One of my Open Source projects uses Packages to build the macOS installer. After GUI-based setup of the pckgproject file, the installer build can be easially invoked from the command line through the packagesbuild command.
I'm now setting up a GitHub Actions based build workflow, running on the GitHub-hosted runners, which should ideally create a ready-to-use installer, so I need to invoke packagesbuild during that process. Unfortunately packagesbuild is not available on the macos-10.15 runner used. I don't see any option to install it during the workflow run, as it's not available through e.g. homebrew but maybe I'm overlooking something. Is there any option I'm overlooking beneath rebuilding my installer and switching to pkgbuild?
Related
I have made a VS2010 project which installs MySQL by calling MySQLInstallerConsole.exe with necessary arguments. This works perfectly however, when I try to run this project's exe as a custom action it executes the exe but MySQL is not installed.
I checked the log file for MySQL exe and it had the following error:
Error 50: Package Name State change request failed.
What does this really mean, any solutions?
There are at least a couple of things that have something to do with this:
I believe the MySQL installer is an MSI file according to this:
http://dev.mysql.com/downloads/installer
You cannot install an MSI from a custom action of a Visual Studio setup because recursive calls to MSI installs are not allowed.
In a Everyone install custom actions run with the SYSTEM account, and installs are not often expecting this if they want to install files in user-profile folders because the SYSTEM account doesn't have any.
You need to bootstrap this somehow, and I think people are often using the WiX Burn bootstrapper for that. You define the ProductCode, name etc of the product and define how to install it.
I'm creating a yeoman generator for my web projects.
But I wonder how I can try and test my changes before publishing it?
Since I have installed it once, it will not run my local development version, instead it runs my installed version.
Any suggestions on how can test-run my local development version?
I finally found some information on how to accomplish this:
if you wish to develop on the generators code base, and debug locally, a common way to do so is to rely on npm link
git clone the generators repo locally
cd into that repository and run npm link. It'll install required dependencies and install the package globally, using a symbolic link to your local version.
If you want to install sub generators, you need to do so in the context of a yeoman-generator package linked earlier. Cd into the sub generators package you have cloned locally and run npm link.
We now have everything linked and known on the system, we now need to link the sub-generator repo into the parent one, yeoman-generator cloned and linked in step 1 & 2.
https://github.com/yeoman/generator/wiki/Testing-generators
EDIT:
Updated link for info: https://yeoman.io/authoring/index.html
If by "running locally" you mean the ability to test your generator and its flow you can simply do this.
In your project directory folder run npm link. If this passes in flying colors, go to step 2.
Open a terminal and cd into the folder you wish to initiate a project.
Run yo generator-theNameOfYourGenerator. This will run your generator.
I have a solution that is primarily developed in Visual Studio 2012. I would like to develop in MonoDevelop without major incompatibilities.
Thus far, I have installed mrward's nuget addin for MonoDevelop and things work if I manually add each package in packages.config through that interface. However, this is highly onerous. This addin doesn't have support for automated package restore as of this writing.
I downloaded nuget.exe from CodePlex ("NuGet command line utility", as it's labeled). I use a simple find/xargs combination to install all required packages:
find . -name packages.config | xargs -I '{}' mono nuget.exe install '{}'
This creates several dozen directories in the directory from which it is run instead of putting things under packages/ as expected, and it also doesn't touch the project files so MonoDevelop still thinks that it should be looking for package references in the directory from which MonoDevelop was started.
I therefore opened MonoDevelop from the working directory that contains all of these package folders, and I still get invalid references. I think this is probably because the project is looking for package_name/ reference, but the folders are name package_name.version/ in the working directory.
Any suggestions for a sane, simple way to interact with this solution? I'm next going to try modifying my shell command so that it automatically drops to project/packages and runs nuget from that directory.
Did you try using the -o command line parameter with NuGet.exe? You can use that to get the packages to install into a particular packages folder.
The NuGet addin for MonoDevelop supports package restore from version 0.6 or above. Right click your project and select Restore Packages. This will download all the packages defined in your packages.config for all projects in the solution. It uses NuGet.exe to do this.
Another way to get this working is to use the custom NuGet MSBuild target so the package restore happens at build time when using xbuild. It would require some manual editing of project files though. Under the covers the custom MSBuild target just uses NuGet.exe with a similar command line to what you have already just with the output directory option specified. So I would try the command line approach since that will be less work.
You would have to get the following files from the NuGet repository on codeplex:
NuGet.exe
NuGet.targets
NuGet.config
Put these in a directory somewhere. Typically these are put in a .nuget directory in the same directory as your solution file. Then you need to edit your project files to include the NuGet.targets file and also define the SolutionDir property. So something like this:
<SolutionDir Condition="$(SolutionDir) == '' Or $(SolutionDir) == '*Undefined*'">..\..\</SolutionDir>
<Import Project="$(SolutionDir)\.nuget\NuGet.targets" />
You will also need to enable package restore on your machine. You can do this using the NuGet addin for MonoDevelop in the Options dialog. Under Linux this is available from the Edit menu under Preferences. Then look in the NuGet - General options and there is a checkbox for enabling package restore.
