This is an error generated when compiling my project in AppHarbor.
Researched the restrictions of AppHarbor but could not find anything.
Today my application has the following structure:
Sql Server (Add-on AppHarbor)
Css and Javascritp minify by Bundle (Microsoft solutions)
3 projetos (Web->MVC4, Data->Class library, Services->Class library)
Work in Medium Trust environment
Full build report:
http://paste.ideaslabs.com/show/LHZBHhKnBC
You're trying to use Sql Server compact ... remove those dependencies and use plain ol' ado.net
Related
Using the google play games services and Admob plugin I cannot seem to get my back to build into an APK and I am receiving the following error message. Any help would be appreciated as I have removed some duplicate files but I am still failing to find why I cannot build my app.
I am using Unity Unity 5.6.1f1 (64-bit).
game services: https://github.com/playgameservices/play-games-plugin-for-unity Ads: https://github.com/unity-plugins/Firebase-Admob-Unity
CommandInvokationFailure: Unable to convert classes into dex format.
C:/Program Files/Java/jdk1.8.0_121\bin\java.exe -Xmx2048M -Dcom.android.sdkmanager.toolsdir="C:/Users/Jack/AppData/Local/Android/sdk\tools" -Dfile.encoding=UTF8 -jar "C:\Program Files\Unity\Editor\Data\PlaybackEngines\AndroidPlayer/Tools\sdktools.jar" -
stderr[
Uncaught translation error: java.lang.IllegalArgumentException: already added: Lcom/google/android/gms/internal/zzbyb;
Uncaught translation error: java.lang.IllegalArgumentException: already added: Lcom/google/android/gms/internal/zzbyc;
2 errors; aborting
]
Some Background
When Unity builds your project for Android, it invokes several tools from the Android SDK. As part of that process, it converts all of your native (Java) code for Android into a file format called DEX (Dalvik executable).
All of your Android plugins get built up into a single package using that tool. The problems start when a few plugins have the same compiled Java code (classes) in them. This will cause the DEX tool to fail with an error like the one you're seeing:
Uncaught translation error: java.lang.IllegalArgumentException:
already added: Lcom/google/android/gms/internal/zzbyb;
Uncaught translation error: java.lang.IllegalArgumentException: already added: Lcom/google/android/gms/internal/zzbyc;
This means that classes with those names were already included in another library, and so they cannot be added again.
Possible Causes
As explained above, this error occurs when you have duplicate plugins in your project; this can happen in different scenarios, here are some examples:
The same plugin is included more than once, under different folders of the project.
The same plugin is included more than once with different versions.
A plugin contains other dependencies "embedded" inside it, but these dependencies are already included in the project in some form.
How To Fix
You should look up duplicate Android plugins in your project and eliminate them (keep only 1 copy). From the error message you posted, the issue here is related to Google play services libraries. You should look into that (libraries named play-services-xxxx.aar).
Paid Help (Shameless Plug)
I provide a professional service for fixing this exact kind of issue. In case you (or anyone else) are not able to resolve such an issue themselves, feel free to contact me and get it solved.
I was facing the same problem after add admob in my project, here are 2 possible solutions:
Go to the file >> build settings and change the option internal to gradle
Open libs folders separately of both Facebook SDK and Google Play services and delete the matching files
It would be something like (support v4)
Google ads 11.2 has errors. It reuses certain jar classes .
On its own, it isnot an issue
Once you add in another google class, ie firebase, your done for.
There is a fix though. Google put out firebase 4.2 and ads 11.4
I'm building some thrift code on windows which has the following dependencies:
open ssl
thrift
Setup:
Windows 7 64-bit
VS2013
Goal:
32-bit thrift app
I compiled open ssl using these instructions (32-bit release) which produced libeay32.lib and ssleay32.lib. Next I created a VS2013 project to build the thrift library that depends on open ssl. On compilation, this project produced libthrift.lib which is (as far as I can tell) statically linked:
All good so far...
