I can't say I agree with you here, if anything FPGAs and general purpose microprocessors go hand in hand. It would be an absolute game changer to be able to literally download hardware acceleration for a new video codec or encryption algorithm. Currently this is all handled by fixed function silicon which rapidly becomes obsolete. AV1 support is only just now appearing in mainstream chips after almost 8 years, and soon AV2 will be out and the cycle will repeat.
This is such a severe problem that even now, (20+ year old) H.264 is the only codec that you can safely assume every end-user will be able to play, and H.264 consumes 2x (if not more) bandwidth compared to modern codecs at the same perceived image quality. There are still large subsets of users that cannot play any codecs newer than this without falling back to (heavy and power intensive) software decoding. Being able to simply load a new video codec into hardware would be revolutionary, and that's only one possible use case.
> It would be an absolute game changer to be able to literally download hardware acceleration for a new video codec or encryption algorithm
That relies on "FPGAs everywhere", which is much further out than "GPUs everywhere".
I'm not sure where the state of the art is on this, but given the way that codecs work - bitstream splitting into tiles, each of which is numerically heavy but can be handled separately - how is development of hybrid codecs, where the GPU does the heavy lifting using its general purpose cores rather than fixed function decoder pipeline?
But why would it be amazing? The alternative right now is that you do it in software and just dedicate a couple of cores to the task (or even just put in a separate $2 chip to run the decoder).
Like, I get the aesthetic appeal, and I accept that there is a small subset of uses where an FPGA really makes a difference. But in the general case, it's a bit like getting upset at people for using an MCU when a 555 timer would do. Sure, except doing it the "right" way is actually slower, more expensive, and less flexible, so why bother?
...which are playing back video, so they're likely blowing most of their power budget on the display and on radio. I guess my threshold of "amazing" is different. Again, I'm not denying some incremental utility in specialized uses, but most of the time, it just doesn't seem to be worth the pain - especially since nothing about the implementation will be portable or maintainable in the long haul.
In the same vein, no one is writing a smartwatch software stack in 100% bare-metal assembly, although in the hands of a capable developer, I'm sure it could prolong battery life.
And you think that a downloaded codec on an FPGA would perform anywhere close to custom silicon? Because it won't; configurability comes at a steep cost.
Why not? The firmware was already public at one point. If people are analyzing your app to find an S3 bucket full of firmware, I'd assume they'd have a pretty good reason to go through the effort.
Doesn't matter really, keeping blobs hidden doesn't actually do anything except make it slightly harder to analyze the software. Making all blobs easily and readily available is exactly what I want the vendor to do. Black boxes don't make things secure.
I agree with you in principle here, but to play devils advocate, $1,000,000 isn't a whole lot of money. A worker will make around that much at $25,000 a year over 40 years. If we have to keep money/capitalism, the limit should probably be around 10-15 million. That's still pretty high, but not egregious. Give or take ~40yrs on a high FAANG salary ($375k/yr). Still firmly upper middle class IMO.
I don't mean earnings over a lifetime or career, but currently. A worker making $25,000 a year will still probably never see a million dollars regardless of the limit. Maybe everything above that is taxed 100%. I don't know.
But the point is kind of to eliminate the upper classes and scale the economy back into the reach of most people. So there would be no FAANG salaries. The cost of everything (healthcare, education, housing) would go down. It would place a hard limit on political influence that isn't too far out of reach of current Congressional salaries and would probably limit pork barrel politics and insider trading as well. It would end inherited wealth and maybe even limit the length of copyright.
That's an admittedly naive and utopian view and I'll admit there are bound to be complexities and externalities I'm not taking into account because I'm not an economist. But it's either that or we seize the means of production and put the rich to the guillotines until the sewers choke on their blood. And then something something luxury space communism.
That's because the data is inherently flawed. The poverty line this year is $15,650 for an individual. That's not poverty, that's destitution. From personal experience, living in WV, you cannot survive on that amount of money without either sleeping under a bridge/in a car or dumpster-diving/shoplifting all of your meals.
Folks say just get on food stamps or medicaid, but it's not that simple. At that level of destitution you may not have a phone, an address, basic ID/documentation, or even a means of getting to the office to apply. Means-testing makes the process so drawn-out and convoluted, that many folks (including myself) don't even bother, because there are more immediate things to worry about (once again speaking from experience).
After years of destitution, I finally managed to make a bit more than twice the FPL and I was still struggling (but significantly better off). Just recently I lost that job for reasons outside my control and after my unemployment runs out I'm back where I started. Everyone I know has a similar story. Any data that says that poverty is decreasing in the US is detached from reality.
Thanks for your story and I'm sorry it's been such a struggle to stay on your feet. I agree that too many people never make it outside of their social and economic bubble and never bother to see what reality is like for most Americans.
I'd argue it's probably time to drop 32-bit x86 support, but the rest of this stuff is arbitrary and doesn't have any tangible benefit except conveniently providing hardware manufacturers with an excuse to unload new hardware onto people when there's nothing wrong with what they have. (not to mention, pardon the conspiracy theory, they're probably trying to use the TPM to turn the PC into a smartphone-like platform)
It's surprising that when we had Win7 they did that brief "XP Mode" experiment with some virtualized-penalty box.
Why didn't that go further? Presumably virtually any x86-64 box currently in circulation would be fast enough to run a VM running a full copy of 32-bit XP/Win7/Win10, or even a full carousel (or download store) of DOS and early-windows releases. It could be the most compatible Windows ever, solving the weird "64-bit systems can't run some 16-bit apps" gotcha and perhaps allowing some way to bridge in support for devices that can only be driven by old 32-bit XP drivers.
Here's a significantly more credible (stacksmashing) video that demonstrates how ineffective some TPM implementations are. If the TPM was integrated into the CPU die, this attack would likely not be possible. https://www.youtube.com/watch?v=wTl4vEednkQ
Despite the TPM being a pretty good and useful idea as a secure enclave for storing secrets, I'm concerned that giving companies the ability to perform attestation of your system's "integrity" will make the PC platform less open. We may be headed towards the same hellscape that we are currently experiencing with mobile devices.
Average folks aren't typically trying to run Linux or anything, so most people wouldn't even notice if secure boot became mandatory over night and you could only run Microsoft-signed kernels w/ remote attestation. Nobody noticed/intervened when the same thing happened to Android, and now you can't root your device or run custom firmware without crippling it and preventing the use of software that people expect to be able to use (i.e. banking apps, streaming services, gov apps, etc.).
Regardless, this is more of a social issue than a technical issue. Regulatory changes (lol) or mass revolt (also somewhat lol) would be effective in putting an end to this. The most realistic way would be average people boycotting companies that do this, but I highly doubt anyone normal will do that, so this may just be the hell we are doomed for unless smaller manufacturers step up to the plate to continue making open devices.
I've looked into this fella before because he didn't pass the smell test. He's running a grift selling schlocky cell phones and cloud services. His videos are excessively clickbait-y and show minimal understanding of the actual tech, it's more or less concentrated disinformation and half-understood talking points. GrapheneOS devs also had something to say about him: https://discuss.grapheneos.org/d/20165-response-to-dishonest...
I own one of these devices (pinephone) and it is legitimately not good enough for day-to-day use (despite the incredible efforts of the people who are working on it's software). I only use my phone for locally-stored music, text-only web browsing and calls/SMS. The Pinephone cannot perform any of these tasks competently. The thing it does best is playing music, but this drains the battery. It will not reliably place/recieve calls/texts (and 911 doesn't work IIRC). It can barely handle basic web browsing. KDE on this device literally pegs both CPU cores to 100% all of the time. Phosh is better but still dog-slow. This is the case even with the many years of improvements the community has been making to these devices. It used to be significantly worse, and the software is monumentally better than it ever has been. I love this device, and it deeply saddens me that it has such major flaws.
All of the current Linux phones have major showstopper issues, and saying we're complaining about them being "unable to run modern PC games" is a strawman. The simple fact of the matter is there are no decent mobile Linux options available.
The most endemic problem right now is "Linux" phones that use crummy forked vendor kernels and Halium. For all intents and purposes, these devices are trapped in time and can't meaningfully get software updates for major system components. The 2 decent Halium-free options, the Pinephone and the Librem 5, both still use downstream kernels, and the Pinephone's kernel is maintained by 1 person in their spare time. I think it's apparent that this is not sustainable, and one can't reasonably expect megi to maintain this device forever.
As sad as it makes me feel to say this, I don't foresee these problems improving for a long time. As of now, I remain stuck with a Moto E6 from 2019 (Android 9.0) as it seems to be the final device ever produced with a replaceable battery, headphone jack, SD card slot, and screws instead of glue.
I don't understand what you're talking about. SXMo (https://sxmo.org/) is fast on Pinephone. Even Phosh is pretty usable. Firefox with NoScript is more than good enough to browse web sites with pictures.
Also, Librem 5 is much faster than Pinephone, and I've been using it as a daily driver for quite some time already.
Man, I just want to get a rapsberry pi and screw together a touch display screen with some sim attachment as my phone.
Or a device which can just take a X server running on the same port of sorts but I have found that sure you can do something like it, but its gonna be of inferior / subpar than a phone but definitely possible.
If you wait around to be purist on this issue all day, nothing will ever change. Something like e.g FuriLabs is good for growing the ecosystem and getting people actually exposed to something other than iOS/Android.
Halium is a hack around crummy vendors doing sub-par work. It is technically impressive but it doesn't resolve the underlying issue that the crummy vendor kernel will never be updated. Saying that Halium is not a good enough solution in the long-term does not make one a purist, it's a simple fact. Devices that rely on Halium are dead-on-arrival.
Devices that rely on Halium actually work. I’ll take that over your perceived “dead on arrival” status right now - especially since by the time anything in that ecosystem changes I’ll be ready to swap devices again anyway.
Except they don't, not really at least. You can't even run a Wayland compositor unless it's hacked up to support Halium, and that's just the tip of the iceberg. It makes sense when you want to run some better userspace on an existing Android device where having a proper hardware support is absolutely hopeless, but in the end it's not that much better than running your stuff in Termux or WSL and you'll find yourself limited as soon as you actually try to do something unorthodox with it.
I'm typing this on a device that doesn't rely on Halium and which actually actually works, without being confined to what distro maintainers happened to manage to hack up or reimplement, so it's not like there are no alternatives.
And yet I've been using these devices for 17 years now (first Neo Freerunner, then Nokia N900, now Librem 5) and they've been good enough for day-to-day use. With some compromises, sometimes effort, maybe not for everyone, but they sure were usable by a determined person who cares.
I do have a replaceable battery, headphone jack, SD card slot and screws. I do some Web browsing, reliable calls/SMS, playing music for hours. It's starting to get a bit slow and old over the years, but I still see no reason to switch to any less user-respectful device.
What I worry about is whether there will be an upgrade path within the next decade. So far there was the Liberux campaign, and it failed. I already had to use an Android device as a secondary phone for 2-3 years before I got my Librem 5 because the N900 eventually aged too much to be usable for the Web and there was nothing on the market that could properly replace it. I don't want to need to do that again.
PinePhone is a low-end device with no support other than what you get from the community. It was a good option for those who couldn't afford anything else and wanted to invest their time and skills instead of money, but there are no miracles. The community of people who did actually care turned out to be small enough that you can still find some low-hanging fruits to work on today - and that's the thing I wanted to point out. I see lots of people who talk about how much they want Linux phones, but it's a tiny subset that actually acts like it. They won't fall from the sky - not when the sales of existing devices can't finance developing their successors.
Which software stack were you using on the Neo Freerunner that was usable as a phone and had working power management?
I tried to use a Freerunner as a phone for well over 2 years before I gave up and just bought another nokia. As far as I'm aware, it was never really usable as a phone, partly due to the power management never really working properly (there was a point where we finally got power management and a battery life of >4hrs, but the phone often wouldn't wake to ring when somebody called). When using several of the available distros I was frequently mocked by my friends for using the "echophone", due to their own voice being echoed back at them, making it extremely disconcerting to talk to.
I tried a bunch of different distros. And I spent hours and hours and hours trying to tweak settings and test to eliminate the echo. qtmoko was the best distro IIRC, but it had its own issues.
To say that "they sure were usable by a determined person" severely overstates the usability of the freerunner IMO - I'll be extremely curious to hear about the software stack that you characterise as "usable", particularly with regard to the ability to make and receive calls and the ability to have the phone on standby for more than about 4 hours away from a charger.
I used SHR (initially Om2007.2, but switched after a few months as it wasn't maintained anymore). Echo could be eliminated by configuring Calypso modem's DSP and IIRC FSO distros did it by default at some point. Buzz and not waking up to ring (the infamous bug #1024) were hardware issues on early units and could be fixed pretty easily by anyone who knows how to use a soldering iron (I didn't back then, so a friend did it for me). There was a software workaround as well, though at a cost of elevated power usage in suspend. I don't remember exactly how long it lasted on battery, but it sure did last a day at school. A quick search through my e-mail archives shows people on mailing lists talking about 100 hours in suspend with modem deep sleep fixed and about 70 hours with it disabled (though I can see someone complaining in one mail that they couldn't reach more than 50 hours), but of course it could quickly burn through the battery when under active use - especially with Wi-Fi on, as I remember its power saving mode to be quite flaky.
Freerunner was the roughest of these devices, but that was more than 15 years ago. Things have changed meanwhile ;)
I tried SHR too. That original 2007.2 distro that it shipped with was almost usable as a phone before OM released the much worse one.
Interesting to hear, I never managed to get anything like that many hours out of mine - as I say I never managed a full day because it wouldn't wake from sleep to ring. And I spent a LOT of time trying to eliminate the echo but never quite managed it (though I think it might have been gone in qtmoko, it's been a long time so hard to remember exactly).
Still I'm glad to hear that it was usable for someone, I guess.
> Things have changed meanwhile ;)
I wish. But my experience with the pinephone was somehow even worse.
Yeah, Om2008 was a disaster. I liked Om2007.2 as a user, but as a developer I can see why it was abandoned. Eventually it was FSO what made the phone actually solid and with proper foundations. If your device shipped with 2007.2 still, it must have been one of the earliest ones, so you've got the whole set of hardware bugs that were fixed in later batches (but so did I).
Still, Freerunner, while usable, required plenty of patience. My current experiences with Librem 5 are so much better - but whenever I play with a PinePhone it does somewhat remind me of my old Freerunner (which still works, BTW!).
Yeah I was part of a group that was involved with a bulk shipment of one of the very first batches, we got all the warts.
Another thing that's worth mentioning was that at first, openmoko were very much over-selling the capabilities and readiness of the device: The Freerunner was initially supposed to be a "consumer-grade" device, with the neo1973 being the prototype / developer version. When I first contacted openmoko I was told that it would be totally usable as a phone out of the box with all the phone functionality you expect. They walked back on those claims and updated their website/wiki pretty quickly after the device actually came out. But not before a bunch of us had handed over our cash.
I'm glad to hear that the stack did get to a somewhat usable state. And I'm even more glad to hear that the librem is better. The experience I had with the Freerunner put me off foss phones for a long time, and the pinephone....didn't help. Maybe I'll take another look at the librem.
(which still works, BTW!)
Yeah my freerunner still mostly powered up the last time I tried, a year or two ago. I think maybe nand had corrupted and I might have had to re-flash it, or something like that. It wasn't in a healthy state but seemed to be mostly OK with a bit of tinkering.
It does deserve some credit as a cool little portable linux device - once I gave up on using it as a phone, I hacked it into a pretty useful GPS and music player device. I was still using it to record GPS tracks for trips in 2016, and I was running it as a second display attached to my workstation for some time after that. It did last quite well, I do have to give it that... But what I bought was supposed to be a phone.
I have to second this. I've bought two of these devices over the years: first the Neo Freerunner and then a Pinephone Pro.
I spent over two years persisting, trying to get the Freerunner to a state where it was usable as a phone. Openmoko were more interested in rewriting from scratch and making sure it had pretty animations than things that some might consider more important, like working power management and phone calls.
For a long time I called the Freerunner "the worst phone ever made"...
...but then I bought a Pinephone. Which couldn't even play mp3s without stuttering - something even the freerunner could manage over a decade earlier. Don't get me started on the "quirkiness" of trying to use it to make and receive calls. Also the keyboard attachment I bought with it never worked. I tried multiple distros and whatnot, but I didn't get to spend a huge amount of time experimenting, because less than a month after I started to try actually using it, I dropped it, and it was so fragile that the screen was destroyed, despite me having bought a screen protector for it.
I've looked at a lot of these devices over the years and been tempted many times. I was very put off by the freerunner experience. The pinephone experience was actually almost impressive that it managed to be somehow worse.
I've just been scanning the postmarketos wiki looking at how that works with a few different devices. The number of devices that have some feature like calls / gps / camera / etc "partially working" is dismaying, particularly for open devices like the pinephone and librem.
Personally I switched to using lineageos on phones a long time ago. It's not ideal but at least it's usable as a phone.
reply