GitHub CoPilot Reviews

I'm using GitHub Copilot as a plugin for IntelliJ IDE. Specifically, I'm using the Community Edition of IntelliJ, not the commercial version. However, our organization provides us with a commercial license for GitHub Copilot.

I can upload code snippets or class files and ask for solutions to improve them, which works well. It's a great productivity tool that saves me time searching websites. While there are alternatives like ChatGPT and other AI tools, the advantage of the tool is that it is built into the IDE.

One drawback is that the solution sometimes suggests unwanted code, especially if I accidentally press the tab. This doesn't happen often. Sometimes it seems to understand my code, but other times it doesn't. This inconsistency is confusing.

I suggest adding a feature to write code directly in the IDE from chat responses instead of copying and pasting. GitHub Copilot can generate code based on method comments in the IDE, but chat responses require manual copying. A feature flag to allow direct writing from chat to IDE would streamline the process.

The product should also reduce its price so that it can improve the community engagement. 

I have been using the product for three months. 

As for performance and usage, I've only encountered connectivity issues once or twice. It responds well about 99 percent of the time. 

Regarding stability, there have been a couple of instances where the chat plugin got stuck. However, I don't see this as a major concern. It might improve over time, but it's not a significant issue.

The tool's deployment is mostly straightforward. The plugin installation is simple. However, there's one confusing part during the login process. When you log into your GitHub Copilot account, it redirects you to a website that asks for a device ID or something similar. I can't remember exactly what it asks for, but this ID is shown in the IDE popup. It's a bit confusing and takes a few minutes to understand this workflow. It would be helpful if they added a message saying you can find the device ID in your IDE popup. Adding more instructions to that screen could improve the login flow since it redirects you to a website to log in.

The ROI decreases if questions are too long. If it automatically accessed the content or context, questions could be asked easily with a few lines of code or statements. Currently, providing all details takes a long time, reducing productivity. Since it resides in the IDE, it should ideally have access to the codebase and know more about the context where questions are asked rather than starting from scratch.

Each user needs their license, whether it's a big team or a small team.

It's a helpful tool, but you need strong programming knowledge to use it effectively. If you don't understand Java or your objectives, GitHub Copilot might give wrong answers.

It is good at what it's designed for when handling complex coding scenarios or languages. However, there are limitations. The main issue is similar to other AI agents - requiring skilful prompting to get relevant answers. If GitHub Copilot had access to the entire codebase we're working on, it could provide more relevant and productive outcomes. But I'm not sure how security and privacy aspects would be handled in that situation.

The tool tends to give atomic answers rather than addressing very complex, multi-class-spanning questions. If it had access to more context, it might be able to provide better answers for complex scenarios involving multiple classes or large code structures.

I rate the overall solution an eight out of ten. I can recommend it, given it's affordable, because there are many free alternatives. They may need to introduce a community edition or similar option to compete and remain relevant.

GitHub CoPilot accelerates productivity. It is an easy-to-use solution that is also helpful in saving the team's time. 

The solution is expensive. It should also support other programming languages. 

I have been using the solution for two years. 

I rate the tool's stability a nine out of ten. 

My company has five to 10 users for GitHub CoPilot. 

The product's installation is straightforward. 

GitHub CoPilot can be deployed in-house. 

GitHub CoPilot's pricing is reasonable. Our licensing costs were initially monthly, but then we switched to yearly payments. I rate the tool's pricing an eight out of ten. 

I rate the solution a nine out of ten.