Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

GitHub Copilot is now available for free

github.com

Post reply on HN
Print view View original

111–120 of 412 posts

Re: GitHub Copilot is now available for free

#111

Post by SkyPuncher »

joshstrange wrote: post #50
JadoJodo wrote: post #35

Earlier quoted context omitted.

Has anyone who uses an IDE (e.g., JetBrains, not a code editor) moved to Cursor? I've downloaded it a few times because everyone raves about it, but I've always come back almost immediately because editors can't reliably make changes across projects (among many other things)... What am I missing? FWIW I use GoLand w/ Supermaven, currently.

This is the main sticking point for me, I'm not leaving JetBrains anytime soon. GitHub Copilot + Aider handle my needs beautifully and while I wish Aider had deep IDE integration I can work around that (Yes, I know about the "AI!" command thing, it's a cool idea but sucks in practice). Aider in browser mode has pretty much replaced my back and forth to ChatGPT/Claude's web UI to the point that I'm considering going A…

Jetbrains is still the leader in all of the small details that make navigating and working a code base easy.

Copy and paste workflow is a minor slowdown, but nothing compared to things like smart links in terminal, auto detection of run configurations, etc, etc.

Re: GitHub Copilot is now available for free

#112

Post by crucialfelix »

orra wrote: post #105

I like the mediocrity of the example they headlined: the user asks for unit tests. Copilot writes only two test cases, so not exactly great coverage. Plus, the test cases uses python's unittest, which isn't as slick as pytest.

You can write a prompt stating how tests should be writen in your project. It goes on some .vscode file. It's consulted anytime it's writing a test, even in a conversation. I state where my factory fixtures are to be found.

Re: GitHub Copilot is now available for free

#113

Post by sionisrecur »

orra wrote: post #105

I like the mediocrity of the example they headlined: the user asks for unit tests. Copilot writes only two test cases, so not exactly great coverage. Plus, the test cases uses python's unittest, which isn't as slick as pytest.

Well they didn't ask for pytest tests. I wonder if AIs are like bored developers, doing the minimum necessary to complete the task instead of going the extra mile and make things better. Perhaps that will be the differentiator in the future. Or they just need to adjust the pre-prompt.

Re: GitHub Copilot is now available for free

#114

Post by xandrius »

beepbooptheory wrote: post #64

Working with people who use this stuff a lot has made my current job just so so much harder in every way its astonishing. I used to solve problems with code, now I feel like hermeneut or dream analyzer: absent of human intention, codebases quickly become these weird piles of different idioms, even without considering the hallucinations (those have definitely cost me a few sleepless nights now either way). But I am ju…

I hear you. There was something special about the old days when programming was all about taking your time, thinking through every step, and truly understanding what you were building. I miss the days of punching cards — there was a certain simplicity to it. You’d write your code, feed it into the machine, and if it broke, it was your fault. There was no hiding behind tests or CI/CD pipelines, no auto-fixes or layers of abstraction. It was just you and the machine, and every bug was a lesson you had to learn the hard way. The feedback loop was slow, but it was real.

Now, everything feels automated, fast, and often a bit too dumb. Sure, it’s easier, but it’s lost that raw connection to the work. We’ve abstracted away so much that it’s hard to feel like we’re truly engineering something anymore — it’s more like patching together random components and hoping it holds. I think we lost something when we all started staring at screens all day and disconnected from the hands-on nature of building. There's a lot of slop now, and while some people thrive on that, it’s not for everyone.

Re: GitHub Copilot is now available for free

#115

Post by scosman »

thot_experiment wrote: post #33

If anyone is looking for a free/local alternative Continue + Ollama is acceptable. If you're just doing run of the mill programming it will work well out of the box. I'm glad it's open source so I was able to fix most of the issues I had with it and now my copy is in a great place. The documentation is in places many versions behind the actual code so it can be tough to figure out how to set things up when you're ven…

Acceptable UX maybe, but the gap between Sonnet 3.5 and open models isn’t worth it. I know people are going to pitch qwen coder 72b, but it’s still a long way off on benchmarks and my time matters more.

