Working on the GUI implementation of pluggable agents, I realized that users creating agents would probably not use our naming convention and that I would have to allow for it.
That led me to realize that it might be a nice feature for each of our agents to have a different name... like net20-agent, netcore31-agent, etc. That way users could tell immediately which agent had been selected to run their tests when a problem arose.
I'm initially calling this an idea and a breaking change. I think it's at least "a little bit" breaking anyway. Not sure if it's sufficiently breaking to make it wait to 4.0 @nunit/engine-team What do you think?
Working on the GUI implementation of pluggable agents, I realized that users creating agents would probably not use our naming convention and that I would have to allow for it.
That led me to realize that it might be a nice feature for each of our agents to have a different name... like
net20-agent,netcore31-agent, etc. That way users could tell immediately which agent had been selected to run their tests when a problem arose.I'm initially calling this an idea and a breaking change. I think it's at least "a little bit" breaking anyway. Not sure if it's sufficiently breaking to make it wait to 4.0 @nunit/engine-team What do you think?