You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@shindig.apache.org by Henning Schmiedehausen <he...@schmiedehausen.org> on 2010/05/25 19:50:27 UTC
java/common/src/test/java/org/apache/shindig/common/JsonSerializerTest.java
Hi,
I am using a slightly newer json.org library that has a bunch of bug
fixes applied to it (which is a collection of various point-in-time
changes to the json.org codebase). This actually serializes the "null"
value in the JSONArray out, which in turn breaks the
serializeJsonArray() test. I was wondering whether this behaviour
would be a problem for the consumers of that class? I "fixed" this by
changing the expected reply for our internal Shindig branch but I am
not sure whether there might be another problem lurking under that.
-h
(Yes, I plan to push our changes to github)
Re: java/common/src/test/java/org/apache/shindig/common/JsonSerializerTest.java
Posted by Han Nguyen <hn...@us.ibm.com>.
Hello,
We have been working collaboratively with the WINK team to leverage the
same json lib for Shindig that is clean of "This software shall be used
for Good, not Evil" issue.
By the end of this week, the WINK team will made the code drop of the Json
lib in Apache, then my team can start to update the poms to
point to this jar + code clean up and get the patch ready for Shindig
trunk. I will open a JIRA for this later today if there's no objection.
Thanks,
Han
From: Paul Lindner <li...@inuus.com>
To: dev@shindig.apache.org
Date: 05/25/2010 02:24 PM
Subject: Re:
java/common/src/test/java/org/apache/shindig/common/JsonSerializerTest.java
Some people have privately mentioned that the json.org code base is
unacceptable due to it's unenforceable license which mentions: "This
software shall be used for Good, not Evil".
See
https://issues.apache.org/jira/browse/WINK-159
https://issues.apache.org/jira/browse/HBASE-2383
http://jira.codehaus.org/browse/JETTISON-66
If we decide we're going to change this we should make that decision now
before 2.0.0 is released.
On May 25, 2010, at 10:50 AM, Henning Schmiedehausen wrote:
> Hi,
>
> I am using a slightly newer json.org library that has a bunch of bug
> fixes applied to it (which is a collection of various point-in-time
> changes to the json.org codebase). This actually serializes the "null"
> value in the JSONArray out, which in turn breaks the
> serializeJsonArray() test. I was wondering whether this behaviour
> would be a problem for the consumers of that class? I "fixed" this by
> changing the expected reply for our internal Shindig branch but I am
> not sure whether there might be another problem lurking under that.
>
> -h
>
> (Yes, I plan to push our changes to github)
Re: java/common/src/test/java/org/apache/shindig/common/JsonSerializerTest.java
Posted by Henning Schmiedehausen <he...@schmiedehausen.org>.
Hello Paul,
I am aware of the license. If you consider sticking with the json.org
library, let me point you at http://github.com/ning/org-json which is
our fresh-off-the-press-just-opensourced fork of the library from
$DAY_JOB that is kind of usable at this point. It even has a test
suite.
We (as in Ning) tried to give our patches back to json.org but they
never answered our mails. Also, their "releases" are kind of
arbitrary, so the library most people use (the 2008... one from repo1)
has a bunch of bugs that bit us.
We moved most of our internal stuff to Jackson but json.org is still a
nice small and usable library. That said, ripping either this one or
json-simple out and using either or would be an overall win for the
code base. :-)
-h
On Tue, May 25, 2010 at 11:23, Paul Lindner <li...@inuus.com> wrote:
> Some people have privately mentioned that the json.org code base is unacceptable due to it's unenforceable license which mentions: "This software shall be used for Good, not Evil".
>
> See
> https://issues.apache.org/jira/browse/WINK-159
> https://issues.apache.org/jira/browse/HBASE-2383
> http://jira.codehaus.org/browse/JETTISON-66
>
> If we decide we're going to change this we should make that decision now before 2.0.0 is released.
>
>
> On May 25, 2010, at 10:50 AM, Henning Schmiedehausen wrote:
>
>> Hi,
>>
>> I am using a slightly newer json.org library that has a bunch of bug
>> fixes applied to it (which is a collection of various point-in-time
>> changes to the json.org codebase). This actually serializes the "null"
>> value in the JSONArray out, which in turn breaks the
>> serializeJsonArray() test. I was wondering whether this behaviour
>> would be a problem for the consumers of that class? I "fixed" this by
>> changing the expected reply for our internal Shindig branch but I am
>> not sure whether there might be another problem lurking under that.
>>
>> -h
>>
>> (Yes, I plan to push our changes to github)
>
>
Re: java/common/src/test/java/org/apache/shindig/common/JsonSerializerTest.java
Posted by Paul Lindner <li...@inuus.com>.
Some people have privately mentioned that the json.org code base is unacceptable due to it's unenforceable license which mentions: "This software shall be used for Good, not Evil".
See
https://issues.apache.org/jira/browse/WINK-159
https://issues.apache.org/jira/browse/HBASE-2383
http://jira.codehaus.org/browse/JETTISON-66
If we decide we're going to change this we should make that decision now before 2.0.0 is released.
On May 25, 2010, at 10:50 AM, Henning Schmiedehausen wrote:
> Hi,
>
> I am using a slightly newer json.org library that has a bunch of bug
> fixes applied to it (which is a collection of various point-in-time
> changes to the json.org codebase). This actually serializes the "null"
> value in the JSONArray out, which in turn breaks the
> serializeJsonArray() test. I was wondering whether this behaviour
> would be a problem for the consumers of that class? I "fixed" this by
> changing the expected reply for our internal Shindig branch but I am
> not sure whether there might be another problem lurking under that.
>
> -h
>
> (Yes, I plan to push our changes to github)