[qcc] TODO?

[Rob Landley]

On 10/21/2015 06:22 AM, Рысь wrote:
> You're talking about android like it's The Only Future and we can't
> escape.

No, I'm talking about smartphones like they're the only future we can't
escape unless you want to follow the old stuff turning into "big iron"
and getting kicked up into the server space yet again.

I'm talking about android like it's shipping over a billion units per
year, which it is:

http://www.cnet.com/news/android-shipments-exceed-1-billion-for-first-time-in-2014/

The entire PC industry's installed base is less than the annual android
unit shipments. Strategically, it's the fourth generation of mainframe
-> minicomputer -> microcomputer -> smartphone. Fighting over PCs in
201x is like fighting over minicomputers in the 1990's. It doesn't
matter who wins, that ship is sinking.

I gave a talk about this in 2013, on the need to make phones into
self-hosting development environments you can rebuild the OS image on,
not just passive data consumers. This is the most thorough "why I am
doing what I'm doing" explanation I've got so far:

Video: http://youtu.be/SGmtP5Lg_t0
Outline: http://landley.net/talks/celf-2013.txt

Getting Android's attention turned out to be possible:

https://lwn.net/Articles/629362/

But it is not the _only_ smartphone OS, I'm also trying to make toybox a
good fit for Tizen for example:

https://lwn.net/Articles/616315/

I keep trying to tell people about this. I was interviewed about it on a
podcast last year:

https://linuxluddites.com/shows/episode-11/

Before all that I did a writeup on the aboriginal linux about page:

http://landley.net/aboriginal/about.html#selfhost

> I wonder why such a hype to a corporation-influenced binary OS.

Because if you try to make your own phone ala openmoko, your unit volume
is so low it prices itself out of the market, and then you can't get a
data plan for it because the providers don't notice (or care) you exist.
I'm not the FSF, I care about trying to improve the world we've got
rather than imagining a cloud castle and insisting everyone live there
with no path from here to there.

The PC didn't start out open, it was _opened_ by cloners like compaq who
survived lawsuits to defend their right to clone it. The open S/100
systems of the day died out due to lack of interest (and yes they went
16 bit with Z8000 and 8086 cards and such: nobody cared). They didn't
take an open system and make it successful, they took a successful
system and levered it open. That's what worked last time. Strategically
they focused on the PC because it had the smallest number of proprietary
bottlenecks (mostly just the BIOS). This time I'm focusing on Android
for similar strategic reasons, making multiple billions of existing
Android devices programmable and rootable is easier than trying to make
firefox phones or ubuntu phones ever matter to anybody.

By the way, my $DAYJOB is trying to fix the long-term "corporation
influenced binary hardware" problem at a much deeper level than all this:

https://lwn.net/Articles/647636/

Rob

 1445476869.0