In my typical development process I start from:
spin new -t http-rust -a example
cd example
spin watch
Then I alternately edit the code and test the app, e.g. using curl.
Whenever I make a mistake which results in compilation process failure, spin prints the error to stderr and keeps previous version of the app running. This requires me to constantly watch the output from spin or rely on IDE to highlight the problem. Otherwise I may end up testing outdated version of the code, which later leads to invalid reasoning.
The improvement that I would consider is terminating the app before starting compilation. This way any attempt of testing the app would reveal the compilation issue.
In my typical development process I start from:
spin new -t http-rust -a example cd example spin watchThen I alternately edit the code and test the app, e.g. using curl.
Whenever I make a mistake which results in compilation process failure, spin prints the error to stderr and keeps previous version of the app running. This requires me to constantly watch the output from spin or rely on IDE to highlight the problem. Otherwise I may end up testing outdated version of the code, which later leads to invalid reasoning.
The improvement that I would consider is terminating the app before starting compilation. This way any attempt of testing the app would reveal the compilation issue.