Well, hmm. Nothing at all wrong with this change, but I'd like to
discuss the pattern.
So, the way alternative languages are implemented, is that if I am only
concerned with *one* language, *all* languages have to be loaded in
memory. Wouldn't it make more sense to split all these files into
Inline.
On 05/03/2015 06:25 AM, adri...@apache.org wrote:
Author: adrianc
Date: Sun May 3 11:25:08 2015
New Revision: 1677386
URL: http://svn.apache.org/r1677386
Log:
Trivial code cleanup in ObjectType.java - no functional change.
Modified:
ofbiz/trunk/framework/base/src/org/ofbiz/base/u
On 05/04/2015 08:36 AM, jaco...@apache.org wrote:
Author: jacopoc
Date: Mon May 4 13:36:09 2015
New Revision: 1677597
URL: http://svn.apache.org/r1677597
Log:
Removed commons-cli jar that was not used: barcode4j has an optional dependency
on it but only if used from the command line (which is
On 04/29/2015 01:10 AM, adri...@apache.org wrote:
Author: adrianc
Date: Wed Apr 29 06:10:01 2015
New Revision: 1676674
URL: http://svn.apache.org/r1676674
Log:
Fixed a bug in UtilMisc.java where the List returned by the availableLocales()
method included an empty Locale - causing errors in Fre
http://docs.oracle.com/javase/7/docs/api/java/util/Locale.html#getDisplayName()
==
depending on which fields are specified in the locale. If the language,
sacript, country, and variant fields are all empty, this function
returns the empty string.
==
I'd really like to know what locale definit
Ok, let's try this again.
==
* rm -rf ~/.m2
* mvn clean
* rm -rf ~/.m2
* mvn clean
* mvn package -DskipTests
* rm -rf ~/.m2
* mvn package -DskipTests
* rm -rf ~/.m2
* mvn install -DskipTests
==
These are the scenarios I have tested. The final step copies all the
artifacts into ~/.m2/repository
With this commit, all optionally detected libraries/features during the
compilation with ant are now being done with maven. A "maven clean" and
"maven package -DskipTests" is runnable with tools/startofbiz.sh.
I have not run system integration tests(aka, any test that requires a
running ofbiz
Replying to my own commit; please, double check your changes. The
problems I fixed below are just sad.
ps: I'm working on maven profiles, that detect optional jars. I went
and fetched payflow.jar, and when I made it available, I got compile errors.
On 04/27/2015 10:25 AM, doo...@apache.org
This is an example of why having a HUGE commit is wrong. This problem
listed below was not discovered, because there were too many changes to
review. My maven branch at OFBIZ-6271 is doing things in a much better way.
ps: I discovered this because I am attempting to compile each and every
co
On 04/23/2015 10:13 PM, Adam Heath wrote:
Replying to my own commit; this is a major milestone, as mentioned in
the commit log.
Self replying again. Note, that everything that is in this branch is
only a week old. I have only been working with maven directly for that
time. Just *1
Replying to my own commit; this is a major milestone, as mentioned in
the commit log.
Internal and external components can still make use of the wildcard
includes on the lib folders; there is no current example in this branch
that shows this, however.
Once all this conversion is done, the to
I'm waiting on my local git(svn fetch) to finish working, so it might be
a bit before anything new lands in this branch.
On 04/20/2015 06:02 PM, doo...@apache.org wrote:
Author: doogie
Date: Mon Apr 20 23:02:47 2015
New Revision: 1675001
URL: http://svn.apache.org/r1675001
Log:
Begin work for
From a single linux-based machine, I can now do:
../../ant [options] [targets]
or
wine cmd /c ../../ant [options] [targets]
The later is used to test underneath wine-i386(aka, 32-bit).
On 06/26/2014 11:15 AM, doo...@apache.org wrote:
Author: doogie
Date: Thu Jun 26 16:15:03 2014
New Revision
t. But I don't know how to make this available. Any
suggestions?
On 06/25/2014 04:53 PM, Adam Heath wrote:
Nicely done. But now cobertura 1.9.4.1 produces byte-code that fails
to verify under 1.7. :(
Don't revert this change, obviously; this is a bug in cobertura.
However, I tried
Nicely done. But now cobertura 1.9.4.1 produces byte-code that fails to
verify under 1.7. :(
Don't revert this change, obviously; this is a bug in cobertura.
However, I tried to update it yesterday, but it pulled in a whole bunch
of other deps, so I might need to play some more.
On 06/25/20
There was a taskdef for sonar, but not for ivy. So, it was pure luck
that it was actually working. The ivy installation docs recommend
manually creating the taskdef.
The issue that made me change all this, was that when the ant lib folder
is the same as the classpath being used at build time
David E Jones wrote:
>
> Not sure if I like this... My preference would be that each build file
> only owns its directory and below. This change makes a lower level build
> file manage something higher level.
>
> What if the framework or base directories move, or we want more build
> flexibility?
17 matches
Mail list logo
