Use correct ollama client in ChatEdit screen. Fix #76

[habaneraa]

A simple fix for issue #76 .

In addition, I think maybe we can extract these local variables to a global client variable, which naturally functions as a singleton object.

oterm/oterm/ollama.py

Lines 22 to 24 in 147cd2f

	client = AsyncClient(
	host=envConfig.OLLAMA_URL, verify=envConfig.OTERM_VERIFY_SSL
	)

oterm/oterm/ollama.py

Lines 39 to 41 in 147cd2f

	client = AsyncClient(
	host=envConfig.OLLAMA_URL, verify=envConfig.OTERM_VERIFY_SSL
	)

[ggozad]

Thank you for this!
Some (minor) comments:
I am happy with a singleton here but then having the OllamaLLM in parallel seems like a missed opportunity. So I would either:

1. move the functionality for list, show to OllamaLLM and create a singleton out of it.
   or 2) remove OllamaLLM altogether and use your singleton directly.
   Happy to discuss if you think differently, I don't have strong opinions on this.

Could you also add an entry to CHANGES.txt with your username and a description?
As soon as we merge I will release :)

[ggozad]

And of course I should have coffee before opening my mouth...
We don't want the singleton here for OllamaLLM. Perhaps best if list, show etc become static methods on OllamaLLM?

[habaneraa]

You are right. Static methods are absolutely better here.

Note that I kept those "list" and "show" requests to be synchronous. Just like before. I observed that the app will result in a strange NoScreen exception if an async call in on_mount() method of ChatEdit finishes late... (To reproduce, add await asyncio.sleep(1.0) to it). Currently I have no idea why a slow request would lead to NoScreen exception. Will investigate this later. Anyway this PR works well.

[ggozad]

Thank you so much, this looks great! I am way from a computer atm, but will merge and release tomorrow!