Of course they do. See Zuckerberg's episode on Joe Rogan.

The problem with the internet is that comments like this appear to carry equal weight to comments from people who actually work in this sector.

That "If" is carrying a lot of weight.

I think the point was a single file for bit by bit replication. Managing infrastructure isn't good ROI here.

The single file bit-by-bit replication is only possible if those bits still exist and you have guarantees of it. Those bits are a hidden dependency unless you mirror them. Without a mirror it seems like replication will work most of the time, but will eventually fail in ways you don't want.

Your comment seems to have nothing to do with the original comment.

I sense a lot of frustration in your comment. Perhaps it would do you well to journal.

This is not a Show HN

How's OmniWM?

I've used paperspoon but it's very laggy. Paneru is better but 3 finger swipe up or down shows all windows on all workspaces which makes it impossible to pick the window you want. (which paperwm didn't do)

I used OmniWM for a year (and even sponsored it) and functionally, it was great. However, the project velocity was too much (partially AI-driven, partially the inability to say no) and I was constantly forced to keep up with breaking config changes (just use an LLM to fix your config — a GH issue literally said that) and Sequoia compatibility was added, then taken back away, which was the last straw for me. I switched to the fork Nehir and things… just work. It’s got less going on, but it’s very stable and nice.

This is my concern, OmniWM is bundling more and more functionality in, the lead dev is a bit too keen on using tokens. I may move to Nehir similarly.

OmniWM is pretty snappy! Lots of unhandled cases that aren’t even really edge cases. Floating windows and custom menus from the toolbar are poorly handled, as is window minimization (the BSP space for a minimized window remains held). Seems to be moving in the right direction, though.

Reminder: Cloudflare doesn't have limits on these server so one day they hit you with a "Hey, you're using more than you should, pay $x,000 or we'll shut you down within 48 hours"

> We're prototyping hard spending caps now, with early availability in Q4 2026

That's for paying customers. Free customers have no spending caps, they'll only tell you how much you have to pay them immediately after you're past their secret internal caps.

Yeah I asked Bill Gates what he codes with in our private chat and John Carmack says he used all of them at the same time but only to check his own code while Torvalds told me he only uses it for side projects.

Nasty. My point was about.. this morning my machine crashed. Pi + Deepseek v4.1 Flash misdiagnosed it, and Opencode + Deepseek v4.1 didn't.

3 people were killed accidentally during IDFs attempts to free the hostages. That's not the same thing as "shot all of the hostages they did release". In fact it's the opposite. Because those people were not released, the IDF went in to try to get them.

Citation needed. Since there are plenty of freed hostages that have done interviews, this seems falsified on its face.

Like I told you, you can easily ask chatgpt for citations if you wanted the truth but you're more interested in hasbara.

Maybe you should try that too since chatgpt pretty clearly says you're wrong. 3 hostages were shot during a rescue attempt from friendly fire. That's a very different thing from what you said.