3 ms·
> when trying to track down buggy behaviour, it's very useful to know that some component couldn't have been the problem. > However, writing only pure code is
by mrkeen 4y ago
> when trying to track down buggy behaviour, it's very useful to know that some component couldn't have been the problem.
> However, writing only pure code is quite limiting. Also, we still need to audit all IO code.
Quite the turn-around.
- belmont_sup 4y agoI read the first statement as being able to do IO judiciously with a subset of IO work, e.g. network, filesystem. The second is no IO.