Re: GitHub Copilot is now available for free

#116

Post by thot_experiment »

meta_x_ai wrote: post #81
thot_experiment wrote: post #33

If anyone is looking for a free/local alternative Continue + Ollama is acceptable. If you're just doing run of the mill programming it will work well out of the box. I'm glad it's open source so I was able to fix most of the issues I had with it and now my copy is in a great place. The documentation is in places many versions behind the actual code so it can be tough to figure out how to set things up when you're ven…

A software developer's time is much more precious than wasting time on sub-optimal models. Open Weights models has it's place (in training custom agents and custom services), but if you are knowledge worker, using a model even 5% less than SOTA is extremely dumb

100% disagree with this take, the flexibility in controlling the prompt leads to QwenCoder2.5-32b outperforming gpt-o1 and claude sonnet 3.5 for nearly everything that I use it for (true for Gemma-27b and llama3.3-70b, though in this context I'm almost always using the former). A specialist model that's specifically prompted to do the correct thing will outperform a SOTA generic model with a one size fits all system prompt. This is why small autocomplete models can very obviously outperform larger models at that specific task. I am speaking 100% from experience and ignoring all benchmarks in forming this view btw, so maybe it's just my specific situation.

Also, in general I don't find the difference between SOTA models and local models to be that significant in the real world even when used in the exact same way.

Re: GitHub Copilot is now available for free

#117

Post by olup »

jasondjk wrote: post #76
olup wrote: post #26

I use supermaven and cline with my own API key, a setup superior to cursor imo. Tried to go back to gh copilot yesterday but couldn't bear it for a full workday, and reverted to my previous arrangement.

This looks really interesting, cursor has been way better than copilot for me but supermaven looks great. I went down the rabbit hole with: https://www.youtube.com/watch?v=zLQuBSuzu2w&t=661s Your setup sounds interesting. What sort of API key do you use?

We have an openai account for the company, so I mainly use gpt4o or 4o mini with supermaven and cline. I think Claude 3.5 works even better.

Re: GitHub Copilot is now available for free

#118

Post by thot_experiment »

TacticalCoder wrote: post #44
thot_experiment wrote: post #33

If anyone is looking for a free/local alternative Continue + Ollama is acceptable. If you're just doing run of the mill programming it will work well out of the box. I'm glad it's open source so I was able to fix most of the issues I had with it and now my copy is in a great place. The documentation is in places many versions behind the actual code so it can be tough to figure out how to set things up when you're ven…

On what kind of hardware/GPU are you running that locally?

2x 1080Ti, ~25t/s

Re: GitHub Copilot is now available for free

#119

Post by vr46 »

JadoJodo wrote: post #35
asdev wrote: post #5

I ditched Copilot for Cursor and never looked back. Cursor might be the only other AI product with product market fit than ChatGPT, it's that good

Has anyone who uses an IDE (e.g., JetBrains, not a code editor) moved to Cursor? I've downloaded it a few times because everyone raves about it, but I've always come back almost immediately because editors can't reliably make changes across projects (among many other things)... What am I missing? FWIW I use GoLand w/ Supermaven, currently.

Yes, I finally decided not to renew my über-cheap JetBrains subscription in January and shift to VSCode + Copilot with Claude, then tested Cursor and dropped Copilot. Cursor is still a bit frustrating and gets trapped in circular reasoning a lot, but its features are good, like creating files.

Re: GitHub Copilot is now available for free

#120

Post by matsemann »

JadoJodo wrote: post #35
asdev wrote: post #5

I ditched Copilot for Cursor and never looked back. Cursor might be the only other AI product with product market fit than ChatGPT, it's that good

Has anyone who uses an IDE (e.g., JetBrains, not a code editor) moved to Cursor? I've downloaded it a few times because everyone raves about it, but I've always come back almost immediately because editors can't reliably make changes across projects (among many other things)... What am I missing? FWIW I use GoLand w/ Supermaven, currently.

Actually writing code is a small part of what I do in the IDE. So I'm not keen on jumping ship to a whole new editor and lose all the IDE stuff. But I do think IDEs need to step up. The editors will get the IDE features slowly (or be able to work around some of them, like how an IDE can know your database layout and help you write sql, the AI helpers can make educated guesses as well and get close), so the IDEs don't have a big moat.
Post reply on HN