Differences with the latest Copilot updates

I just took a few minutes to try Copliot again. Been a few weeks, nothing substantially new or super useful over Cursor, in fact I immediately found a few things lacking:

  1. Couldn’t @Folder, picking #workspace is not very effective either
  2. No @Web or pasting of a link to scan sites.
  3. The tab feature has gotten a lot better at guessing, but didn’t once offer to jump to another file. I mean whos gonna open the CHANGELOG.md and suggest what to add… Am I expected to do that manual work myself now psshhh :rofl:
  4. Copilot is locked to OpenAI, doesn’t use Claude or other models at all. This is to its deficit. Claude Sonnet is VERY VERY good at code, actually better than o1 from my tests.
  5. Compose can work within an existing project to build new features, but not sure what you expect given its purpose. Are you expecting it to run an entire update cycle without your input? That sounds wonderful, and when Copilot offers that I’m all in (oh wait I pay for them all anyways haha).

To your point though as soon as you mentioned Copolot getting new features I jumped to VS Code to try it.

I think this is a cat & mouse game for the next 5 years at least. No real clear winner will emerge any time soon, they will leap frog each other, adding new better integrations, copying the others and then inventing the next level repeating the cycle.

Strap in, luckily you can literally swap back and forth between Cursor & VS Code in the same project with no issues. So my take… USE BOTH :wink: