So (unsurprisingly) I modified the image to make my point. I don’t know whether you will find it convincing with this revealed: I flipped the image because I too understand now (I did not when these switches were first introduced) that right = on. And I also desaturated the image to mimic colourblindness (achromatopsia). (Indeed, when they were first introduced I remember them being two shades of grey…)
I believe that, but… why should that be the case? It’s a convention without a reference or motivation. I remember when they were new and I really had nothing to refer to and just having no idea whether they were on or off. The fact that I’ve got used to them now isn’t really that forgivable. In contrast, checkboxes not only have convention behind them, but mimicked filling in paper forms which many people were and are familiar with anyway. The idea that “filled” is on and “empty” is off seems inherently more intuitive even if you’ve never filled in a form if you just know that the concept of forms that you fill in exists.
GNOME settings pages don’t have “apply” buttons. When do you think a selection from a drop-down or numerical selector takes effect?
Yes, most settings now take effect immediately, and that’s great. (I think KDE still prefers a separate “apply” step though). That is (still) separate from the decision of how to style something which turns something on and off, which is what I’m complaining about. I don’t take any issue with having things apply immediately.
You raise visual impairment, which is exactly why I’m complaining. Look at the image and tell me, which of the controls is on, and which is off?
Taking effect instantly is not really indicated by the control shape; it’s indicated by whether or not the form has a button equivalent to “apply”. Settings pages with checkboxes that applied immediately have been common for years; this distinction is not nearly as clear cut as you make out. I suspect what is going on is that both toggle switches and the removal of a separate apply step has gone on gradually at the same time.
But a good thing to think about is all the other controls: drop downs, text entry boxes, date pickers - these have no second version which might apply instantly or not! So it’s a mistake to think that information is conveyed by the look of the control.
Eh? What do radio buttons have to do with anything?
The styling of a UI element - whether it’s a box that gets an X or tick in it, versus a little thingy that moves left and right - is wholly unrelated to any aspects of implementation, including whether the effect happens immediately or not.
Checkboxes that look like left/right toggle switches are the worst. And the only way to know whether left or right is on is colour?! Can you please get in the fucking sea?
You ask too much.
Sleep on Linux is never quite as reliable as it should be, and I end up with a laptop gently roasting itself to death in my bag more times than I’d prefer.
From my limited journal submission experience, some at least just want a PDF. They’ll re-set the thing anyway.
This actually looks good. Though at this point I get annoyed at anything that doesn’t make an effort to be compatible with Markdown - idk if there’s any good reason they chose # as the symbol for introducing code lines, but surely it makes sense with the ubiquity of markdown to make that start a header. Similarly with * for emphasis and whatnot.
But it genuinely looks like a good effort to dispense with the backslashes and braces of LaTeX yet retain a coherent and comprehensive mathematical output.
What is surprising is that it doesn’t appear to ultimately just call through to TeX. That probably means it has a chance of having good error handling and decent performance… I will have to give it a try.
There were already plenty of reliable compilers at least for the main architectures in use. Replace C with Fortran though if you prefer - complaining about python in mission critical software is a brain-dead take that belongs in the bin of history.
It’s not a “narrative”; it’s their experience. I don’t have the same experience, but do have experience of myself and colleagues using LLM agents effectively and doing more work reviewing their output than writing lines of code. Some colleagues are pretty much AI boosters, but most are very aware of its limitations.
Developing mission-critical systems in scripting languages
This is a wild take. If you’d come up in the 80s you’d be complaining about using C instead of hand-writing assembly.
Do you not think rants belong here? What about linking to a rant on a blog?
Looks like the chipset/graphics driver is crashing. That can be because of the driver or the hardware.
It will be hard to diagnose, but you can search for the most detailed of those log lines together with your laptop model and see if that yields anything. There problem is that it’s never possible to know whether you have a software issue or a hardware issue that is exposed by particular software.
You can try installing a completely different os (i.e windows) to see if the same problem occurs - if it does you can be fairly sure it’s hardware.
No u.
An almost-complete lockup on Linux is basically always due to running out of memory and having to hit swap. A system can run at 100% CPU and still be usable, but when it hits 100% memory, it will not be usable. For a desktop system, that means keystrokes, if they are registered at all, won’t be registered until minutes have passed. For a server, it will mean all requests time out.
Unfortunately, Linux’s approach to memory management firstly allows this to happen and secondly fails to solve it once it does happen. What is supposed to happen is that the “OOM killer” wakes up and kills off a process to free up memory. That may theoretically happen if you left the machine on for a year, but what actually happens is that the amount of memory needed to run programs exceeds the amount of physical RAM, but swap is still available, so the OOM killer doesn’t give a shit. At this point many, many operations in programs are taking several orders of magnitude longer than they should do because instead of fetching a value from memory they need to:
So while your PC is running 100-1000x slower than it normally would, the OOM killer is doing nothing. If you manage to consume all your swap space, then, and only then, will the OOM killer wake up and kill something. It may kill the right thing, or it may not.
The modern approach is to use a user-space OOM daemon which monitors memory and swap usage and aggressively kills processes before that happens. Unfortunately, this tends to result in killing your (high-memory) web browser, or the whole desktop session.
Sucks. Get more RAM for your sever maybe.
I think the reason is obvious.
They want to be less dependent on US tech. Using something whose code is hosted on US tech is far less dependent than using something whose entire everything is American.
I’m sure we can all agree that they are not as important as the code, though ;)
Even if you’re skeptical about the AI part, this is desperately needed. Preferential procurement though is probably not enough to make companies think twice about selling out to US big tech - I think we need governments to intervene and prevent such sales in key areas, and I don’t see anything about that in the summaries.