I'm confused why BPF exists in the first place. Can't we just compile kernel modules that hook into the tracing infrastructure?
It seems like a webassembly for the kernel but local software has the benefits of knowing the platform it is running on. I.e. Why compile C code to eBPF, when I can just compile to native code directly?
I can potentially see it solving a permissions problem, where you want to give unprivileged users in a multi-tenant setup the ability to run hooks in the kernel. Is that actually a common use case? I don't think it is.
Your raid array hasn't failed you until it does one day. Make backups, they don't have to be off-site. Don't wait, speaking from irreparable experience.
Another solution to this problem would be to extend DNS with port information. For example, if http for foo.com is hosted on a multi-tenant server, a DNS entry that says "access http for foo.com on port N."
Satisfaction or non-satisfaction is totally orthogonal to whether the world is a "better" place or not. Lots of examples of people who were never satisfied who made the world a better place, e.g. Steve Jobs.
His initial argument is that it's okay to be mediocre because continuous success is just a "paradigm." That much was already clear. His argument falls apart when he tries to justify having enough, since that contradicts his initial argument. If continuous success isn't written in stone, then neither is having enough.
Nokia's anecdotal demise isn't a reason to not seek continuous success. In fact, one could use that same anecdote to argue that having enough leads to failure.
Distance between a founder and the product isn't an argument for not seeking continuous success. Some people prefer that distance and consider separation of labor to be more efficient.
It's clear that DHH has some bias against success. Given articles like this and his public bet that Facebook's share price would fail. It isn't based in logic though, just ideology.
It seems like a webassembly for the kernel but local software has the benefits of knowing the platform it is running on. I.e. Why compile C code to eBPF, when I can just compile to native code directly?
I can potentially see it solving a permissions problem, where you want to give unprivileged users in a multi-tenant setup the ability to run hooks in the kernel. Is that actually a common use case? I don't think it is.