I've upgraded apps from every version since .Net Core 2.0. The upgrades always take a few hours, maybe half a day (with the exception of 2.0 to 3.0 which took a few days - many breaking changes). It's well worth a few hours per year to get the advantages of newer languages features, enhancements in Asp.Net Core, security and performance enhancements in the .Net Runtime, etc.
Is upgrading every 3 years really that bad? As far as I know, they typically aren’t removing features or causing disruptive changes in .NET versions.
They did when switching away from .NET Framework, but this was because they had to reimplement many features from scratch to work on other platforms, and certain parts got left behind.
The last few .NET updates have been pretty much effortless updates for us. These are internal web apps and not using really esoteric C# features but that probably describes a reasonably large percentage of enterprise .NET adoption.
A shift in those cultural dynamics is that you can ship the current .NET LTS with your app (or even STS if you feel like making that sort of security support SLA with your own clients). You aren't relying on their Windows Update habits (or lack of them) or having to install a big .NET installer that might break their other apps.
They may still get left behind on an older version of your software because they want to be, but their relationship to Windows is no longer the big excuse/reason to skip updating to your latest that it used to be.
You are right that it’s on a 2 year cycle. Though there is support for 3 years, so you could safely put off an upgrade for up to a year if there are breaking changes one release.
When I need to send someone an app, I usually use .NET Framework 4.8 because I know it's already installed on every recent version of Windows. This way all I have to distribute is a single very small exe. No need to package a framework with it.
Similar feat with .NET core usually results in a 70-80MB executable.
Yea that's why I write all my programs in C using the win32 APIs. It has a lifetime even longer than .NET Framework 4.8.1! I'll never have to change or adapt to new things. /s
I've upgraded apps from every version since .Net Core 2.0. The upgrades always take a few hours, maybe half a day (with the exception of 2.0 to 3.0 which took a few days - many breaking changes). It's well worth a few hours per year to get the advantages of newer languages features, enhancements in Asp.Net Core, security and performance enhancements in the .Net Runtime, etc.
Is upgrading every 3 years really that bad? As far as I know, they typically aren’t removing features or causing disruptive changes in .NET versions.
They did when switching away from .NET Framework, but this was because they had to reimplement many features from scratch to work on other platforms, and certain parts got left behind.
The last few .NET updates have been pretty much effortless updates for us. These are internal web apps and not using really esoteric C# features but that probably describes a reasonably large percentage of enterprise .NET adoption.
I don't think they introduced any breaking changes after .NET 6
2 replies →
.NET LTS is on a 2-year cycle, isn't it?
I've worked in .NET shops with very niche WPF/WinForms applications where customers were years behind with our software/major .NET Framework releases.
I don't think it's a technical challenge, more a cultural one.
A shift in those cultural dynamics is that you can ship the current .NET LTS with your app (or even STS if you feel like making that sort of security support SLA with your own clients). You aren't relying on their Windows Update habits (or lack of them) or having to install a big .NET installer that might break their other apps.
They may still get left behind on an older version of your software because they want to be, but their relationship to Windows is no longer the big excuse/reason to skip updating to your latest that it used to be.
You are right that it’s on a 2 year cycle. Though there is support for 3 years, so you could safely put off an upgrade for up to a year if there are breaking changes one release.
When I need to send someone an app, I usually use .NET Framework 4.8 because I know it's already installed on every recent version of Windows. This way all I have to distribute is a single very small exe. No need to package a framework with it.
Similar feat with .NET core usually results in a 70-80MB executable.
Is 70-80MB really a big deal to distribute in 2025?
Also if you use trim unused code and compress the executable in my experience it's usually a lot smaller than this.
1 reply →
With all the dependencies that maybe getting an update to the new .Net version, yes.
Yea that's why I write all my programs in C using the win32 APIs. It has a lifetime even longer than .NET Framework 4.8.1! I'll never have to change or adapt to new things. /s
I change things if it’s necessary, not for the sake of using the newest version.
Now I have to update software and all its dependencies without a real benefit.
It’s equivalent of buying a new PC because MS tries to force Windows 11.
Lets waste lots of time and resources for nothing.
Just the amount of perf improvements they have added since .NET 4.8, year after year, is staggering.
Having written .NET 5+ apps, and now being back to maintaining a 4.8 app...
There's a lot of quality of life features of the later versions of C# that we're missing out on. My code was a LOT more streamlined in my .NET 9 app.