Future of SourceBrowser

[alejandro5042]

From the README:

This is a reference implementation [...] We will consider accepting high-quality pull requests that add non-trivial value, however we have no plans to do significant work on the application in its current form. Any significant rearchitecture, adding large features, big refactorings won't be accepted because of resource constraints.

That begs the question: Is there another project that aims to be maintained long term? Is there a "better form" of this source browser that's in the works? Is the way this project is coded incompatible with future C# features? Is it merely time constraints on the maintainer?

I ask because this project seems to have heavy usage in Microsoft source browsers:

• https://source.dot.net/
• https://referencesource.microsoft.com/

We also have a deployment with hundreds of users indexing internal sources. We're incredibly grateful for this project, but wondering about its future :)

[KirillOsenkov]

Unfortunately I don't have a good answer for you right now.

SourceBrowser won't evolve because the static HTML approach is fundamentally limited. I also don't have any time to continue maintaining it, so it's minimal updates once in a while (such as updating to the latest Roslyn).

There is indeed a replacement project in progress now, https://github.com/ref12/codex, but it is not ready for public consumption and is not yet supported. Volunteers are welcome to explore that, but we don't have capacity to support or answer any questions about that for now.

So, to answer your question, it's a severe lack of maintainers time, and it's not officially funded by Microsoft, so it is what it is. It took effort to open source it at all, so I'm happy to hear that at least it's being useful.

Good luck!

[alejandro5042]

I was afraid of that answer 😢

This tool is AWESOME and in my opinion, should get more official Microsoft investment.

I tried the ref12/codex project but was unable to actually get it running. Once I have more time, I'll post there with the issues I've had. Any idea of when that project would be ready for prime time? Months, years?

[KirillOsenkov]

Not any time soon, I'm afraid :| Both SourceBrowser and Codex are side projects that we only work on in spare time.

cc @ljcollins25

[cloudhan]

Defining protocol maybe more important than adding new features. There is other similar software Sourcetrail for other languages. But using it locally sometimes is pain in the ass, that is, your code base is too large to be built and indexed locally. An unified indexer protocol might enables the ability of integrating and leveraging existing efforts.