There is an example project on GitHub created by Jonathan Channon which uses package restore and works when building with xbuild inside MonoDevelop. There is also an issue on GitHub about using NuGet restore on Linux which might be helpful.
Update: 2014-05-14: NuGet addin for MonoDevelop now supports package restore.
I'm following the instructions from here to build MonoDevelop on Mac OS X: Building MonoDevelop on OS X
I can build MonoDevelop just fine. I'd like to also build the MonoMac add in and use it.
The document says:
"To include addins from extras, manually copy their build directories
into the MonoDevelop.app directory, e.g.
cp -r ../../../extras/MonoDevelop.AspNet.Mvc/build/*
MonoDevelop.app/Contents/MacOS/lib/monodevelop/AddIns/MonoDevelop.AspNet.Mvc/"
However, I have no build directory under /extras/MonoDevelop.MonoMac.
I tried to follow this step to build extras:
"To include additional addins from extras in the build, instead of
using the Mac profile, use configure --select to select addins. This
will write the default profile. Next, replace the line in
profiles/default that begins with "main" with the one from
profiles/mac, then run configure again using the default profile."
But it does not give me the option to build the MonoMac extra.
Next, I try to compile the monodevelop.mdw. Again, I can build the main solution (MonoDevelop), but the extras fails with over 1000 errors.
Finally, I tried to take my existing monomac and macdev plugins and copy them into my master monodevelop.app bundle. But this breaks add solution/project dialog whenever I try to select a MonoMac project - it fails to load the Monodevelop.ide DLL.
Any advice on where I am going wrong would be appreciated.
I was not able to make this work following the directions on the MonoDevelop site. Even manually adding the extras/MonoDevelop.MonoMac extra to the profile did not work.
In the end, I build MonoDevelop from source, using the mac profile.
Then I got the latest MonoMac pieces, and build them from source. I used
make update
to install into my installed copy of MonoDevelop.
Then, I manually copied the following packages from my installed MonoDevelop to my newly compiled version from master:
MacPlatform.dll
MonoDevelop.iPhone
MonoDevelop.MacDev
MonoDevelop.MonoMac
Xamarin.Ide
Any everything now works :)
I am new in Hudson.
I would like to execute a 'sourcecodeanalyzer' command in Hudson as Post-build Actions to generate an html report. Please let me know is this at all possible, if yes let me know the Hudson configuration steps to execute the command.
Your earliest response in this regard will be extremely helpful.
Thanks in advance.
Yes, it is almost certainly possible.
You will need to configure the Hudson project to have either a post-build action or a build step that runs your source code analyzer.
You've not stated in your question precisely which analyzer - it may be that Hudson already has a plug-in installed for it, in which case it may be listed on the Config page for the project at the bottom under Post-build Actions.
If not, next check to see if there's a plug-in available for the analyzer that hasn't been installed. From the main Hudson page select Manage Hudson, then Manage Plugins, and choose the Available tab. If there is a plug-in available it's definitely a good idea to use it as they are generally very well integrated with Hudson itself.
As a last resort you'll have to configure a build step to run the analyzer. Configure the project, then choose "Add build step". The drop-down that appears depends on your environment (Windows or Linux) but should include the ability to run a shell command or batch file. You can configure your analyzer there.
(If you're building Windows Visual Studio applications, a more flexible way that I've used is to use the MSBuild plug-in for builds, and have an MSBuild script that builds the application and then runs analysis tools. This can automate pretty much everything: mine builds the application, builds an acceptance test database, runs the acceptance tests and copies the result HTML to a page linked from the project.)
You could create a new job with a "Execute Shell" build step. Type in the command you wish to run in the text box. Then all you have to do is trigger this job by selecting:
"Build after other projects are built"
And select the trigger job from the list.
Hope this helps!
As a follow up to Jeremy's post. If you don't see the ability to add post build steps, you might work with maven jobs. In that case you need the Hudson M2 Extra Steps Plugin. This will give you pre and post build steps.
I use the 'Post build task' plugin to delete some resources after a build. You could call any shell script or command lines. If you want you could make the call depends on some logging output.
there is one best way to solve this:
Upgrade to Fortify SCA 2.6.x (as of writing, latest version is 2.6.5).
Download the Fortify Maven Plugin version 2.6 from https://customerportal.fortify.com and install it into your Hudson server's Maven repository.
Update your project's pom to carry out the Fortify scan. There is an example provided with the Maven plugin.
Currently , I am experiment with sonar plug in. It looks great check the details here
http://sonar.codehaus.org/a-new-hudson-plugin-for-a-closer-integration-with-sonar/
http://wiki.hudson-ci.org/display/HUDSON/Sonar+plugin