Hi Stefan,
Concerning the failing JUnit tests, here is what I did:
·Used recent master version (commit 07e0c1e)
·Changed the target version of dpp.core and Saros/E (right-click >
"Build Path" > "Configure Build Path ..." > Tab "Libraries" > Select
"JRE System Library [JavaSE-1.6]" > "Edit ..." > Change to "JavaSE-1.7")
·Cleaned the workspace and re-build everything àno compiler errors
·Ran the SarosCoreTestSuite àall green
·Ran the SarosEclipseTestSuite à10 errors (SarosSessionManagerTest,
SarosSessionTest)
See below for the copy-pasted error message of the failing test class
SarosSessionManagerTest.
Franz
Details: java.lang.VerifyError: Bad <init> method call from inside of
a branch
Exception Details:
Location:
de/fu_berlin/inf/dpp/project/internal/SarosSession.<init>(Lde/fu_berlin/inf/dpp/net/xmpp/JID;Ljava/lang/String;Ljava/lang/String;IILde/fu_berlin/inf/dpp/ISarosContext;)V
@99: invokespecial
Reason:
Error exists in the bytecode
Bytecode:
0000000: 2a19 062b 1504 1505 2c2d 3a0d 3a0c 360b
0000010: 360a 3a09 3a08 3a07 1300 4fb8 0016 1006
0000020: bd00 0459 0319 0853 5904 1909 5359 05bb
0000030: 0051 5915 0ab7 0054 5359 06bb 0051 5915
0000040: 0bb7 0054 5359 0719 0c53 5908 190d 5313
0000050: 0056 b800 1eb8 005a 3a0f 190f b200 2aa5
0000060: 000a 2ab7 005c a700 1619 0719 0819 0915
0000070: 0a15 0b19 0c19 0db7 005e 0157 b1
Stackmap Table:
full_frame(@105,{UninitializedThis,Object[#103],Object[#75],Object[#75],Integer,Integer,Object[#77],UninitializedThis,Object[#77],Object[#103],Integer,Integer,Object[#75],Object[#75],Top,Object[#4]},{})
full_frame(@124,{Object[#2],Object[#103],Object[#75],Object[#75],Integer,Integer,Object[#77],Object[#2],Object[#77],Object[#103],Integer,Integer,Object[#75],Object[#75],Top,Object[#4]},{})
at java.lang.Class.getDeclaredMethods0(Native Method)
at java.lang.Class.privateGetDeclaredMethods(Unknown Source)
at java.lang.Class.getMethod0(Unknown Source)
at java.lang.Class.getMethod(Unknown Source)
at
org.easymock.internal.ObjectMethodsFilter.<init>(ObjectMethodsFilter.java:55)
at
org.easymock.internal.MocksControl.createMock(MocksControl.java:59)
at
org.powermock.api.easymock.PowerMock.doCreateMock(PowerMock.java:2212)
at org.powermock.api.easymock.PowerMock.doMock(PowerMock.java:2163)
at
org.powermock.api.easymock.PowerMock.createNiceMock(PowerMock.java:187)
at
de.fu_berlin.inf.dpp.project.SarosSessionManagerTest.setUp(SarosSessionManagerTest.java:97)
[...]
*From:*Stefan Rossbach [mailto:srossb...@arcor.de]
*Sent:* Thursday, October 16, 2014 2:23 PM
*To:* Zieris, Franz; dpp-devel@lists.sourceforge.net
*Subject:* Re: [DPP-Devel] Can the IntelliJ plugin target Java 7+
On 16.10.2014 14:09, Zieris, Franz wrote:
Hi there,
I would favor having the same Java support for all Saros versions,
so if we'd actually rely on Java 7 for Saros/I, Saros/E should no
longer support Java 6, too.
I see two immediate technical issues here:
(1)One change that would be necessary is in our test and build
infrastructure ("saros-build"), which currently uses Java 7 for
doing some code analysis, but Java 6 for compilation (am I right,
Arsenij?).
(2)The second one involves our source code itself. Quick test:
Changing the target version for the Core and Saros/E currently
leads to no compilation errors, but some Saros/E JUnit tests fail.
This needs to be investigated first.
Which JUnit tests are failing ? I am running Java JDK 1.7 and get no
failures when running a local regression ?
I have no idea how many Saros (actual or potential) users out
there still rely on Java 6.
Does anyone have any numbers that could suit as a proxy?
Stefan: Do our anonymous statistics contain the used Java version?
Yes they do, only spotted Java 1.7 and 1.8 so far.
But the question is: In which module should Grahams changes be
implemented ? If it is only IntelliJ then their is no problem to use
Java 1.7 for IntelliJ and
Java 1.6 for Eclipse.
Franz
*From:*Graham Allan [mailto:grundlefl...@gmail.com]
*Sent:* Wednesday, October 15, 2014 7:37 PM
*To:* dpp-devel@lists.sourceforge.net
<mailto:dpp-devel@lists.sourceforge.net>
*Subject:* [DPP-Devel] Can the IntelliJ plugin target Java 7+
Hi all,
Is there any reason that the IntelliJ plugin should not target
Java 7+ as the runtime environment?
The specific motivation for me is to be able to use Java NIO's
glob matching, in order to evaluate the rules specified in a
.gitignore file. This was previously fulfilled by JGit, but having
that dependency was seen as undesirable, I think due to file size
concerns. Not having to reimplement glob matching (the equivalent
of the POSIX fnmatch function) would save a lot of effort and
maintenance.
Another approach would be to pull out the part of JGit which did
this, which targeted Java 5 I believe, and include this source
code in Saros. That code is under the Eclipse Public License, I'm
not 100% sure on the legality of that, or the impact of including
something non-GPL in the Saros distributable, but I think that's
valid.
Cheers,
Graham
------------------------------------------------------------------------------
Comprehensive Server Monitoring with Site24x7.
Monitor 10 servers for $9/Month.
Get alerted through email, SMS, voice calls or mobile push notifications.
Take corrective actions from your mobile device.
http://p.sf.net/sfu/Zoho
_______________________________________________
DPP-Devel mailing list
DPP-Devel@lists.sourceforge.net <mailto:DPP-Devel@lists.sourceforge.net>
https://lists.sourceforge.net/lists/listinfo/dpp-devel
