kiitos

38 ms·

571.
by kiitos 3y ago
> Web of trust protocols are a decades-long solved problem Solved in a technical sense, maybe, but not in any meaningful sense. Statistically zero people use any web-of-trust based system for anything useful. But we've already playe
572.
by kiitos 3y ago
Any system that expects to be widely used must delegate trust to some singular and addressable authority, and that authority must be able to remove malicious users (and many other similar things). De-platforming is a feature, not a bug.
573.
by kiitos 3y ago
Yeah, but we course-corrected on those mistakes. That's how civil society works. We (necessarily) delegate trust and authority to higher-order abstract entities (i.e. the state) and make sure we have ways to influence how they behave (
574.
by kiitos 3y ago
Less "adversary" as evaluated by a specific entity, more "bad actor" as evaluated by the collective. Which is the intent. Of course nobody issuing sanctions specifically intends them to be criminal penalties for local (t
575.
by kiitos 3y ago
Sanctions are an important component of our society, in the broadest sense. They are net good, not bad.
576.
by kiitos 3y ago
Identity (in any meaningful sense) must always delegate trust to some kind of issuing authority. If for no other reason than because any humane system must always accommodate users who forget their passwords, lose their private keys, etc. K
577.
by kiitos 3y ago
> The benchmarks show the following results Where are the benchmarks?
578.
by kiitos 3y ago
Redis is an in-memory cache by definition. If you don't want to cache in-memory, then don't use Redis.
579.
by kiitos 3y ago
The entire value proposition of Redis is that it operates out of memory, and therefore has memory-like performance. (edit: And doesn't provide the benefit of, and therefore pay the costs related to, ACID-like consistency guarantees.) I
580.
by kiitos 3y ago
Well, of course! Redis is not (and has never been) a database, it's a data structure server, at best described as a cache. If jobs are added faster than they're removed, this is straight queueing theory 101 -- ideally you'd r
581.
by kiitos 3y ago
In the browser, fine, but what about in the terminal?
582.
by kiitos 3y ago
ULID has many advantages over UUIDv7, including (but not limited to) higher timestamp granularity, and more usable canonical string representation.
583.
by kiitos 3y ago
Literally everybody does this. It is actually a critical property of usable IDs!
584.
by kiitos 3y ago
If logs contain privileged info, then the damage is done the moment they're published. Whether or not they're downloaded to a laptop is irrelevant, the risk impact is the same.
585.
by kiitos 3y ago
You think you've heard it all, but then someone says that introducing undefined behavior improves software quality, and just wow, man, I don't know.
586.
by kiitos 3y ago
I guess this would be relevant if Apple had a monopoly in the phone market.
587.
by kiitos 3y ago
Because if you allow apps that haven't been vetted by Apple to be downloaded and executed via the app store, then grandma is inevitably going to run some malware that drains her 401k to a teenager in Russia.
588.
by kiitos 3y ago
Adding more choices, especially for fundamental stuff like we're discussing here, actually literally does make the device harder to use.
589.
by kiitos 3y ago
That the iPhone is a general-purpose computer is an implementation detail, not a product feature or guarantee.
590.
by kiitos 3y ago
Security is a consideration, but ultimately it's a design decision for the product. Companies can generally make design decisions for their products as they see fit. Apple is no more obliged to provide the escape hatch you describe for
591.
by kiitos 3y ago
If you want to treat your phone like a general-purpose computer, that's fine, but the iPhone doesn't work that way, very much by design. I understand that you want a different user experience, but them's the breaks.
592.
by kiitos 3y ago
> Excuse my French, but uh what? A browser accessing a bank in a Linux virtual machine running on bare metal is by far more secure than desktop MacOS running on bare metal. Facts are most definitely not in evidence for this claim.
593.
by kiitos 3y ago
I actively don't want third-party payment systems on my phone. I also actively don't want the ability to load and run arbitrary third-party applications on my phone. Adding these things would make my user experience worse, not bet
594.
by kiitos 3y ago
> These are not fringe things that are impossible to build. Maybe not, but I'm confident that the system you're describing is impossible to build in a way that is both general and efficient.
595.
by kiitos 3y ago
There are an enormous number of tools that meet these requirements, most obviously Nomad. But really any competently-designed system, defined in terms of any cloud-agnostic provisioning system (Chef, Puppet, Salt, Ansible, home-grown script
596.
by kiitos 3y ago
> We have an on prem elastic db pulling in tens of thousands of events per second. Many kinds of systems can work great at this scale, which is non-trivial but ultimately not particularly large. > It's really just 6 servers with
597.
by kiitos 3y ago
You should never need to sample telemetry data.
598.
by kiitos 3y ago
It is a tragedy of the current generation of observability systems that they have inculcated the notion that telemetry data should be sampled. Absolute nonsense.
599.
by kiitos 3y ago
If you're OK with a single point of failure, sure, maybe.
600.
by kiitos 3y ago
The line is fuzzy. Containers are processes, but sandboxing requires more than just FS/chroot isolation.

More ›