In this episode I show how to run a local LLM, in this case Gemma4, using Ollama. Thanks to the Weekly Dev Chat folks for the help and inspiration.
Ollama:
https://ollama.com/
OpenCode:
https://opencode.ai/
Gemma4:
https://deepmind.google/models/gemma/gemma-4/
Weekly Dev Chat:
https://weeklydevchat.com/
Have a question or problem for a future video? Constructive feedback? Then comment, DM me, or send an email to ask@saturdaymp.com.
[Read More]The system test is finally working!! I go over, again, why it’s important to make sure your tests fail. Now that the test is working I start refactoring it to remove duplicate code.
Spent longer then I would have thought trying to get a before setup method to work in MiniTest. During my investigation I think I got the JetBrains AI to suggest a fix for a MiniTest PR that was rejected which was interesting to see.
[Read More]I get one system test working for generating the todo file and almost get the second test working. The second test that actually tests the bug we fixed. Went down a bit of rabbit hole trying to get the optimal Ruby file/directory manipulation code.
I also spent some time trying out JetBrains new coding AI. It has some issues but also lots of promise.
[Read More]