[hwchase17]

LangChain co-founder here. There's lots of good feedback here (that also resonates with previous feedback) that we're working hard to address. On some key points:

- We genuinely appreciate all the thoughtful criticism and feedback. Our goal is to make it as easy as possible to build LLM applications (both prototypes and production-ready applications), and if we're falling short in an area we'd much prefer to hear it rather than not. We don't have the bandwidth to respond to all feedback directly, but we do (1) appreciate it, and (2) try to address it as quickly as possible.

- Documentation: we've heard this for a while now, and have been working to improve it. In the past ~3 weeks we've revamped our doc structure, changed the reference guide style, and worked on improving docstrings to some our more popular chains. However, there is a still a lot of ground to cover, and we'll keep on pushing. Feedback on which specific chains/components need better documentation is particularly helpful

- Customizability: we need to make it easy to customize prompts, chains, and agents. We're thinking of changes to more easily enable this - better documentation, more modular components. We'll up the priority of this.

- Other tooling: there are general difficulties in building LLM applications that aren't strictly related to langchain, such as debugging and testing. We're working on building separate tooling to assist with this that we hope to launch soon.

[morgango]

I think Langchain is like democracy, everyone complains about it and tries to poke holes in it but it is clearly better than all the alternatives.

Once I got "into" langchain and how it did things my life as a developer got infinitely easier. It is true that it is doing a lot of things that you "could" do elsewhere, but that is kind of the point of a library. For example, it makes it incredibly easy to switch between vector datastores or embeddings, with just a tiny code change. I love that.

Look at how much code it takes to actually get something done. It makes it trival to take a file (or a number of files), chunk them, and load them into a vector store. Sure, I could write and maintain the code to do that, but why?

While I did find it challenging to get started with Langchain, it was more a lack of understanding of the ecosystem than anything else. Great abstractions aren't going to shield me from that without restricting choice. The documentation has improved noticeably in the last few weeks.

Great work, it is very much appreciated by the non-HN crowd. Don't let this feedback get you down.