Ideas to improve the IPFS developer experience (and feedback)

Hi all,

this is a general thread to gather ideas that can improve your experience as contributor to the IPFS ecosystem, as well as to hear your feedback on things we have already done (most things can be reverted).

As you can read on the blog post (, we have already started taking some steps to make the process swifter, more explicit and painless. We will also post any future endeavors in these forums as well to gather feedback. In the meantime, if you have any ideas, feel free to drop them here!

Hi, i met IPFS a year ago, and i remember test it a bit via cli. Yesterday i return to this crazy project and the experience was a lot more pleasant.
The evolution of multiple packages (ipfs-webui, ipfs-pack, ipfs-cluster, ipfs-desktop), the documentation and the reference to makes the first steps easier to understand the idea of the decentralized protocol.
For now, the only inconvenience was that i cannot find a good mobile app to test local ipfs gateway from home.

Hope the best for this project! <3

I have written some R&D style projects with IPFS. I have been absent about a year and I have come back to what seems like APIs that are even more in flux and much documentation and examples that are out of date. I have spent a non-trivial amount of time trying to get peer discovery for pubsub via wrtc signalling working in the latest version of js-libp2p (0.44.x).

Anyway, it is what it is and I am still very excited about working with IPFS / libp2p libraries and I do recognize that none of this is easy. At the end of the day my hat is off to Protocol Labs.

One suggestion I might add is - if possible - can Protocol Labs leverage the github API to understand patterns and trends in the search of its repos? The aggregate data might make some insightful reports on what developers are trying to understand en masse.

Thanks for the suggestion!! We have been looking into how we can use github data (ex for making sure we’re tracking open issues or PRs that need responses to help maintainers get back to folks more quickly), but AFAIK the Github API only has data on aggregated traffic to a repository, not the order in which visitors browse different repositories. We do have work underway in libp2p to merge more repositories together to make it easier to find the relevant code, and IPFS has already archived a large set of deprecated repos - but agree there’s more work to be done here, and really appreciate the feedback. =]