3 ms·
All the open/pro-privacy/pro-consumer cell phones I've seen (pinephone, Fairphone, Librem) never go far enough (non-open cellular modems/wireless chipsets) whil
by autoexec 3y ago
All the open/pro-privacy/pro-consumer cell phones I've seen (pinephone, Fairphone, Librem) never go far enough (non-open cellular modems/wireless chipsets) while also being either insanely priced, unobtainable, barely functional, scammy, broken, or some combination of those things.
I really want a product like that to succeed, but I'm just not seeing anything reasonable being produced.
- NoboruWataya 3y agoProbably because developing and making those things is extremely expensive. The "traditional" smartphone manufacturers have far lower per-unit costs because of economies of scale and because they externalise many of the production costs by using low-paid labour and environmentally destructive practices. (Some "open" phones probably externalise these costs as well, but Fairphone in particular is so expensive because it tries to avoid doing that.) Not to mention that those lower production costs are then further subsidised by the manufacturers' efforts to extract value from the user in other ways. Every time someone tries to make an open/pro-privacy/pro-consumer phone and doesn't produce a perfect product they get heavily criticised, but I think they are up against some fairly fundamental limitations. The fact that there isn't really a satisfactory option out there despite numerous attempts suggests to me that the problem is not with the specific products or companies but something more fundamental. I don't think it will ever be feasible for ethical phones to compete with non-ethical phones on cost.
- autoexec 3y agoI'll agree that major smartphone manufacturers will probably always be able undercut them on price. I'd be more forgiving on cost if they actually delivered a quality product, especially one with open chipsets.
- mixmastamyk 3y agoHow many open modem chipsets that work are there? Honestly even if one exists it’s a luxury until there’s (other) hardware and software that actually works. 1. Get it to work 2. Sell them 3. Profit 4. Reinvest in hardware That’s what Apple did.
- jancsika 3y ago> (non-open cellular modems/wireless chipsets) > barely functional > insanely priced > I'm just not seeing anything reasonable being produced. It's not reasonable to offer an affordable phone that also runs an open source baseband OS that has been tested to work reliably with the most popular carriers.
- okanat 3y agoThat's made almost impossible by the bigger players in the SoC world like Qualcomm and Samsung. They created a multi level oligopoly which can only be challenged with nation state support (like Mediatek and Huawei did). The problem is the drivers and documentation for those chips are very limited and short-lived. One simply doesn't get it without signing an NDA and buying some high thousands. Without documentation it gets harder to optimally program them. On top of that add the know-how. Due to both NDAs and pure difficulty of working with low power systems, it is really hard to hire engineers who are both very knowledgeable and not bound by NDAs and willing to work on community projects that doesn't pay the significant wages that the consumer tech giants do. Desktop Linux can get away with it if the desktop computer consumes slightly more power or laptop battery life is 2 hours shorter. People accept it. For phones any extra operations cause significant power usage. Everything in the stack needs to be optimized for it and regular Linux desktop stuff often isn't. Android has much tighter controls but it is 5-10x harder to program in the lower levels which again limits the supply of the engineers. This is just getting a simple SoC and running Linux on top. Add similar complexities for every single peripheral.
- autoexec 3y ago> That's made almost impossible by the bigger players in the SoC world like Qualcomm and Samsung. They created a multi level oligopoly which can only be challenged with nation state support (like Mediatek and Huawei did). Which is why we need new chips to replace qualcomm/samsung/mediatek entirely in open devices. It's kind of crazy that nobody is willing to even attempt it. I'd gladly take the inevitable massive hit to performance if it meant fully open, documented, and auditable devices.
- giantrobot 3y agoFully open wireless basebands are unlikely if not (legally) impossible. Modern radio front ends have significant software control/definition from the baseband chipset/firmware. They'll happily run outside of regulated limits (ERP etc). The only reason they can get Part 22/15 licenses is because their firmwares are sold locked down and can't be adjusted outside of regulated limits. Anyone that gets halfway interested in trying to make an open baseband, necessitating a custom radio front end, gets quickly uninterested once they look at the development costs and then needing to get a license.
- xethos 3y agoThis [0] is the closest we have so far - almost entirely FLOSS, and it consumes ~1/4 (IIRC) the power of the stock firmware. It's also never going to be the default Pine64 ships with, because no OEM wants to tell the FCC "We asked the users nicely not to send erroneous signals. No, this was not lab-tested, nor verified by the FCC to not interfere with other devices." Considering the legal limitations, I think this is an absolutely amazing point to have gotten to. OP may have issues, but he mentioned even his Android handset had similar issues. I have a pretty good carrier (for a carrier), amd have pretty well no issues using my Pinephone Pro as my actual smartphone. I just carry a spare battery :) [0] https://github.com/the-modem-distro/pinephone_modem_sdk https://github.com/the-modem-distro/pinephone_modem_sdk