71–80 of 101 posts
The original ShowHN: https://news.ycombinator.com/item?id=14497295
Incredible, their website used to both simpler and have screenshots of the actual software[0], now it's just a generic LLM slop website. [0] https://archive.ph/x0sck
What I'm confused about with Monica is this being v3. Didn't Monica already get rebuilt from the original version in 2017, since there's a beta software called Chandler? I know very little about the Monica ecosystem but how come you're throwing away a second version as well? What were the flaws in that iteration?
The version numbering here is an absolute mess. Chandler/Monica 2 is actually Monica v5.0.0-beta, which would logically mean that this newly announced Monica 3 is actually Monica v6.x. Chandler/Monica 2 (v5.0.0-betas) never made it out of beta. It’s had five beta releases but no stable release. The stable 4.1.2 version is actually still the original Monica 1.
I'll take "things you should never do" for $1000 Alex. https://www.joelonsoftware.com/2000/04/06/things-you-should-...
Can't help but notice that this line is not true anymore in 2026: > It’s harder to read code than to write it.
I read this article and the subtext in my head is “I pointed Claude at my ten year old code in a sandbox and it couldn’t handle adding the features I wanted. So for v3 I will make it better for agentic development.” I’ve become so cynical this year.
The code base and logic is complex enough that it's hard to fix bugs and add features nowadays since they're few and far between. The test coverage is the only thing that makes it remotely possible.
I've considered doing the same, rewriting it so that AI can reason about it better. I think it would be a net positive compared to where it is today. But it's also stable, as I mentioned, so I don't have enough motivation to do it.
Just saying, I can see myself doing what this person is doing.
I read this article and the subtext in my head is “I pointed Claude at my ten year old code in a sandbox and it couldn’t handle adding the features I wanted. So for v3 I will make it better for agentic development.” I’ve become so cynical this year.
Why is it bad to make a codebase better for agentic development?
The sort of "eternal addition" glomming of code that leads to tears on older codebases... I don't really see why agentic development avoids this. And I think there are arguments that it _accelerates_ it (but I'm pretty open to nuance on that side of the fence)
I read this article and the subtext in my head is “I pointed Claude at my ten year old code in a sandbox and it couldn’t handle adding the features I wanted. So for v3 I will make it better for agentic development.” I’ve become so cynical this year.
What’s cynical about this? Everybody is modernizing their codebases for agentic development.
After screaming into my pillow about the recent Facebook f**ery and the destruction of the individual and their relationships, I had a moment of appreciation and human-ness when I read, "Built because my memory is bad.; I created Monica because I kept forgetting details about people I genuinely cared about." That's the kinda thing we should be building and supporting together. I may start using it, even if that's a h…
> I now see relationships as their own domain rather than an attribute attached to a contact I mostly skimmed this announcement, but this seems pretty obvious to me? Relationships form a graph, so it seems reasonable to use graph modeling techniques to handle them. At the least RDF triple style [1]. One thing I'd also strongly consider, in this age of LLMs, is how much you can store in text. A personal CRM seems like…