go-ipfs is MIT licensed open source software. We welcome contributions big and small! Take a look at the community contributing notes.
Please make sure to check out the issues. Search the closed ones before reporting things, and (if you can!) help us with open ones.
Please note that the go-ipfs issues are only for bug reports and directly actionable features. Check the IPFS Community guide on reporting issues if your question doesn't fit as a bug report or an actionable feature, and our guide on opening issues if you are not sure how to make an issue here.
If you're looking to help out, head to the captain's log and try picking up an issue from there.
Please look and conform to our Go Contribution Guidelines.
- See the dev pseudo-roadmap.
- Please adhere to the protocol described in the main ipfs repo and specs (WIP).
- Please make branches and pull-request, even if working on the main repository.
- Ask questions or talk about things in Issues or #ipfs on freenode.
- Ensure you are able to contribute (no legal issues please-- we'll probably setup a CLA).
- Have fun!
All commits in a PR must pass tests. If they don't, fix the commits and/or squash them so that they do pass the tests. This should be done so that we can use git-bisect easily.
We use CI tests which run when you push to your branch. To run the tests locally, you can run any of these: make build
, make install
, make test
, go test ./...
, depending on what you're looking to do. Generally go test ./...
is your best bet.
If you are working on a new feature, prefix your branch name with feat/
. If you are fixing an issue, fix/
. If you are simply adding tests, test/
. If you are adding documentation, doc/
. If your changeset doesn't fall into one of these categories, use your best judgement and come up with your own short prefix.
After that, try to signal what part of the codebase this branch is working on. For example, if you are adding a new test to the DHT that tests for nil providers being returned, then test/dht/nil-provs
would be acceptable. If your changes don't fall cleanly in a single module, you can use a more general descriptor, or leave it off in favor of a slightly more wordy description.
Please also try to keep branch names around or under 20 characters. It keeps things a little cleaner overall. Also try to avoid putting issue numbers in branch names, it takes up space without providing any immediately relevant context about the changeset.
A few examples of good branch names:
feat/cmds/object-diff
- For a Pull Request that adds an
ipfs object diff
command.
- For a Pull Request that adds an
test/dag/cache-invalid
- For adding tests around the merkledag's cache invalidation code.
doc/unixfs/pkg-desc
- For a branch that adds or improves the package description in unixfs.
Commit messages must start with a short subject line, followed by an optional, more detailed explanatory text which is separated from the summary by an empty line. We use GitCop to check that commit messages are properly written. It checks the following:
-
The first line of a commit message, called the subject line should not be more than 80 characters long.
-
The commit message should end with the following trailers:
License: MIT Signed-off-by: User Name <email@address>
where "User Name" is the author's real (legal) name and email@address is one of the author's valid email addresses.
These trailers mean that the author agrees with the developer certificate of origin and with licensing the work under the MIT license.
To help you automatically add these trailers, you can run the setup_commit_msg_hook.sh script which will setup a Git commit-msg hook that will add the above trailers to all the commit messages you write.
See the documentation about amending commits for explanation about how you can rework commit messages.
Some example commit messages:
parse_test: improve tests with stdin enabled arg
Now also check that we get the right arguments from
the parsing.
License: MIT
Signed-off-by: Christian Couder <chriscool@tuxfamily.org>
and
net/p2p + secio: parallelize crypto handshake
We had a very nasty problem: handshakes were serial so incoming
dials would wait for each other to finish handshaking. this was
particularly problematic when handshakes hung-- nodes would not
recover quickly. This led to gateways not bootstrapping peers
fast enough.
The approach taken here is to do what crypto/tls does:
defer the handshake until Read/Write[0]. There are a number of
reasons why this is _the right thing to do_:
- it delays handshaking until it is known to be necessary (doing io)
- it "accepts" before the handshake, getting the handshake out of the
critical path entirely.
- it defers to the user's parallelization of conn handling. users
must implement this in some way already so use that, instead of
picking constants surely to be wrong (how many handshakes to run
in parallel?)
[0] http://golang.org/src/crypto/tls/conn.go#L886
License: MIT
Signed-off-by: Juan Benet <juan@ipfs.io>