Forget it - we've not even got upgradable laptops, and those are far simpler to make upgradable than phones/tablets.

Phones are precision designed - every component is carefully selected based on physical size, thermal output, etc. and there's thousands of man-hours of work involved to get everything fitted into the smallest possible volume without anything overheating/shorting out.

Hardware is upgrade-able on laptops. At least, all the laptops I've dealt with. Now, it's not AS upgrade-able as a desktop, but you can get higher capacity RAM chips, bigger harddrives, etc.

See, I know that, but I also think that you could sacrifice some slimness/compactness for up-gradable components and there would be a good amount of market for that if the OS was good and the feature was marketed right. (I mean, commercials almost write themselves - show person frustrated by their phone being outdated 6 months after they get it - BAM - show happy customer with upgrade-able phone.)

To be clear, my contemplation about this had it starting as providing either replacement motherboards or replacement SoCs and flash storage (if the storage is separate from the motherboard/SoC respectively, the OS could be made to check the storage for a backup image of the SoC's storage when booting on a blank SoC, and the SoC could theoretically be used to store at least basic settings and the like when you swap the storage, though you'd have to use other storage for the rest of the data). You wouldn't be replacing RAM/CPU by itself, because that WOULD cost a shitload in R&D to develop right without it being huge, and you'd need to make your own hardware at that point, but I think replaceable motherboards/storage/antennas (upgrade from penta-band to whatever if you need more, upgrade to a 4G or 5G antenna when your carrier rolls that **** out, etc), that would be a start. I think it's genuinely doable, even for a start-up, if you manage to get just a handful of skilled embedded device hardware engineers, some good marketing, and a ****-ton of money for the hardware contacts with various factories.

So, like I said, even if serious discussion and intent started now, it'd probably take a couple of years to get to the point where you could get this off the ground, but as much as I realize it's not at all an easy concept, I think there's a small but realistic chance to make something like this work, and it sure as **** will be even less likely come to us from the large established companies out there.

True - and you can upgrade the sd card in your phone (which is the equivalent of the hard drive). I agree, making memory upgradable should be doable, but it took quite a while for laptops to start doing it as well.

I suspect that the costs won't work out. Even with laptops nowadays it's rarely worth upgrading the memory after purchase - by that time a new laptop is not much more expensive and far more powerful. Having the option for more memory initially would certainly be an improvement though!

I think you ought to talk to a skilled embedded device hardware engineer first. It certainly sounds a reasonable idea but I suspect the situation's far more complicated than it looks to a layman.

agreed. I do upgrade parts of my laptops, but it's just like buying extra batteries and sd for my phones.

What you could do is to try to define a standard form factor for a range of mobile mother cards and screens, a produce bodies that can accomodate them.
This way you could have some kind of customization, and you could change the MB/screen from time to time, or the body (w/ w/o hwk, extra space for battery, etc..)

But I think there is no way you can change the main component of the MB (CPU, GPU, ram and any other chip) without loosing lots of space and efficiency.

Maybe some king of micro PCIe could be added to the MB too, so you could add compass/gps or fancy things like that, that have small impact on tdp/volume.

@eredon: right, screens too, both the display and the touch processing part (ideally giving users options of resistive and capactive, ideally with the stantum screens if that's close to affordable). That's basically what I was thinking it would have to be if it got that far - you make the cell-phone body, use mass-market SoCs, flash storage chips, and sensors, and the only custom part would have to be the motherboard, which would just need to fit, along with all the possible 'expansion' slots (for if/when some new fancy sensor technology comes out, like gyroscopes being all the rage now), within the shape of the housing. What this would mean though, is that you can offer both 'build-it-yourself' sort of ordering for the device to begin with, and of course future upgrade-ability to users who want to take tiny screw-drivers to their expensive mobile electronics.

Well, SD card is one thing, but I think they'd be slower than a more 'proper' flash storage chip. Probably goes back to the talk-to-a-skilled-embedded-device-engineer-first point, since I don't know how much faster than microSD card you could make the bus of a freely end-user upgrade-able chip.

