I like the idea to be fair, but it's a concept that could never work commercially.
I don't even know where to begin tackling the faults to be honest.
Firstly, anyone who buys into the project technically is putting themselves at a disadvantage in the market. Before you even have a product to sell, you need everyone who is invested to agree on a set of standards. This can take years. By the time the standards are achieved, the technology in those standards are all those years behind anyone producing a competing phone.
This is the big one really. Look at USB. It's been with us for over 20 years now and it took ages before 1.1 worked as a standard (1.0 was released but didn't work, so everyone had to go back and redesign their USB stuff). We're now only just seeing the third iteration of USB and while it's a standard and much better than USB 1.1, it's so much slower than other data connection types used proprietorially.
The reason is because when you make an iteration in the standards to improve speed, compatibility or to take advantage of new hardware developments, you have to gather everyone around again and start on your standards once more. You need to have complete backwards compatibility and a system that is guaranteed to work for everyone.
This limits technical advances. If you develop a new 128 bit CPU system say, then the current 32 or 64 bit system needs to be retained. That means that somewhere you need to be converting the 128 bit CPU transfer to 32 bit or 64 bit. Or 32 bit to 128 bit (or whatever
). That means you're never getting the most out of the current hardware as you're always tied to previous legacy systems.
Basically, all that means that you're always going to be anywhere between 3 and 10 or 20 years technologically behind your competitors.
Let's assume that in five years time, computing power in this system is good enough so that you don't worry about being behind. So it's fine to produce these archaic blocks of interchangeable phone. It could happen, so let's look at the next stumbling block; again, from a technology point of view to start with.
Operating Systems are the biggest issue now. People love to moan about Windows, and how it's huge, clunky, unstable and resource hungry. There's a reason for that - to make it easy for the end user, Microsoft essentially sends Windows out ready to deal with any hardware you may throw at it. It has to be bulky and clunky, because it's being everything to everyone.
Windows crashes a lot because there will be some variations of hardware, that when put together cause driver/software compatibility issues. They may be very subtle and hard to track down (a crash every Wednesday at 11:00 p.m.), or disastrous from the start with the hardware not booting at all.
Linux has overcome this issue by having a base set of vital "core" systems in place, designed to work generically. It means they don't work particularly well optimised (so again, you'll be getting hardware not running at current speeds) - so a Linux user then loads their own specific drivers and software to run their machine on a "I need it now" basis. IT means you don't have the jack-of-all-trades big lump you get in Windows, but it does mean you need users who really know what they're doing. It's too complicated for anyone other than the computing hobbyist or professional. So basically, you can't produce that as a phone.
Android currently sees issues, with each manufacturer having to tweak Android for their own particular flavour of phone. For each different model, they need a slightly different version of Android. You simple couldn't have this with an infinite number of possible hardware configurations, so the "Windows" option would be the only one.
We'll assume then that because everything is so powerful and coding is so wonderful and standards so standard, that we can actually overcome the operating system problems. We then have to deal with the final nail in the coffin - form factor.
Look at your current smart phone. I'll bet it's thin. It's thin because every component from screen to battery to motherboard to speaker and mic are designed to fit in as small a case as possible. If you look at the variety of phones in the world, they come with different screen sizes and thicknesses. For this system to work, you'd need to have a standard form factor. Imagine buying it in bits, and the screen was 12 pegs by 30 pegs, but you have a CPU module 11 pegs wide and 15 down. Your GPU module is 5 wide but 16 high. You can't combine them.
So you'd have to divide up the form factor into module size standards and areas. The GPU is always 4x4, the CPU 6x6, the Wifi 2x2, etc. They would then have to fit into specific points on the year of your modular peg board. Every phone would have to be exactly the same length and width to accommodate this. So the first standard would be "what size is the unit going to be?" How then, do you deal with people who want smaller phones, or larger ones? Your competitors are offering a full range of form factors, but you have one... For 20 years...
Next you look at thickness. You'd need an interface board between screen and backplane (the module area for the components). It would have to be sturdy and be able to cope with both front and back connectors being pushed into it. The key word here is "sturdy". That immediately means that things are going to be chunky. I reckon you'd be looking at 5mm, even if you used an edge connector for the screen. Add on the backplane components and your phone is now a decent sized brick. I haven't even started to cover trying to fit a battery in there (at the moment, they are designed to sprawl as much as possible and are shaped to fit into nooks and crannies in the case design, maximising the power containment of the battery). A round AAA battery wouldn't have the power, so you're really looking at a piggyback battery on the modular components. That could add another 5mm to the thickness. Then you need to start looking at binding the whole thing together to lock solidly so that it doesn't twist or fall apart.
The physical form factor is the final, absolute end for the idea. People want design, they want aesthetics and it's simply impossible for anything in a small form factor to possess this if it's modular. You could round off the screen, or the modules edges, but it would have to be a standard design - rounded edges on the screen and square edges on the interface/modules would look abysmal. Even modul thickness would have to be standardised, or you'd end up with a lumpy phone. So you're stuck with "this is the brick now, and this is the same way the brick will look in 20 years time.
In the mean time, everyone else has produced something sexy, delicious, modern and ultimately, throwaway for a lower price than you can build a modular phone from scratch - as the manufacturer is sourcing identical components for their entire phone manufacturing, they buy in bulk and get it much cheaper. The finances of that are difficult to explain... Erm...
If you're Samsung and producing the S4, you decide you will sell 1 million units. So you buy 1 million of everything, preferably from the same supplier where you can to get discounts. You then make 1 million units and cost them at 1 million x (manufacturing costs + profit).
If Samsung produced modules, they could produce 1 million screens, 1 million GPUs, 1 million CPUs, etc. Not much different there, only now when they sell, they have say 20 modules to sell instead of a single product. Each needs to be packaged, it needs to be "modularised" - both adding to the cost. There's no guarantee that they will sell in bulk in the same numbers. They may sell 1 million CPUs, but only 10 screens. They then have to cover potential shortfall in sales by increasing prices. So eventually, you actually find that a phone made out of modules comes to a much higher cost put together than one mass produced. You already see it in PC manufacturing which is pretty much the same business model.
The only real advantage is you can mini/maxi your own device configuration. So you can build one out of the cheapest bits which don't concern you to keep costs down and upgrade later if you need to. So you could end up with a device slightly better designed for your wallet and need than one off the shelf. However, it would still be a brick, years out of date and at least the cost of an equivalent model out there.
Phew... I may have covered everything there, and why you'll never see one that crosses company boundaries. You
may see one from one manufacturer, which eases standards and the like, but even then, I can't ever see the cost/benefit analysis working out in favour of the brick over a throw away phone.
Personally, I see something very different (and have for years).
The power of the device is placed into a small, wearable (or easy to pocket) core device. The screen/keyboard/mic/speaker are all remote slave devices. You will be able to mix and match each one depending on what you want. So you may buy an Apple iCore phone, but the screen is from HTC. You wear the iCore on a chain around your neck (due to miniaturisation, they'll become that small and because it's not powering a physical screen, the battery life will be great) and it broadcasts back and fourth the screen data. You pull out your touch screen and it works - think the Wii U here. Broke your screen? Pick up a replacement for £20 and pair it to your phone. You lose nothing and you don't worry about the device too much. Want a faster core phone? Buy one and keep your old peripherals.
That's where things will go if they go "modular". It will be a separation of the processing unit from the input/output devices.
I should have done my dissertation on this :lol:
