I can imagine a few more. Here's one that is 100% effective:
* Wait until your account was created > 18 years ago.
That may not be enough for you, and that's fine, this is not a governmental service or site, with a mandate to serve everyone. You can chose not to use discord, and discord can choose not to have you as a customer. YMMV.
It's the classic "make your problems worse to fix my problems" that you see across tech. Since they don't care if you get locked out or leave their platform, it's designed to make them maximally profitable. It's the same thing you see when sites "offer" to let you do pre-registration work before an event or transaction, even though doing that work at registration is often trivial.
Want to get a RealID? You can spend 30 minutes scanning and uploading your supporting documents through the website, or you can hand your documents to the person at the desk and they'll evaluate it in 30 seconds. Over thousands of people, it's rational for the organization to optimize away that 30 second review (it also gives people a recovery option if they don't bring the right documents.) For the customer who is competent enough to bring the right documents, the latter approach is way more efficient. I regularly have to put myself in the "I'm old and hate technology, so lets do this the manual way at the desk" flow even though I'm totally capable of doing the electronic flow.
RHEL is EOLing CPU architectures that CERN still relies on heavily.
> RHEL 9 mandated x86-64-v2, requiring instruction sets like SSE4.2 and POPCNT, while RHEL 10 targets x86-64-v3. CERN's accelerator infrastructure manages more than 2,200 specialised front-end computers and 17,000 embedded devices, including legacy Core 2-era and custom industrial boards, engineered for 10- to 15-year lifecycles aligned with long-term accelerator shutdown windows. Upstream microarchitecture mandates threatened to abruptly obsolete thousands of functional, purpose-built control nodes.
And they are, in fact, planning to continue to use RHEL and/or AlmaLinux on their desktop and datacenter machines, even going so far as to adapt Fedora's build system to support building Debian packages[1] to leverage their existing RPM infrastructure.
As I understand it (from reading another article[2] posted on HN a few days ago), this change applies purely to the layer of machines that more or less directly interface with their experimental hardware, which has requirements more akin to industrial controllers than workstations or servers when it comes to the expectation of long service life with limited windows for scheduled maintenance.
> This is just vandalism from badly supervised ‘agents’ which don’t know what they are doing or why. You could set this up with a short perl script, and the human setting it up would be held responsible for the spam - why is this different when it’s AI agents set up by a human and allowed to post to the internet at large?
It's not, but the courts and the legal system move slowly by design. There is absolutely legal risk for OpenAI here that will not close until the Statue of Limitations has expired.
Yes, but it's a different group of Americans, and the ones who are currently thriving in this broken system will fight tooth and nail to preserve their way of life.
And most of that income is salaries to existing healthcare administrators. Quickly ending the employment of the people working those jobs would be a significant shock.
Yeah, the only place I use SLOG is with Optane devices in front of SATA/SAS flash. Anything that needs the speed of SLOG shouldn't be on rust anyways.
I would be in a different world if my system was backing banking or other loss critical data. For all my workloads, losing 5 seconds of writes in a crash is NBD.
Isn't that kind of the point of the article? The random reads required for metadata traversal during scrubs will kill performance for sequential reads and writes as well if you don't have the IOPS to back them up, building a pool without taking this into account will make regularly scheduled scrubs a pain point.
I schedule scrubs during idle time (10pm Friday - 6am Monday), so the metadata hit doesn't affect usage, but YMMV if you don't have an idle time. I also don't generally let them get that full.
If I was building the array new, I'd probably use a special VDEV for metadata, but they didn't exist when I built my last one of these.
Disk failures don't neatly wait for idle times so investing in a special vdev (which can be added to an existing pool as the article points out) should be well worth it.
* Wait until your account was created > 18 years ago.
That may not be enough for you, and that's fine, this is not a governmental service or site, with a mandate to serve everyone. You can chose not to use discord, and discord can choose not to have you as a customer. YMMV.
reply