@tk i mean you can hold on to it but you'll have to balance the size of the USB stick with the year readers became available for it :gutkato_sciencisto:
I think I saw my first USB ports on a Packard Bell in high school in the late 90s. When I went to University I went all in on USB .. Microsoft USB natural keyboard, with the hub in the back, and my mouse connected to the keyboard. I don't think flash drives were common yet .. that would be a few more years. We still burned CDs/DVDs.
@sun@djsumdog@icedquinn@tk can you replace the ram and storage on them i can see maybe finding a no hard drive no ram model one day. but i don't know how idoit proof macs are these days.
I think RAM is soldering on for all the M* chips ... I think storage is too because of the encryption chip. Once x86 pulls ahead on power consumption, Apple hardware is going to be fairly garbage tier. Hopefully more laptops will support the CAMM2 standard and there will no longer be an excuse for soldering on ram.
Early USB devices suffered from various compatibility issues. They often don't work even after you installed the correct driver via a disk/disc. @icedquinn@tk
The reviews on AMDs new release have been pretty lackluster, but a lot of people have been critical of the review methodology for not correctly addressing power. These are desktop chips too. I'm curious how the laptop chips will do when they come out.
Meanwhile the Snapdragon/Win boxes seem to be a comparability shit show as well.
One of the big issues is there are only two manufactures of x86 chips. You need an Intel license for 32-bit and an AMD license for the EM64T instructions. VIA did have a license they bought from Cyrix and it's still held by some company (I think in China or Taiwan) and they do make some (disappointing) chips.
IBM, Fujitsu and others had Intel licenses, but they all stopped making chips back in the 286/386 era (and I doubt any of them could get the AMD license even if they wanted to).
So there aren't enough companies around to even try to make competitors on the same instruction set. The last big one was Transmeta way back. But if it was more open or easier to license, I think we would see at least one or two startups trying to make M* level performance on the x86 ISA.
@djsumdog@tk@icedquinn@iceloops VIA chips still are being made by some other country yeah, they only exist for industrial automation equipment. Some of them are like 800Mhz and they actually run even worse than their clock speed suggests.
oh and they also have an entire second set of instructions on it. VIA at one point tried to combine x86 and their own architecture and there are undocumented op codes that let you flip .. and they can be called in user-mode allowing a regular user application to get ring-0/protected mode access .. and these chips are embedded into point-of-sales devices and kiosks .. and it's a hardware issue and there's no fix 🤣 (it was covered at DEFCON a few years back)
@djsumdog@tk@icedquinn@iceloops interestingly it was actually discovered by a different set of researchers years before the defcon paper but they weren't as good at self-publicity!
it's not clear what the fuck the extra risc cpu would even be good for lol
@icedquinn Interestingly… I don't think I would use it. Like it would have to be somewhat useful for me both present and childhood, and like I feel like I knew my way around internet enough to find what I needed. That said maybe the only thing would be to hoard some more, maybe present me giving a list of net-media which got lost in $current_year.
@djsumdog@tk@icedquinn@iceloops@sun >I think storage is too because of the encryption chip. Mac Studios use the m.2 connector for flash modules, it's superficially similar to m.2 SSDs but most of the controller functionality is on the M* SoC and naturally they are encrypted and paired to the device.
>Once x86 pulls ahead on power consumption I'm still waiting for an x86 SoC to not lose to M1 on any efficiency-related metric by my testing. They've had years at this point, and node advantages now, but when I grabbed a Ryzen 7840U, set it to match observed M1 MBA power, and threw some of my leetcode-tier solvers at it, it was considerably (~1.2x runtime) slower.
Yes, the 7840U is much faster (~0.5x runtime) at matched power on AVX-style workloads. No, x86 is not competitive on CPU-style CPU workloads. It's ridiculous to not have a clean sweep against chips made by the same foundry 3 years prior on a less-refined node, or at least be equal on the worst cases.
One thing about the PC architecture (and that's very specific. PS4 is not a PC even though it's x86. You can't boot it like a PC and most of the hardware is ass crazy) is the standard made running Linux incredibly easy. Sure your video or audio or phone-modem might not work, but it would almost always boot to console.
I really worry about Linux getting left being with the Apple/M* or Microsoft/ARM offerings. You can take almost any Lenovo/Dell/HP offering and there's a good 80% chance all the hardware on it works with mainline Linux (unless it's super new and has like Wi-Fi7 or something). I don't want to see us getting to the point where Linux users are having to get very specific lappys just to run the OS we want.
@djsumdog@tk@icedquinn@iceloops@sun Yeah it's a bad situation for sure. Embedded-style lockdown has made its way to the mainstream computing devices, and while Android's custom ROM scene was never as free as the PC ecosystem, that has been regressing a lot over the past decade too.