This is the Dropbox problem. People are willing to pay for convenience, and tech folks tend to underestimate how much convenience comes from seemingly simple solutions

this is convenience for tech folks and the price isn't a few bucks a month but 1000x.

It's like if Dropbox was an rsync server (no app) and it cost $10,000 a month for 1TB of space. Think it would still take off?

It may not be their duty to filter it out, but it should definitely be their duty to not take money to bump it to the top of their results. Let the algorithm dump random unlinked medium posts on the 5th page where they belong

Poe's Law strikes again

Telegram which famously didn't have _any_ end to end encryption for ages, and even now only has very limited opt-in "secret chats"?

Yeah Telegram only has 1:1 opt-in E2EE, that you can't use across your devices, so either you or your buddy quickly gets tired of whipping out their phone when they're sitting at their laptop, and just replies you through Telegram's non-E2EE cloud chats, and that's the backdoor. The user activated it. It's "their fault".

It's a bit vague, but the idea is right. If your SaaS is built with AI, then any customer you have can also build it with AI, and whatever they build is going to be better suited to their needs and will run cheaper because they aren't paying your margin. AI skews the build vs buy curve massively, because it makes building so much easier

This completely ignores that a lot of products distill expertise into something manageable for the end user.

And that the actual act of these 3rd parties offering said products maintains not only the software, but the knowledge required to build it.

But not everyone else is the problem? OTel works fine for humans. Sometimes AIs are just shit

It's not a new thing to bring up that OTel is difficult to get correct. This was a criticism levied before the AI era.

That is a wild claim my dude. Some of the comments here would challenge the claim that otel has worked pretty well for humans.

How are those "bugs" not immediately disqualifying? "Move fast and break things" is not an acceptable strategy for controlling 2 tonne bricks hurtling down the freeway

Homebrew has been using git in the backend to manage its database of package formulas since its inception. No reason it wouldn't work here as well

Homebrew's built a package manager on top of git. I'm talking about platforms that generate built artifacts and have package managers with dependency resolution to fetch them.

Anecdotally, Claude Bug Bot has actually been super impressive in understanding non trivial changes. Like, today, it noted a race condition in a ~1000 line go change that go test -race didnt pick up. There are definitely issues though. For one, it's non deterministic, so you end up with half a dozen commits, with each run noting different issues. For a second, it tends to be quite in favour of premature optimisation. But over all, well worth it in my experience

I haven't used the bug bot, but I like asking claude code to just review my PR in the command line. Yesterday it found a bug in a data structure I was implementing (it didn't support ZSTs properly). Of course, the fix it suggested was completely wrong, but what are ya gonna do. Still saved me from embarrassing myself before asking for a review

Pinning certs has generally been discouraged for a while afaik. It's pretty trivial to bypass, at least on Android where you can side load easy, and it's a pain in the ass to manage with a huge potential to just take down your app if you mess it up