(Considering) Moving .net4 -> .net5. End of Windows XP/Vista support.
Posted: Fri 2020.05.29, 22:09
Subj says it all - I'm considering to stop supporting Windows XP/Vista. This will not happen tomorrow or any time soon, I might even discard this move at all, at the moment I'm just considering it.
.net4, used by Launcher (it uses not only .net4, but that's not relevant in this context), was released back in 2010. My main reason to not upgrade the framework to the newer versions was the cornerstone of Windows XP support. Apart of better asynchronous code support, newer versions do not contain a whole lot of improvements, that would justify the drop of Windows XP support (IMO).
Not that long ago Microsoft has decided to reunify, previously split .net, into a single framework under the name of .net5, which is scheduled for release sometime in November 2020, with further versions increase by +1 each following year (e.g.: .net6 in November 2021, etc.).
You could ask "Why are you even considering to move, while everything works fine?", which will be a valid question to ask. I cannot give you a definitive answer on this one, yet. There are a lot of claims, regarding .net5, but how valid they are I'm yet to find out. In theory it should simplify cross-platform development (Mac, Android, Linux support for Launcher? who knows...), reduce the amount of resources used by the framework (this one I highly doubt...), higher performance, better interoperability (while I have no doubt about other platforms, I'm not so sure about Windows), and more. Also if all claims are valid, then someone might actually make it usable on Windows XP (not that I think that someone will do so, but just in theory). This move will also allow me to add some new features in the future, that might be extremely difficult to create from zero on the current framework.
As I've already mentioned - I'm not sure yet if I'll eventually move or not, right at this moment I'm downloading preview version of the upcoming .net5 and will be looking at it over some period of time. My main concern is over how Windows-specific things are used there, and how easily it'll be to convert whole project from .net4 to .net5. I'm in full control of most of the codebase used in Launcher, and I plan to replace those few non-Microsoft 3rd party libraries in the nearest future anyway, but that still leaves Microsoft itself, so I'll be inspecting changes and how those will reflect onto their use by Launcher. If everything will be easy-peasy then my only concern will be Windows XP/Vista support, which is one of the main reasons for this thread to begin with. If there will be a whole lot of other problems with move to .net5 then I'll make sure to notify you and I might skip this move even if everyone will be OK with the drop of XP/Vista support, so it's too early to say anything definitively, but having said that - preview is not a final version, so things might change, not even to mention future .net6 and so on and so forth, at some point this question will arise again, the question of whether I should end the support of Windows XP/Vista.
tl/dr:
I'm considering to end Windows XP/Vista support in like 6 months or so, if everything will be great with .net5
Vote on the pool whether you're using Windows XP/Vista and Launcher on it, this will influence on my decision. Feel free to leave your thoughts and concerns in this thread.
.net4, used by Launcher (it uses not only .net4, but that's not relevant in this context), was released back in 2010. My main reason to not upgrade the framework to the newer versions was the cornerstone of Windows XP support. Apart of better asynchronous code support, newer versions do not contain a whole lot of improvements, that would justify the drop of Windows XP support (IMO).
Not that long ago Microsoft has decided to reunify, previously split .net, into a single framework under the name of .net5, which is scheduled for release sometime in November 2020, with further versions increase by +1 each following year (e.g.: .net6 in November 2021, etc.).
You could ask "Why are you even considering to move, while everything works fine?", which will be a valid question to ask. I cannot give you a definitive answer on this one, yet. There are a lot of claims, regarding .net5, but how valid they are I'm yet to find out. In theory it should simplify cross-platform development (Mac, Android, Linux support for Launcher? who knows...), reduce the amount of resources used by the framework (this one I highly doubt...), higher performance, better interoperability (while I have no doubt about other platforms, I'm not so sure about Windows), and more. Also if all claims are valid, then someone might actually make it usable on Windows XP (not that I think that someone will do so, but just in theory). This move will also allow me to add some new features in the future, that might be extremely difficult to create from zero on the current framework.
As I've already mentioned - I'm not sure yet if I'll eventually move or not, right at this moment I'm downloading preview version of the upcoming .net5 and will be looking at it over some period of time. My main concern is over how Windows-specific things are used there, and how easily it'll be to convert whole project from .net4 to .net5. I'm in full control of most of the codebase used in Launcher, and I plan to replace those few non-Microsoft 3rd party libraries in the nearest future anyway, but that still leaves Microsoft itself, so I'll be inspecting changes and how those will reflect onto their use by Launcher. If everything will be easy-peasy then my only concern will be Windows XP/Vista support, which is one of the main reasons for this thread to begin with. If there will be a whole lot of other problems with move to .net5 then I'll make sure to notify you and I might skip this move even if everyone will be OK with the drop of XP/Vista support, so it's too early to say anything definitively, but having said that - preview is not a final version, so things might change, not even to mention future .net6 and so on and so forth, at some point this question will arise again, the question of whether I should end the support of Windows XP/Vista.
tl/dr:
I'm considering to end Windows XP/Vista support in like 6 months or so, if everything will be great with .net5
Vote on the pool whether you're using Windows XP/Vista and Launcher on it, this will influence on my decision. Feel free to leave your thoughts and concerns in this thread.