Posts by fromxyzzy 155 publicly visible posts • joined 25 Jan 2019
> Want to talk an American out of government health care? Just tell them it will be as efficient as the Department of Motor Vehicles, as compassionate as the Internal Revenue Service, and as loving as the police.
And as with the 'death panels' panic from nearly 20 years ago, make absolutely sure that they don't figure out that this is how it already works with private institutions in charge except with much less accountability.
It very nearly has, there have been probably a few dozen people caught or killed already who were either planning to do something violent or actually did do something violent. The only thing that has really convinced the AI companies to work to reduce the sycophancy of their newer models is the fact that there have been several people who have been pushed towards attacking the AI companies themselves because they shut down their AI girlfriends.
If you want to know what the future looks like, keep an eye on the interviews that these execs are doing where they say they want an end to affordable computer parts and a permanent price hike. Then look at what has happened in the automobile market in the US, where people are either keeping cars for 10+ years or taking out 8-year loans.
The only real hope is that the Chinese spool up new factories and fill the void, cutting the Taiwanese/Koreans/etc manufacturers off at the knees when the AI bubble bursts. Barring harmful (to local citizens, not China) tariffs, this will happen in the automobile market as inexpensive and well-made BEV and PHEV cars undercut US and EU car makers who want to sell only high-margin and low quality products, leading to the same effect the US auto markets saw in the 70s when they had to buy in to foreign companies and captive imports or go out of business.
Just an astonishing lack of foresight, or hindsight.
As far as I can tell, Apple's real advantage in Arm has come from the fact that they design their own licensed Arm cores while Microsoft is trying to do the Android-style thing and sell Windows on Arm to manufacturers. So whenever some manufacturer wants to sell a Windows on Arm machine they need to make sure it runs on every part of their system or buy an existing/license a reference system design and repackage it. It's far more expensive to design their own cores and complete systems (and get Windows working 100%) so only the biggest chip players have tried (like Qualcomm's new Snapdragon laptops) and the off-the-shelf designs are apparently not very fast and don't get updated for years.
So Apple's vertical integration and long experience with iOS makes their shift to Arm comparatively simple and allows them to optimize and improve things incredibly quickly while the Windows world just isn't iterating anywhere near as fast because existing x64 laptops are Good Enough.
WindowMaker is one of my go-to desktops for low (computing) power machines, I've run it happily on a DEC 3000/400 with 64MB RAM in the past. I'll keep an eye out for Gershwin in the future, I tried GhostBSD a few years ago but it didn't seem to like the old laptops I put it on for testing and I moved off it pretty quickly.
Cassettes are an easy way to Make something physical and tangible, which feels mechanical and requires at least a reasonable amount of effort. About 20 years ago at the nadir of cassette sales, they were quite popular among underground and avant garde music communities because they were tangible but immediately reproducible (with a dual-well tape deck) and the time and effort of making one gave them value over the flood of CDRs that every bedroom musician with a copy of Ableton was handing out at the time.
Companies are making new cassette players, check out the KLIM brand. Unfortunately, there's only one or two manufacturers making tape mechanisms anymore so a vintage name brand one will probably still have better audio reproduction.
The old audio equipment maker Alesis (RIP) made a pair of extremely user-friendly (for the time) MIDI music making devices, a drum machine called the HR-16 after it's amazing-for-1987 16-bit sounds and a MIDI sequencer called the MMT-8. Orbital used a trio of the MMTs as the heart of their live setup for years.
Both were loved by musicians for many years after their release because in the pre-internet age, Alesis had in their wisdom seen that a big flat empty spot on top was necessary to fit the PCB in the case, and decided to give it a flip-top with a little manual explaining all of the essential functions. So, if you bought one used for 50 currency units or whatever, you could immediately get to terms with it.
Akai had a help system in some of their sampler OS systems as well, which wasn't terribly useful but an impressive inclusion for one of the more arcane bits of technical equipment ever produced.
We've abandoned a lot of the UI/UX lessons learned through experience, and ironically it was Apple who both did a huge amount of impressive work (building on IBM's internal work I believe) on strictly codifying their UI elements in order to ensure a consistently usable system across applications, and then with iOS, totally destroying any sense of consistency in interaction and hiding every aspect of the real system from the user. I run an old iBook with MacOS 9 for legacy software and tinkering and the only things that don't adhere to the Apple UI guidelines are video games which were virtually always originally for Windows/DOS. I have an iPad in the loo and I can perform the exact same swipe motion on it 5 times and get 5 completely different results for no discernible reason.
Kids that are growing up on iPad and other touch-screen devices are being taught that tech devices are magic boxes that act in ways that you can never understand because they don't respond consistently and they hide every aspect of the underlying system. Honestly, it's primed them for the advent of AI as well, where they simply trust what the magic box tells them is true and are flummoxed when told that the magic box is wrong, unreliable, and they've failed because they just expect systems to work without understanding how.
This is fantastic news, I've been so fed up with the Windows 11 that was automatically installed on my main media PC a few years ago that I hardly use it at all. It's a laptop with an RTX2070. If I can get native drivers on it I can probably get some more use out of it than I have for a while since getting anything done with Win 11 is a fight.
iOS has tried to do this as well, it fights you hard to keep you from seeing the filesystem structure. They claim it's for 'security' (which surely means 'so you can't jailbreak it') but it only really makes sense if you understand it as being philosophical.
Makes it impossible to use iPhones/iPads to get any serious work done.
The 'turn on and go' and 'jump into basic' thesis seems to be completely covered by RISC OS, right? Boots instantly, hit a key to reach BBC BASIC, runs on Pi.
Weird I searched all the comments and only one person mentioned it. Checking the Raspberry site, I also see they've removed it from their downloads page and from the list in the Imager app. Perhaps that'd be the place to start.
Ultimately, the point is that you can base your entire company infrastructure on an open source software tool or library, but there's no obligation for the maintainers to actually fix it when it's broken. If there's a zero day, they aren't obligated to do anything about it, especially not on anyone else's expected timeline. It is the perfect example of 'a lack of planning on your part does not constitute an emergency on my part'.
The people who are advocating for this kind of financial support for open source developers to maintain the software recognize this and are trying to provide an incentive, not because of a moral obligation but because to not do so leaves them totally at the mercy of the maintainers, who may consider that maintenance a low life priority. If that software is mission critical for your business, it's a smart business decision to convince those devs that it should be a much higher life priority for them, perhaps even their full time job.
Apple was working on computer-grade ARM CPUs for a long time before they switched off of Intel, and may not have ever made the switch had Intel not followed a diminishing price/performance curve into oblivion for the foreseeable future. Sure, Apple functionally has infinite money right now, but it's not as if Qualcomm can't finance parallel RISC-V development as a moderately short term drop-in replacement insurance policy if ARM tries to pull something similar.
RISC-V's issues at the moment in comparison to ARM would probably be more related to the swift development of newer standards of implementation, to the point where the newest performant RV chips are already either falling behind or not managing to live up to expectations. See the SpacemiT K1/M1 RVA22 situation. It doesn't help that most RISC-V implementor/manufacturers at this point are somewhere on the level of Rockchip, Mediatek or Allwinner in the ARM world. Building actually compliant and performant RV SOCs should not be a problem for larger manufacturers who are able to take their time getting product to market.
Chrome is particularly bad about this, which means that every web thing that relies entirely on Google's frameworks like Steam and those stupid Electron containers may have problems (even Edge because MS have absolutely no foresight), but much like the folks in the comments telling people to switch to Linux: Firefox still works on Win7 without issue.
I recently switched to Floorp, a Firefox fork with a silly name and an obsession with privacy. All the updates, none of the weird nonsense Mozilla is trying to sneak in now that Google has totally abandoned any pretense towards privacy and user control and Mozilla thinks they can get away with it.
As much as I appreciate and support what Ocean Cleanup is doing, this is the only realistic solution. Plastics are not recyclable and never were, and no amount of innovation will make them sustainable.
https://www.theguardian.com/us-news/article/2024/sep/09/us-voters-distrust-plastics-manufacturers-claims
"Research published earlier this year found that plastic producers have known for decades that plastic recycling is too cumbersome and expensive to ever become a feasible waste management solution, but promoted it to the public anyway."
"In fact, just 5% of plastic waste generated by US households in 2021 was recycled, one study found."
