dogline

This is a breakdown of how I built a book scanner using my iPhone camera, and the problems and issues I ran into. With pictures. Normally, book scanning can take excessive time and ideally, a scanning rig, but AI and OCR can handle page skew and motion pretty well now.

Remember, if all you need is a git remote to push to somewhere via ssh, `git init --bare` on your server should be available and is solid if you don't need a web interface. Now you've got something you can push to on a shared server.

Unless you really also want a web interface, issue tracking, et. al. Then, use any of these fine tools. Go wild!

Good, rational points from an expert. Not AI slop. It's silly that we have to separate this now.

Well, this post made me reassess using zswap vs zram. I've expanded my zswap and disabled zram. Not sure if that makes any difference to others, but it's probably good for me.

Wait, what? This sounds important, but I can't quite parse it.

In my case, I'm just trying to make a list of official sources, so there's minimal data that can go stale. Although I'm using Claude to help put this together, I'm reviewing most everything, or I gave Claude the list I want to include from my own searches.

Is there liability for posting something like this? I don't think so, but can't say that I'm certain. It would be sad if there is liability.

UDF: User Defined Function

Except for history, I don’t find much that stops you from switching back and forth on the CLI. They both use tools, each has a different voice, but they both work. Have it summarize your existing history into a markdown file, and read it in with any engine.

The APIs are pretty interchangeable too. Just ask to convert from one to the other if you need to.

I think it's becoming clear we're not anywhere near AGI, we figured out how to vectorize our knowledge bases and replay it back. We have a vectorized knowledge base, not an AI.

With lots of people showing how Saas apps can be easily written these days, I'm not as interested in those articles, as people showing off new ideas of what I can do with these new found abilities. This is cool.

I would drive hours to get to the nearest Fry's to me, to pick up some new gear. Being able to browse everything and look around was great. For me, online ordering of parts probably hurt Fry's, but the real reason was after a while, you were never sure if the video card you were buying was new, or actually a return item, and after a couple times having to drive all the way back for something that was missing parts, the whole thing just seemed way too risky. Amazon and Newegg nailed that door shut.

You're right, and it's a good idea. The summary started out small, as a header to the actual daily pages, but then I realized I could have AI do a lot more work here, including silly things like collect weather references and assemble them together. My prompt kept getting bigger to find trends in the data. But, it takes away from the view-ability of the site, which is not good.

LLM's ability to take 7400+ handwritten entries and try to make a narrative out them is amazing. With all of the AI experiments on HN lately, we're figuring out the power of LLMs, but it most cases, it still needs a human refining touch, and we need to remember that. Or else it just looks like AI slop.