3 ms·

I used it 3 years ago or so, but ultimately decided against it and went back to Python/Go for creating my IoT integrations. It works (worked?) well, but had a

I used it 3 years ago or so, but ultimately decided against it and went back to Python/Go for creating my IoT integrations. It works (worked?) well, but had a few "gotchas" like where your workflows are stored, configuration, logging, monitoring of application status, and general NPM upgrade madness. There is (probably) also the general issue of 4th gen development tools (point & click) where maintaining a codebase is rather tedious. I've worked with WebSphere Message Broker in the past, and major upgrades could be a chore. I'm aware most of the issues i had have been fixed in later versions, but i've not had the need/urge to go back and try it again. That doesn't mean others shouldn't.

Tajnymag 8y ago
Interesting, I've always had way more trouble with pip than npm/yarn ecosystem.
- 8fingerlouie 8y ago
  Pip is easy once you start using requirement.txt files and virtualenv/docker/jails/whatever. it all boils down to "pip install --upgrade -r requirements.txt" Where most people start having trouble is when they install some packages through rpm/apt/pkg and then install others through pip, and there are shared dependencies between the two sets of packages with different versions.