5 ms·

The Copilot-Cli is not so bad, https://github.com/features/copilot/cli https://github.com/features/copilot/cli

by azaras 8mo ago

hpdigidrifter 8mo ago
Can't speak for copilot but Gemini cli is unbelievably bad compared to Gemini web. CC has some magic secret sauce and I'm not sure what it is. My company pays for both too, I keep coming back to Claude all-round
- mcintyre1994 8mo ago
  Claude Code is one of a very few AI tools where I genuinely think the people at the company who build it use it all the time.
  - danw1979 8mo ago
    He does ! https://x.com/bcherny/status/2007179832300581177 https://x.com/bcherny/status/2007179832300581177
  - taude 8mo ago
    watch the interviews with Boris. He absolutely uses it to build CC.
  - vdm 8mo ago
    s/AI//
  - giancarlostoro 8mo ago
    They absolutely do, the CEO has come out and said a few engineers have told him that they dont even write code by hand anymore. To some people that sounds horrifying, but a good engineer would not just take code blindly, they would read it and refine it using Claude, while still saving hundreds of man hours.
    - lelanthran 8mo ago
      > They absolutely do, the CEO has come out and said a few engineers have told him that they dont even write code by hand anymore. To some people that sounds horrifying, but a good engineer would not just take code blindly, they would read it and refine it using Claude, while still saving hundreds of man hours. TBH, that isn't sustainable. Skills atrophy. At some point they are going take the code blindly. Considering what they have said in the past about agentic code changes, they are already doing just that - blindly approving code from the agent. I say this because when I last read what one of their engineers on CC tweeted/posted/whatever, I thought to myself "No human can review that many lines of code per month"[1]. --------- [1] IIRC, it was something stupid like 30kLoc reviewed in a month by a single engineer.
      - kamaal 8mo ago
        >>Skills atrophy. I keep telling my friends while experienced devs feel extremely productive. The newer ones will likely not develop skills needed to work with finer aspects of code. This might work for a while, but you do a year or two of this, and then as little as a small Python script will feel like yak shaving.
  - MarcelOlsz 8mo ago
    I know they do because of how painfully awful the Claude web/Claude desktops uxui is, as well as performance.
- michaelcampbell 8mo ago
  I would love to hear/see a definitive answer for this, but I read somewhere that the relationship between MS and \A is such that the copilot version of the \A models has a smaller context window than through CC. This would explain the "secret sauce", if it's true. But perhaps it's not and a lot is LLM nondeterminism mixing with human confirmation bias.
k__ 8mo ago
It's sluggish in GitHub Codespaces, as it has so many animations.
taude 8mo ago
Agreed. I was an early adopter of Claude Code. And at work we only had Copilot. But the Copilit CLI isn't too bad now. you've got slash commands for Agents.MD and skills.md files now for controlling your context, and access to Sonnet & Opus 4.5. Maybe Microsoft is just using it internally, to finish copying the rest of the features from Claude Code. Much like the article states, I use Claude Code beyond just it's coding capabilities....
- yesiamyourdad 8mo ago
  Same situation, once I discovered the CLI and got it set up, my happiness went up a lot. It's pretty good, for my purposes at work it's probably as good as Claude Code.
tveita 8mo ago
The Copilot IntelliJ integration on the other hand is atrocious: https://plugins.jetbrains.com/plugin/17718-github-copilot--your-ai-pair-programmer https://plugins.jetbrains.com/plugin/17718-github-copilot--y... I'm amazed that a company that's supposedly one of the big AI stocks seemingly won't spare a single QA position for a major development tool. It really validates Claude's CLI-first approach.

The Copilot-Cli is not so bad, https://github.com/features/copilot/cli https://github.com/features/copilot/cli

by azaras 8mo ago

The Copilot-Cli is not so bad, https://github.com/features/copilot/cli https://github.com/features/copilot/cli

Can't speak for copilot but Gemini cli is unbelievably bad compared to Gemini web. CC has some magic secret sauce and I'm not sure what it is. My company pays for both too, I keep coming back to Claude all-round

mcintyre1994 8mo ago
Claude Code is one of a very few AI tools where I genuinely think the people at the company who build it use it all the time.
- danw1979 8mo ago
  He does ! https://x.com/bcherny/status/2007179832300581177 https://x.com/bcherny/status/2007179832300581177
- taude 8mo ago
  watch the interviews with Boris. He absolutely uses it to build CC.
- vdm 8mo ago
  s/AI//
- giancarlostoro 8mo ago
  They absolutely do, the CEO has come out and said a few engineers have told him that they dont even write code by hand anymore. To some people that sounds horrifying, but a good engineer would not just take code blindly, they would read it and refine it using Claude, while still saving hundreds of man hours.
  - lelanthran 8mo ago
    > They absolutely do, the CEO has come out and said a few engineers have told him that they dont even write code by hand anymore. To some people that sounds horrifying, but a good engineer would not just take code blindly, they would read it and refine it using Claude, while still saving hundreds of man hours. TBH, that isn't sustainable. Skills atrophy. At some point they are going take the code blindly. Considering what they have said in the past about agentic code changes, they are already doing just that - blindly approving code from the agent. I say this because when I last read what one of their engineers on CC tweeted/posted/whatever, I thought to myself "No human can review that many lines of code per month"[1]. --------- [1] IIRC, it was something stupid like 30kLoc reviewed in a month by a single engineer.
    - kamaal 8mo ago
      >>Skills atrophy. I keep telling my friends while experienced devs feel extremely productive. The newer ones will likely not develop skills needed to work with finer aspects of code. This might work for a while, but you do a year or two of this, and then as little as a small Python script will feel like yak shaving.
- MarcelOlsz 8mo ago
  I know they do because of how painfully awful the Claude web/Claude desktops uxui is, as well as performance.
michaelcampbell 8mo ago
I would love to hear/see a definitive answer for this, but I read somewhere that the relationship between MS and \A is such that the copilot version of the \A models has a smaller context window than through CC. This would explain the "secret sauce", if it's true. But perhaps it's not and a lot is LLM nondeterminism mixing with human confirmation bias.

k__ 8mo ago

It's sluggish in GitHub Codespaces, as it has so many animations.

taude 8mo ago

Agreed. I was an early adopter of Claude Code. And at work we only had Copilot. But the Copilit CLI isn't too bad now. you've got slash commands for Agents.MD and skills.md files now for controlling your context, and access to Sonnet & Opus 4.5. Maybe Microsoft is just using it internally, to finish copying the rest of the features from Claude Code. Much like the article states, I use Claude Code beyond just it's coding capabilities....

yesiamyourdad 8mo ago
Same situation, once I discovered the CLI and got it set up, my happiness went up a lot. It's pretty good, for my purposes at work it's probably as good as Claude Code.

tveita 8mo ago

The Copilot IntelliJ integration on the other hand is atrocious: https://plugins.jetbrains.com/plugin/17718-github-copilot--your-ai-pair-programmer https://plugins.jetbrains.com/plugin/17718-github-copilot--y... I'm amazed that a company that's supposedly one of the big AI stocks seemingly won't spare a single QA position for a major development tool. It really validates Claude's CLI-first approach.