Next, I build my thrift app., again, linking statically:
However, when I run my thrift app., I get the following error that complains about incompatible image types (x86/x64):
To find the cause, I ran dependency walker on my app only to find that it's pulling in 64-bit ssl dlls (!) that came with the 32-bit (!!!) mercurial installation:
My understanding is that my application should only link against libthrift.lib which has already statically linked the open ssl stuff (above).
It seems my understanding is not up to scratch but I'm stumped as to where I'm going wrong...
I am working on a Silverlight web application in Visual Studio 2010.
I recently made the mistake of updating to the latest version Silverlight when prompted through the browser. Consequently, I was unable to build (or even load) my Silverlight application.
To get around this, I installed the Silverlight 5 Developer runtime - which enabled me to build and deploy the application.
However, when I deploy to a locked down test machine, and run the web application, I get an exception something like the following:
"Debugging resource strings are unavailable..."
The test machine appears to have the 'standard' 64-bit Silverlight runtime installed, version 5.1.20513.0.
It appears that the runtime on the test machine is the 'standard' runtime, not the 'Developer' runtime. Because the test machine is locked down, I am unable to install the 'Developer' runtime - so I am forced to find a setup that will both compile on my development machine, and will run on test.
I have tried several combinations of reinstalling Silverlight, with various back versions - but nothing seems to work.
So I am in a quandary:
Compiling on the dev machine won't work without the 'Developer' runtime.
The application cannot be run on the test machine without the Developer runtime being present - which cannot be installed (and shouldn't be necessary - in principle - given that it was working before!).
Is there a 'trick' here that I am missing?
Thanks.
Apparently, I was missing a (required) setting in the Web.Config file.
The "Debugging Strings not available" appears when there is a fault in the application, but the relevant fault code(s) cannot be translated into something more human-readable, (due to the absence of the required DLLs or some such).
After having updated the Web.Config, the problem disappeared.
It turns out that it was a Key/Value pair that was missing from Web.Config:
<add key="DashboardDataRetrievalMethod" value="UseSummaryTable" />
The application was specifically looking for this, and complained with a rather generic error message with no stack trace etc.
I've downloaded (cloned the repository) of script# from https://github.com/nikhilk/scriptsharp
Only I can't figure out how to get the source to compile.
I have installed ScriptSharp 0.7.3.0 as it seems to have a dependency on the installed directory to be present in the build.
I have looked around in the source code for some instructions and on google to no success.
Edit
Seems that the Libraries won't compile for me.
In particular the CoreLib is requesting the a reference to 'mscorlib' be added
BTW I really recommend using script# for complex projects with rich web client experiences.
Especially coupled with knockout.js
I had no problem compiling the whole solution. (v 0.7.3) In fact the CoreLib project seems not having any external reference at all. (CoreLib compiles to mscorlib.dll). You could try to download the complete source package again, open and compile the ScriptSharp solution.
I'm trying to profile a C++ project in Embarcadero RAD Studio 2010. To do this I wanted to use AQTime, but I'm running into a bit of a problem.
I can compile and run the application in release mode. But when I compile in debug mode, using the settings mentioned at http://smartbear.com/support/viewarticle/18053/, I get an access violation.
... faulted with message: 'access
violation at 0x062324bd: read of
address 0x62324bd'. Process stopped.
Use Step or Run to continue.
It seems to come when I load a bunch of dll's at the startup. But since it goes well in release mode I can't seem to figure out what could be the cause.
When the access violation occurs I get thrown out into assembly and that isn't one of the languages I'm fluent in ;)
EDIT : When scanning the .exe with Dependency Walker I get a message saying that the following files can't be found
CC32100MT.DLL
INET140.BPL
RTL140.BPL
VCL140.BPL
IESHIMS.DLL
Could this be the problem somehow? Are these debugspecific or is Dependency Walker not giving me correct information? The same files are said to be missing when I try a release compiled version to.
I'm running on Windows 7 x64, if that could be part of the issue. I have had problems before with the symlink-look-alike (user/AppData/Local...) that MS used for some folders. Notably when I ran an apache server and the htdocs folder actually wasn't located where the server thought it was (and where it appeared to be) :)
Have you tried disabling dynamic rtl which can be found in the C++ builder linker options pane?