This is figure is not a result of people not sending plastic to be recycled, it is the result of plastic being expensive to recycle at best if not literally impossible to reuse in new products after the first generation.
SCO never really innovated or improved Xenix/SCO Unix/OpenServer after they took it over. They even wound up purchasing Unixware from Novell and doing very little with that other than maintenance. The driver situation was particularly appalling.
The reason is that they were only interested in massive corporate distribution and the hefty maintenance contracts that came with. By the 90s, Xenix and descendants provided the point-of-sale backbones for a lot of retailers and lots of commodity small-to-medium business servers. However, they put no thought in to potential competition and just assumed that customer lock-in would ensure at least a couple more decades of easy money, just as it had up until that point. They did not think that companies would ever be willing to consider using Linux, even as companies like VA Linux and Red Hat started selling and supporting hardware for the growing market and IBM began spooling up programs to investigate the potential.
So within a few years of their full takeover of the license for Xenix they were on the back foot, then absolutely on the skids, then reduced to desperate lawsuits accusing Linux of intellectual property theft.
Well said. There is one very particular thing that would have wound up being the MSX's Achilles' heel had the standard reached the sort of market penetration that Microsoft hoped (comparable to IBM compatibility - remember that this was not just a hardware standard but a software/OS standard designed by Microsoft and ASCII Corp).
The MSX memory map was left relatively free and open in order to accommodate varying amounts of RAM across model lines and allow compatibility between low-end and high-end systems. Many of the very earliest systems had 8k of RAM, which was pretty much useless but it kept costs down, allowed some BASIC programming and let you play cartridge-based games. However since the MSX standard didn't hit the market until 1983, RAM was cheap and these low end systems didn't really appeal to anyone, so dictating a minimum of 32K if not 64K would have allowed them to nail down a standardized memory map.
The result of the memory map is that different regions (and occasionally different companies within the same region) wound up settling on different memory maps. Now, the MSX standard made allowances for this and properly written software generally shouldn't have problems, but by-and-large what really happened was that software makers just assumed that all MSX machines used the same map as the dominant manufacturer in the region. This meant that, unlike MS-DOS software, a significant amount of MSX software just assumed the RAM was at a certain 'Slot' in the internal mapper and then crapped out when it turns out something else was mapped there. At the time this would have been inexplicable from a consumer standpoint and likely resulted in complaints to the software company, who probably would have just dropped the product and any future product intentions.
This gets even more fun when you take into account the other big highlight of the MSX standard, and the reason for the international popularity that wound up highlighting this problem - non-roman character sets. Since one of the main companies involved in the original standard (and the biggest market) was Japanese, they implemented a fairly thorough Kanji input system. However this wasn't quite standardized, so when they made Korean and Arabic MSX systems, things also broke in new and interesting ways thanks to the memory mapping.
My MSX2+ is a Korean one that I imported a few years back when someone found a cache of them in an abandoned school. Great specs for the system, but despite having 128k it has a very odd memory map that means that software is a crap shoot.
There's a more in-depth write-up here: https://www.msx.org/wiki/MSX_compatibility_problems
The MSX still has a massive following in the Spanish retro-computing community and a significant fanbase in the Russian retro community.
I am firmly convinced that changes like this are the product of middle managers desperate to justify their jobs, which should in truth never be anything more than maintenance. Maintaining something that works fine doesn't get you a bonus or a promotion in a dysfunctional company (something that absolutely should get you paid and respected in a company that isn't poorly managed). Breaking something that works fine, in a flashy way, will net you that bonus and recognition.
The problem is that when the company/person the code is lifted from looks for someone to sue, the buck stops with the user of the LLM.
So your theoretical engineer could get the blame and get hung out to dry, but if you use an LLM and it steals code, you're that engineer and you probably don't even realize it.
AI code is unreliable on both a functional and a legal level.
When I started getting in to building hobbyist computers a few years back, the CP/M version in ROMWBW was the only real way to go, so I have several such systems sitting around. I don't have much use for them (the building was the fun part, fighting with CF card formatting was not) and there seem to be far more options now, but I still keep a little Easy_Z80 plugged in via USB-Serial just to poke at it now and then.
