floribe2000
SerialPort ParityReplace on Linux
I have an issue with the use of SerialPort on linux.
If parity is set to none, everything seems to work just fine. However, I need to connect to a device using a serial connection where the parity bit is used to indicate a specific byte (some embedded device stuff, there's no way to change that).
On windows, I can ignore the parity errors caused by this by setting ParityReplace to 0, however, on linux this has no effect.
It is not possible to just set parity to none, because I need to specifically set the parity bits when sending messages to the embedded devices.
From what I can tell, it appears this is not implemented. Is this correct or am I missing something here?
If this is indeed the case, are there any known workarounds for this problem? Unfortunately, I was unable to find anything about this problem.
21 replies
❔ C# Analyzer with nuget dependency
I'm trying to pack a simple Analyzer as NuGet package. My problem is including a utility library that provides some useful extensions used across multiple projects. This library is available as nuget package.
Is there any simple way to include that library in my analyzer package, ideally without putting the dll into the analyzer package but as some kind of package reference, similar to how nuget dependencies on normal libraries work?
If not, what is the recommended approach here?
2 replies
❔ Default behavior of ComVisible attribute
While updating a legacy .NET framework application to net6, I ran into a question where I was unable to find an answer for:
What's the default behavior if I simply omit the ComVisible attribute from the AssemblyInfo? The doc says its default value is true but does that only apply when the attribute is set or will it also apply when I don't set the attribute at all? Does the compiler implicitly add anything like that?
8 replies
Branching and release strategy for multi-project repository
I am currently working on improving the branching and release strategy for one of my projects.
At the moment, we use feature branches to develop new features, merge them into dev and whenever we want to publish a new release, we just create a PR from dev to main.
This worked fine so far as we only published a desktop app. However, now we also have a web app in the same repo. It's all in the same repo because both desktop and web need to share a lot of code and it was considered to be too difficult to split it into different repos.
The goal is to have independent and automated releases of both desktop and web app.
My idea for this is to still use a development branch where feature branches get merge into. Then I thought about creating separate release branches for web and desktop based on this dev branch for each version. I thought about using some tool like Nerdbank.GitVersioning to automate the version management for those releases.
Is that a reasonable approach or is there maybe something better that I just missed?
4 replies