Conversation

Notices

1. Embed this notice
   obrhoff (obrhoff@chaos.social)'s status on Friday, 14-Feb-2025 20:21:26 JST obrhoff

   Zed Editor now comes with completion model based on Qwen2.5-Coder-7B.

   For some reason its not offline and needs Github? Not sure what I think about it.
   #zededitor #programming

   https://zed.dev/blog/edit-prediction

   • Embed this notice
     Daniel (djh@chaos.social)'s status on Friday, 14-Feb-2025 20:59:02 JST Daniel

     in reply to

     @obrhoff

     "Edit prediction won't be free forever, but right now we're just excited to share and learn."

     https://zed.dev/blog/edit-prediction

     🦧

     Not sure these small AI models are there yet in terms of quality / latency to run them offline. Especially when most folks don't have powerful GPUs in their dev laptops. Mac folks are probably best positioned here.

   • Embed this notice
     obrhoff (obrhoff@chaos.social)'s status on Friday, 14-Feb-2025 21:03:12 JST obrhoff

     in reply to

     • Daniel

     @djh There is a open PR regarding ollama models. Not sure, if its ever get merged.
     https://github.com/zed-industries/zed/issues/15968

   • Embed this notice
     obrhoff (obrhoff@chaos.social)'s status on Saturday, 15-Feb-2025 00:46:13 JST obrhoff

     in reply to

     • Lucas ✦

     @lucas I checked out Nova, and it made a solid impression but I was not sure if it will survive on the long term and will get the support it needs to thrive. (for example a Golang Debugger is very important to me)

     Good to hear, that you even move on from it to Zed.

   • Embed this notice
     Lucas ✦ (lucas@post.lurk.org)'s status on Saturday, 15-Feb-2025 00:46:14 JST Lucas ✦

     in reply to

     @obrhoff have been playing around with Zed since a week now and have to say I quite like it.

     Will probably replace Nova for me.

   • Embed this notice
     obrhoff (obrhoff@chaos.social)'s status on Saturday, 15-Feb-2025 17:42:38 JST obrhoff

     in reply to

     • Marcin Krzyzanowski

     @krzyzanowskim This might be interesting in this context. https://github.com/bernardo-bruning/ollama-copilot

   • Embed this notice
     Marcin Krzyzanowski (krzyzanowskim@mastodon.social)'s status on Saturday, 15-Feb-2025 17:42:39 JST Marcin Krzyzanowski

     in reply to

     @obrhoff it is impossible to guarantee low latency with local model on average developers machines. and that pretty large model. That may be critical factor.