My biggest hope for Elon’s Mars plan is the chance to create a whole new Calendar system that makes sense. Like honestly. 7 days in a week? Random days per month? Uneven quarters? Who the hell decided to put the leap day in February! Clearly it should be at the beginning/end of the year. The western Calendar is nuts.
Why doesn’t every month have 30 days with the last day of the quarter having 31? Ohh leap year? December 32nd or January 0.
Little bit of both. Pi still uses a sort of unique boot sequence due to it’s heritage. Most devices will have the CPU load the bootloader and then have the OS bring up the GPU. Pi sort of inverts this, having the GPU leading the charge with the CPU held at reset until after the GPU has finished it’s boot sequence.
Once you get into the CPU though the Aarch64 registers become more standardized. You still have drivers and such to worry about and differing memory offsets for the peripherals - but since you have the kernel running it’s easier to kind of poke around until you find it. Pi 5 added someone complexity to this with the RP1 South Bridge which adds another layer of abstraction.
Hopefully that all makes sense. Basically the Pi itself is backwards while everything else should conform. It’s not Arm specific, but how the Pi does things.
It’s not about putting data centers into orbit. It’s about the cost-yield inversion to data centers cooling infrastructure that happens at terawatt scale. All things being equal - a chilled circuit performs better and produces less heat than a hot one. There is a high up front cost to pre-cooling but if you can get in the -60C range, and stay there, you can increase performance and cut energy costs.
When they say data centers in space - they mean data centers you can’t get to because they are flooded with ultra cold dielectric fluid and it costs tens of millions of dollars to bring them back up to human temperatures.
Right now it’s not worth the hassle. At terawatt scale it’s almost mandatory.
When you walk down that line it’s pretty close to putting them in space. No access. Super cold. No air. Tiny, insulated capsule. Thermal management hell. They’ll be buried in mines though, not launched into orbit.
It’s just corporate propaganda to simplify an otherwise insane situation.
This is why I hate digital thermostats. With the old classic round Honeywell thermostats you could turn the dial a fraction of a degree when nobody was looking and "boil the frog" to get a reasonable temperature. With digital thermostats, you can only change the temperature in discrete steps which will be immediately noticed.
Use home assistant, and program in a second stealth thermostat controlled by the first, that allows you to 'nudge' the values.
It's what I did, not because of relationship reasons, but the hvac and furnace thermostat disagreed on what temperature 23C should be so I had to tweak it.
The flip side is that, if you do hammer out an agreement on what the thermostat should be set to, with an analog thermostat, you can have arguments about whether it is actually set to that.
"We agreed it would be set to 74!"
"It IS set to 74!"
"No, it's set to like 74.2 or 74.3 or something! The little pointer is not pointing directly at 74, and you know it!"
I have an analogue thermostat in my home, but vacations (in rental properties) with the in-laws turn into thermostat wars. I particularly don't appreciate the ones that use proximity sensors to light the thermostat display's backlight. Whoever came up with that idea was a genuine asshole.
Besides, would you really break off a relationship over something so petty as temperature preference? The people who find somebody who's literally perfect for them must be very rare, I think most people have to make small sacrifices and concessions.
I agree, everyone makes small sacrifices and concessions to the people they live with, and I would never break up with someone over such a small issue as temperature preference. But trying to trick your partner or housemate into thinking you haven’t changed the temperature? That’s the kind of strategy you use when you’re stuck with someone you can’t communicate with, or don’t respect enough to want to communicate with, or have given up on communicating with. At that point I’d be packing my things.
A younger me would have had the same gusto. Age has taught me that attempting to improve the AC, in ways that my family can neither appreciate or understand, is merely going to lead to disaster.
I don’t really know who Tim Bray is and until now I had never been to Grokipedia. I don’t really like Grok - I tried Superheavy and it was slow, bloated and no better than Claude Opus.
But I have a bad habit of fact checking. It’s the engineer in me. You tell me something, I instinctively verify. In the linked article, sub-section, ‘References’, Mr. Bray opines about a reference not directly relating to the content cited. So I went to Grokipedia for the first time ever and checked.
Mr. Bray’s quote of a quote he quote couldn’t find is misleading. The sentence on Grokipedia includes 2 referencee of which he includes only the first. This first reference relates to his work with the FTC. The second part of the sentence relates to the second reference. Specifically on Grokipedia in the Tim Bray article linked reference number 50, paragraph 756 cleanly addresses the issue raised by Mr. Bray.
After that I stopped reading, still don’t know or care who Tim Bray is and don’t plan on using either Grokipedia or Grok in the near future.
Perhaps Mr. Bray’s did not fully explore the references or perhaps there was malice. I don’t know. Horseshoe theory applies. Pure pro- positions and pure anti- positions are idiotic and should be filtered accordingly. Filter thusly applied.
Relevant text: Serving as the FTC's infrastructure expert, he testified on technical aspects such as service speed and user perceptions of responsiveness, assessing potential competitive harms from reduced incentives for innovation post-acquisition; his declaration, referenced in court filings, emphasized empirical metrics over speculative harms.[49][50]
Paragraph 756: Tim Bray, the FTC’s proffered infrastructure expert, opined that “[u]sers’ perceptions of how quickly an online product responds to requests is an important component of the quality of their experience,” and that the delay between a user request and an online product’s response is commonly referred to as latency. Ex. 288 at ¶ 98 (Bray Rep.). Mr. Krieger testified that Instagram saw a “significant latency reduction post-Instagration,” a term referring to Instagram’s migration to Meta’s data servers. Ex. 153 at 76:24-77:5, 287:3-20 (Krieger Dep. Tr.). He prepared a presentation in 2014 stating that there was a “75% latency reduction in our core ‘hot path’ in rendering feeds” after the integration.
