←back to thread

Show HN: I built an AI that turns GitHub codebases into easy tutorials

(github.com)
923 points zh2408 | 5 comments | 19 Apr 25 21:04 UTC | HN request time: 0.81s | source
Show context
ryao ◴[19 Apr 25 22:53 UTC] No.43740101[source]▶
I would find this more interesting if it made tutorials out if the Linux, LLVM, OpenZFS and FreeBSD codebases.
replies(3): >>43740141 #>>43740152 #>>43743972 #
1. zh2408 ◴[19 Apr 25 23:02 UTC] No.43740152[source]▶
The Linux repository has ~50M tokens, which goes beyond the 1M token limit for Gemini 2.5 Pro. I think there are two paths forward: (1) decompose the repository into smaller parts (e.g., kernel, shell, file system, etc.), or (2) wait for larger-context models with a 50M+ input limit.
replies(3): >>43740668 #>>43741788 #>>43745188 #
2. rtolsma ◴[20 Apr 25 00:46 UTC] No.43740668[source]▶
You can use the AST for some languages to identify modular components that are smaller and can fit into the 1M window
3. achierius ◴[20 Apr 25 05:50 UTC] No.43741788[source]▶
Some huge percentage of that is just drivers. The kernel is likely what would be of interest to someone in this regard; moreover, much of that is architecture specific. IIRC the x86 kernel is <1M lines, though probably not <1M tokens.
replies(1): >>43743038 #
4. throwup238 ◴[20 Apr 25 11:09 UTC] No.43743038[source]▶
The AMDGPU driver alone is 5 million lines - out of about 37 million lines total. Over 10% of the codebase is a driver for a single vendor, although most of it is auto generated per-product headers.
5. ryao ◴[20 Apr 25 17:36 UTC] No.43745188[source]▶
The first path would be the most interesting, especially if it can be automated.