Well... I think you're right, but I think that's mainly a carrier-subsidized-phones-on-contract issue for embedded devices. Because at least in the US where I'm based, the biggest problem is that you could no doubt make it cheaper than the newest-hardware-phone, but you probably won't be able to make it cheaper than getting a new contract + upgrade phone. On the other hand, I'm hoping that if something like this ever got off the ground, enough effort would be put into making the OS and devices damn good, to the point where you'd rather keep the current device instead of upgrading to a whole new device. On the other hand, I'm also not saying the upgrade-able hardware 'feature' needs to be the defining and primary point of the devices made in this hypothetical openly-developed cell phone manufacturer. Focusing on delivering a great phone for the end-users with all the power user freedoms of a proper Linux stack and all the versatility of a desktop OS (out-of-the-box USB/bluetooth mouse/kb/printing support, full office suite based on Open/Libre Office with a finger-friendly UI, etc), now THAT would be the main goal, at least in my mind if I ever got enough people interested to get this off the ground. The hardware upgrade-ability would be nice, but I'm willing to admit that is niche enough of a concept that it may take a while to catch on, and if I ever had the opportunity to be involved in a project like what we're discussing, I wouldn't risk sinking the startup before it gets off the ground properly by pushing too much for the hardware upgrade-ability. I think there's a demand for that too, and it should be met eventually, but the devices wouldn't take off regardless of their hardware advantages if the OS wasn't something amazing or at least good (of course, one of the advantages of a truly open development process, if done right, is that community involvement would make the OS far more amazing than the company itself could, as we've seen with the N900, and probably the previous Nxx0s, but I wasn't around for those).[/quote]

I agree (joerg, you there? We need your genius in this field). Though, I ALSO don't see this as happening immediately, especially if I'm the only person who has any serious interest in doing this. On my own I'm sure as hell not raising enough funds in the upcoming year or too - if it looked like there was a real chance of it happened I'd put in everything I could, but as a college student who has an okay paid internship now (and may or may not be able to continue working part time for the same people over the college year that's about to start and in the future), I would currently be a drop in the bucket on the financial side of things if hardware needs to be acquired, built, tested, people hired, servers/site/etc set up.... 'course, I'd do a shitload for free, but until my coding gets better (which, realistically, it would fast if this started up full steam), that makes me a drop in the bucket there too.

/me wonders if I'm just giving this idea and myself bad PR by having mentioned it at this stage. The dominant reason I said nothing so far all the times it has come up in my head is that I know there's a very very small margin for error in pitching ideas like this, even into the anonymous mass that is forum readers. Currently it's a vision I think is realistic, but I know I can't get it off the ground myself for years, and probably would still need a couple of years even with significant rallying around the idea. I suppose I don't like all this depressing sentiment that has taken hold here lately, that Maemo is dead, we can't hope for MeeGo to be anything anytime soon due to lack of hardware maker support, and all that's left for so many of us to be happy with our devices is cling to stocked up second hand N900s and hope something free-enough comes around some day soon.

I know there's plenty of people here, and PLENTY of people everywhere if they think about all the little things they'd be happy to have in their devices, who can agree with the general gist of the visio: a mobile device line and OS to go with it that doesn't treat the user like a prisoner, maintains a good level of non-tech-savvy surface simplicity and ease while providing normal-computer functionality that phones could EASILY have (i.e. printing-to-any-usb/bluetooth printer, from normal user-space programs, out-of-the-box or from available-from-the-get-go-to-install-packages), while leaving the underlying power and flexibility of the full Linux stack and open source development available for those who want it.

Making that a reality has to start somewhere, so I might as well try throwing it out there now.

Certainly, at least TABLETS should be upgradeable.

Did ysss and vi get their n950s? I seem to be getting offers.

not me..
that last one seems pretty questionable if you ask me ;(

Whoever is sending those offers, can please stop spamming my YouTube account. Maybe I need to put the No in larger font size...