NUnit-Agent has an odd argument structure as compared to normal command-line apps. The first two args are positional and required... if they aren't present, the process just fails without explanation.
We felt OK doing this because the agent process is only started by TestAgency and the agent and agency are maintained in the same project. There's even a comment in the agent code explaining this logic.
With 4.0, this will change in a few ways. With pluggable agents, agent and agency code may not be maintained in the same repo any longer. We are also likely to start agents in other ways than through test agencies.
I propose rationalizing the agent a bit by making all the accepted arguments named options: agentID for the first one, agencyUrl for the second. The code is already present for named arguments is already present.
I don't actually think this is breaking, but 4.0 it seems like a good point at which to make the change anyway.
At the same time, we may want to make the argument processing more like what we do in the console... accepting / to start an option on windows and sing-letter abbreviations for some of the options.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4