DeepWiki – Which repo would you like to understand?
(deepwiki.com)
16 points by pr337h4m 6 hours ago
I also asked Devin a question:
Can I use a star-tree index with upserts?
I was pleasantly surprised by the answer:
https://deepwiki.com/search/can-i-use-a-startree-index-wit_a...
Briefly: "No, you cannot use a star-tree index with upserts in Apache Pinot. This is a built-in restriction that is explicitly enforced in the system."
Why This Restriction Exists
When upserts are enabled for a table in Pinot, the system explicitly checks and prevents the use of star-tree indexes. This validation happens during table configuration validation.
The code in TableConfigUtils.java explicitly enforces this restriction:
[snip]
At cursory glance, I did not detect a hallucination. The answer was true (AFAIK), clear, and objective. I also see that you can peek into other resources to get some additional contextual information.
Very impressive work. by PeterCorless 2 hours ago
Hey I'm Silas, I worked on this. Lmk if you have any questions!
by silasalberti 6 hours ago
Great work Silas! Can this also be trained on externalized sites, such as human-written docs, our web site, public-facing Google docs, etc?
by PeterCorless 2 hours ago
Currently it uses only the codebase itself. However we‘ve been thinking about adding other source like docs. Can’t promise any timelines yet
by silasalberti 1 hour ago
This might be a bit off-topic but is this custom-built documentation or did you use a template? I've been on the hunt for pretty documentation tools but have only come up with Mintlify.
by cuuupid 4 hours ago
This is entirely custom built but we‘re also fans of Mintlify. We use them for docs.devin.ai
by silasalberti 1 hour ago
Is it custom all the way down or based on something like starlight? I really like Mintlify but $200 per doc site hurts for internal tools.
I can see a lot of merit in custom building though especially with making it easier to dump it into `llms.txt` or exposing a search for AI. Hoping that's where DeepWiki is headed :) by cuuupid 8 minutes ago
This seems like a really cool project and could be really useful in getting to know a new codebase. How do you ensure that the output in the wiki does not contain hallucinations?
by hiatus 6 hours ago
When working with LLMs you can never exclude hallucinations entirely but we‘ve been carefully tuning the system over time and found it to be pretty high signal! The reason we display the code snippets is to make it easy to double check with the source
by silasalberti 1 hour ago
• Apache Pinot - https://deepwiki.com/apache/pinot
• Apache Pulsar - https://deepwiki.com/apache/pulsar
• Apache Superset - https://deepwiki.com/apache/superset
• Grafana - https://deepwiki.com/grafana/grafana
• K8sGPT - https://deepwiki.com/k8sgpt-ai/k8sgpt
Seems, at least at a cursory glance, to produce decent output. by PeterCorless 2 hours ago