The "open hardware" phone project that's had the most interest

[Raphael Wimmer]

On Sat, 05 Oct 2013 17:04:24 +0200, Pascal Gosselin  
<pascal at aeroteknic.com> wrote:

[...]
> If technically feasible, this project I believe stands the best chance  
> of obtaining funding as the concept has wide appeal.

Interesting long-term vision: maybe.
Short-term replacement for GTA04: no.

I posted this on the OpenPhoenux list recently:

There are numerous threads on Reddit that explain very well why this is  
not feasible [1,2,many]
The (imho) most important ones in short:
* Mechanical modularity increases size and cost, makes it harder to create  
a beautiful phone, and is not in the manufacturer's commercial interests.  
Therefore, it would be hard to find companies/customers to build/buy this.
* Many current components are highly integrated - SoCs, sensor ICs,  
Display/Touchscreen, etc. Making these modular would require development  
of many new components (and would increase size, cost, power consumption).
* The concept assumes that all components use a common communication  
backplane. This is not feasible, as a variety of voltages and  
communication protocols are in use in a typical phone (I2C, SPI, UART,  
USB, various display protocols, etc.). Many components need very short  
connections to the CPU/GPU/whatever without crossing other PCB traces. It  
is not realistic to make this work with a generic communication backplane.  
Proper heat dissipation for CPU/GPU is another problem.

In summary, while it is certainly feasible to build a modular phone (look  
at David Mellis' DIY cellphone [3]), doing so for current hardware would  
involve major engineering effort (== design and manufacture dozens of new  
chips) and would result in less stable, more expensive, and less beautiful  
phones requiring more power.

Raphael

[1]  
http://www.reddit.com/r/gadgets/comments/1m6y1q/that_phonebloks_things_annoyed_me_so_here_are_17/
[2]  
http://www.reddit.com/r/explainlikeimfive/comments/1m4pmy/eli5_why_is_phonebloks_a_bad_idea/
[many] http://www.reddit.com/r/all/search?q=phonebloks&restrict_sr=on
[3] http://hlt.media.mit.edu/?p=2182



-- 

Dipl.-Medieninf. Raphael Wimmer
Wiss. Mitarbeiter / Research Assistant
Doktorand / PhD student

Ludwig-Maximilians-Universität München   E-Mail: raphael.wimmer at ifi.lmu.de
LFE Medieninformatik                     Skype:  real_raphman
Amalienstr. 17 / Raum 206                WWW:     
http://www.medien.ifi.lmu.de
80333 München                            Tel:    +49 (89) 2180-4659
Germany                                  Fax:    +49 (89) 2180-99-4659