vim/YouCompleteMe
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
cbc6c07a1a
YouCompleteMe/CONTRIBUTING.md
Val Markovic b2120c9647 Minor formatting changes to docs
2013-07-01 10:55:56 -07:00

4.3 KiB
Raw Blame History

Writing good issue reports

First things first: the issue tracker is NOT for tech support. It is for reporting bugs and requesting features. If your issue amounts to "I can't get YCM to work on my machine" and the reason why is obviously related to your machine configuration and the problem would not be resolved with reasonable changes to the YCM codebase, then the issue is likely to be closed.

YCM compiles just fine; the build bots say so. If the bots are green and YCM doesn't compile on your machine, then your machine is the root cause. Now read the previous paragraph again.

Realize that quite literally thousands of people have gotten YCM to work successfully so if you can't, it's probably because you have a peculiar system/Vim configuration or you didn't go through the docs carefully enough. It's very unlikely to be caused by an actual bug in YCM because someone would have already found it and reported it.

This leads us to point #2: make sure you have checked the docs before reporting an issue. The docs are extensive and cover a ton of things; there's also an FAQ at the bottom that quite possibly addresses your problem.

Further, search the issue tracker for similar issues before creating a new one. There's no point in duplication; if an existing issue addresses your problem, please comment there instead of creating a duplicate.

Lastly, make sure you are running the latest version of YCM. The issue you have encountered may have already been fixed. Don't forget to recompile ycm_core.so too (usually by just running install.sh again).

OK, so we've reached this far. You need to create an issue. First realize that the time it takes to fix your issue is a multiple of how long it takes the developer to reproduce it. The easier it is to reproduce, the quicker it'll be fixed.

Here are the things you should do when creating an issue:

  1. Write a step-by-step procedure that when performed repeatedly reproduces your issue. If we can't reproduce the issue, then we can't fix it. It's that simple.
  2. Create a test case for your issue. This is critical. Don't talk about how "when I have X in my file" or similar, create a file with X in it and put the contents inside code blocks in your issue description. Try to make this test file as small as possible. Don't just paste a huge, 500 line source file you were editing and present that as a test. Minimize the file so that the problem is reproduced with the smallest possible amount of test data.
  3. Include your OS and OS version.
  4. Include the output of vim --version.

Creating good pull requests

  1. Follow the code style of the existing codebase.

    • The Python code DOES NOT follow PEP 8. This is not an oversight, this is by choice. You can dislike this as much as you want, but you still need to follow the existing style. Look at other Python files to see what the style is.
    • The C++ code has an automated formatter (style_format.sh that runs astyle) but it's not perfect. Again, look at the other C++ files and match the code style you see.
    • Same thing for VimScript. Match the style of the existing code.

  2. Explain in detail why your pull request makes sense. Ask yourself, would this feature be helpful to others? Not just a few people, but a lot of YCM’s users? See, good features are useful to many. If your feature is only useful to you and maybe a couple of others, then that’s not a good feature. There is such a thing as “feature overload”. When software accumulates so many features of which most are only useful to a handful, then that software has become “bloated”. We don’t want that.

    Requests for features that are obscure or are helpful to but a few, or are not part of YCM's "vision" will be rejected. Yes, even if you provide a patch that completely implements it.

    Please include details on exactly what you would like to see, and why. The why is important - it's not always clear why a feature is really useful. And sometimes what you want can be done in a different way if the reason for the change is known. What goal are you trying to accomplish?