you'd have to get a h
r).
Decloaking... ;-)
Might as well deprecate the Script task now. Or do you plan to limit Java
implementation of tasks similarly?
- Sam Ruby
Peter Donald wrote:
>
> BTW who is the moderator for this group?
[EMAIL PROTECTED]
[EMAIL PROTECTED]
- Sam Ruby
.org. Cvs is done through
icarus.apache.org.
jakarta.apache.org is an alias to daedalus, as is dev, xml, www, and
several others.
cvs.apache.org is an alias to icarus.
- Sam Ruby
arta-site2\docs\index.html
- Sam Ruby
I found this humorous, and thought other subscribers to this mailing list
might also:
http://www.segfault.org/story.phtml?mode=2&id=3ba7eadc-03cdf340
- Sam Ruby
For reasons I don't understand, the following commit is causing the Expand
task (aka unzip) to produce files of zero length. This ultimately produces
the errors that you can see at
http://nagoya.apache.org/~rubys/gump/rhino.html .
I've reproduced on Solaris, Linux, and Win2K.
- Sam R
ave to move out a few libraries for a 1.1 build.
>
> Not a priority at the moment.
I'll change the project description from to
.
- Sam Ruby
are welcome.
- Sam Ruby
Stephane Bailliez wrote:
>
>
I'll try to debug when I get back on Thursday (assuming it hasn't been
fixed by then).
- Sam Ruby
home/rubys/jakarta/jakarta-oro/jakarta-oro-gump/jakarta-oro-gump.jar
java org.apache.tools.ant.Main -Dbuild.sysclasspath=only run-tests
- Sam Ruby
accumulate problems here.
This passed for me yesterday...
- Sam Ruby
Stephane Bailliez wrote:
>
> Sam, can you please add the following line to the gump jakarta-ant
> descriptor (in test-ant)?
>
> diff -r1.15 jakarta-ant.xml
> 54a55,56
> >
> >
That fixed it. Thanks!
- Sam Ruby
While researching an unrelated build failure, I noticed something odd.
Search for [null] in
http://jakarta.apache.org/builds/gump/2001-08-18/xml-fop.html . Apparently
it has been this way for quite some time.
- Sam Ruby
I'll start a special gump run - that way, if there are any problems, you
can get a second chance! ;-)
It will overlay today's status. I'll probably temporarily disable nagging
- so you need to check on it if you are interested.
- Sam Ruby
should already be there. I'm aware of two
radically different versions of antlr, so I'll hold off on that for the
moment.
- Sam Ruby
or compat to compile, but
not to run.
- Sam Ruby
It should be a straightforward change. I don't honestly care if it makes
1.4 or is post 1.4 (I run from CVS anyway... ;-)).
Xalan1 is essentially retired at this point...the last release of that was
a year ago, and the code has even been removed from cvs.
- Sam Ruby
Paul Hammant wrote:
>
>> Do you know whether BSF supports context classloaders?
>
>Not sure ... Sam Ruby - do you know the answer to this?
What's a "context" class loader?
BSF has no explicit support for class loaders,
There's much I don't know about c
> Fix bootstrap problems.
Thanks. I've cvs update'd jakarta-ant on the gump machine and kicked off a
build (no other updates). Once complete, I'll post the results (this will
have the result of overlaying today's info).
- Sam Ruby
re, after all,
volunteers.
There's no magic here. You give up, and undoutably somebody else will fill
in the gap. You persist, and you will have a direct say in the evolution
of Ant. You persist long enough, then you to can be the subject of
countless e-mails commenting on your lack of responsiveness. ;-)
- Sam Ruby
> Sorry, I'd gone to bed.
No problem...still another 7.5+ hours before the "official" run.
> Anyway it looks good.
Yup!
- Sam Ruby
ong with a test run, check
http://nagoya.apache.org/~rubys/gump/ in a bit. I'm only updating
jakarta-ant and rerunning the build.
- Sam Ruby
rta-tomcat/src/build/manifests/manifest.facade22
http://cvs.apache.org/viewcvs/~checkout~/jakarta-commons/httpclient/src/conf/MANIFEST.MF
- Sam Ruby
o grow,
particularly if now a classloader is recommended...
Perhaps we could come up with a single set of common code which can be used
by multiple projects?
My only concern is the additional effort which will be required to
bootstrap Ant, but I'm confident that's solvable.
- Sam Ruby
ad the compiler itself in its own class
> loader, so that you can throw it all away afterwards. That's the
approach
> we'll take when modifying Jasper.
Would there be a value in a commons-javac? Several Apache projects seem to
have need for this function...
- Sam Ruby
FYI. Bye bye "classic" compiler.
Ant looks safe (it uses reflection), but Jasper looks like it will be
affected.
- Sam Ruby
-- Forwarded by Sam Ruby/Raleigh/IBM on 07/20/2001
07:30 AM ---
Davanum Srinivas <[EMAIL PROTECTED]> on 07/
2001/07/16 20:26:01
@@ -228,9 +228,9 @@
-
-
-
+
+
+
- Sam Ruby
nd
Solaris.
The build of xml-batik fails on all three platforms. Adding a slash as I
indicated in my previous e-mail makes it work again.
- Sam Ruby
lures.
In any case, if the exclude is changed to (and a similar change is made for jacl and
rhino), then these files are excluded as expected with the current (from
CVS) version of Ant.
- Sam Ruby
Conor MacNeill wrote:
>
> Sam, I'd like to use the nightly Gump runs as a testing tool for these
> changes. Let me know if we need to co-ordinate.
That's what it is there for...
If you need any extra support from me, just let me know.
- Sam Ruby
Can I get somebody to look at
http://jakarta.apache.org/builds/gump/2001-07-06/xml-fop.html? It worked
on 07-05, and on 07-06, xalan's check out failed...
- Sam Ruby
Ant team:
Did the logic for resolving the path to the style attribute referenced
in the
arious callXXX methods essentially works out to
be an antcall on the value specified by XXXtarget.
- Sam Ruby
Stefan Bodewig wrote:
>
> Could somebody with moderator power for ant-dev and ant-user please
> throw [EMAIL PROTECTED] from these lists, or at least
> disallow posting?
I've unsubscribed [EMAIL PROTECTED] from both lists.
In general, send these types of requests to [EMAIL PROTECTED]
- Sam Ruby
;alt="Jakarta ORO"" with either a hyphen or
Looking closer, it appears that Ant passes the arguments to the JDK as an
array, so the bug is actually in the JDK :-(.
- Sam Ruby
"http://jakarta.apache.org""";>Jakarta'
I'm looking for suggestions on how best to approach this problem.
- Sam Ruby
1 - 100 of 227 matches
Mail list logo