With these sources in hand I think I am convinced of Tim's original point; I first looked at that ftc document and didn't see anything about user responsiveness in it.
The fact that it's not mentioned and that the main points he made in that summary have nothing to do with user responsiveness is evidence that Grok did a crap job summarizing his role in the FTC's cas, right?
My understanding is that model are already merely a confederation of many smaller sub-models being used as "tools" to derive answers. I am surprised that it took us this long to solve the "AI + Microservices = GOLD!" equation.
What's fun is that I have had Anthropic's AI support give me blatantly false information. It tried to tell me that I could get a full year's worth of Claude Max for only $200 dollars. When I asked if that was true it quickly backtracked and acknowledged it's mistake. I figure someone more litigious will eventually try to capitalize.
Modules provide more than just speed. Compile time benefits are great and the article is right about build-time bloat being the bane of every developer. But modules serve a deeper purpose.
Explicit sub-unit encapsulation. True isolation. No more weird forward declarations, endlessly nested ifdef guards, or insane header dependency graphs. Things just exist as they are separate, atomic, deterministic and reliable.
Modules probably need a revision, and yes, adoption has been slow, but once you start using modules you will never go back. The clarity of explicitly declared interfaces and the freedom from header hell fundamentally changes how you think about organizing C++ code.
Start a new project with modules if you don’t believe me. Try them. That is the largest barrier right now - downstream use. They are not supported because they are not used and they are not used because they are not well supported. But once you use them you will start to eagerly await every new compiler release in hopes of them receiving the attention they deserve.
Use a feature that doesn't really work, after so many years, hoping compilers will actually make it work if people use it - seems like a disastrous plan. People have tried to use modules, and have generally found that they fail, and have dumped them.
It's unlikely at this point modules in their current form will ever be anything but a legacy feature in C++. Maybe someday a new implementation will arise, just like noexcept replaced the failed "throws" declaration.
Thing is (correct.me if Im wrong), that if you use modules, all of your code need to use modules (e.g. you cant have mixed #include <vector> and import <vector>; in your project). Which rules out a lot of 3rd party code you might want to depend on.
you wrong
You can simply use modules with includes.
If you will #include vector inside your purview then you will just get a copy of the vector in each translation unit. Not good, but works. On the other hand. If you include a vector inside the global module fragment, then the number of definitions will be actually 1, even if you include it twice in different modules.
Well, the standard says you can, but it doesn't actually work in practice in msvc, which is the only compiler that's supported modules for over a year.
The only thing that forward declaration buys you is avoiding to include a header, which is not that expensive in C, but thanks to templates something as innocent as including a header can become infinitely expensive in C++.
Forward declaration breaks dependency chains so that you don't need to recompile vast swathes of your codebase anytime an upstream header file is modified.
I would argue they need to go even further - warranty. I’ve seen vehicles 100% maintained by contractors need full engine or power train replacements after 5000 miles of light driving. Never been off road, never been deployed, just shuttling people around the base completely inoperatable after less than 2 years. When you buy a car the manufacturer has some stake in the game still - even the limited warranty ensures that a major failure will cost them and not you. Government contracting doesn’t work like this. They sell a product, sell you service and then sell you parts. If that part is bad you, the American taxpayer, then has to buy another one. There is no such thing as returns or lemon laws when it comes to government purchases. You get what you get and hope for the best.
I am in the military product industry, the military absolutely demands we provide them warranty for the purchase, typically 3 years on our particular products. We sell the same products outside the military with only a 1 year warranty.
And we absolutely accept US government RMAs and replace product under warranty as we get it.
Jeez, we even replace product that they've clearly played shotput with and thrown off cliffs.
What sort of vehicles are those? I thought that the military purchased regular civilian vehicles for light duty road transport outside of combat zones. Do those not come with the regular powertrain warranty?
Not GP, but I suspect the "100% maintained by contractors" part of the sentence has something to do with the explanation to an otherwise absurd situation.
When you consider the sheer size of the US defense budget, and the nature of government contracts in the first place, it would be more surprising if this sort of thing didn't happen at all.
One thing that happens a lot in DOD is that you'll purchase something that, to an outsider, you'd think you just went to a store and bought. But no, the DOD goes through a contractor to procure it and often pushes the warranty/maintenance onto those same contractors instead of the original producer. It results in some odd, and often bad, situations.
You buy a bunch of HPE servers from HP? Nope. You buy them from Fly-by Night Contractor who won some contract and didn't document the HP support contracts they are wrapping their own support contract around so when it gets handed off at the end of their contract, you're SOL and can't get support from FbNC or HP without going through a lot of red tape. And by the time you succeed in identifying who is responsible for what, you're out of support anyways.
It's dumb, and happens way too often. DOD should be purchasing straight from those major vendors instead of purchasing through a contractor.
Not only the DoD by a long shot. I remember idly mentioning in a meeting that the same $50 USB thumb drives that we were buying from our vendor could be purchased at Office Depot for < $10 and without the paperwork and signoffs of generating a Purchase Order.
No one, including the Buyer cared. That was when I learned that buying from an Approved Vendor had nothing to do with what they charged.
In the case of light duty vehicles if the contract includes maintenance services then those are rolled into the contract price, which is presumably the lowest bid. What's the problem?
Exactly, what's the problem. I know more things on the IT side, and there the lowest bidder is always the best at failing, but also the one selected most often because they check enough of the right boxes.
Even the civilian vehicles tend to go through a different fleet vehicle process, sometimes even with different features. Many times fleet vehicles do not get the same consumer grade warranties, even for civilian companies.
Why doesn’t every month have 30 days with the last day of the quarter having 31? Ohh leap year? December 32nd or January 0.
reply