Re: build a hack box of choice?

On the EOL problem [down below from !3runo] ... if you are paying the 
producer for mass quantity, they will be willing to not EOL so fast.  
It's the SOHO vendors that want the churn.  Newer/cheaper quicker.  I 
don't think it's their suppliers especially.  The quantity producer 
likes to have long runs I think, more profit.  Where I work we do custom 
ASICs for things like medical and automotive, one of our selling points 
to these "low tech" customers is we can keep the ability to supply them 
with more for 10 years.  We probably don't care about more than 3 years 
anyway though.

Along with Moore's law...  The good part is these SOHO type boxes are 
now at a point where they can run Linux and tools well without stripping 
things down too bad.  So we don't have to chase Moore's law really.

Right now, there is probably the right reference design out there for 
us.  The down-side...  As soon as NetGear/Linksys/D-link see you making 
money, they will whip up a similar offering.  But I guess it will take 
this to get them thinking.  Just like no enterprise database vendor 
would port to Linux in about 1997, one finally did... and the rest 
scrambled.

-Scott

John T wrote:

>Seems if someone really wanted to accomplish this, the idea would be choose
>a reference design, design the firmware, contract some company in tiawan to
>mass produce the units, with the specific firmware.... in volume, cost would
>probably be $20 per unit or less... then sell them as specifically openAP
>routers complete with source code, make them geek friendly.  I'm sure there
>would be a pretty good market for them... enough so that you could probably
>find a source of venture capital to get it off the ground.
>
>----- Original Message ----- 
>From: "Bruno Lopes F. Cabral" <[email protected]>
>To: "Linux Access Point Development" <[email protected]>
>Sent: Thursday, January 29, 2004 4:58 AM
>Subject: Re: [LinuxAP-dev] build a hack box of choice?
>
>
>  
>
>>Price?
>>
>>about the EOLd, the problem is Moore's law. one year-1 1/2
>>after the release of a hardware revision, components start
>>to be difficult to find, newer powerfull versions come to
>>live and so on
>>
>>so consider 18 months as the maximum lifetime for new
>>soho-routers. you will still be able to find them up to 6
>>months after that, but they will be mostly factory outlets
>>
>>Cheers
>>!3runo
>>
>>Francois Gurin wrote:
>>    
>>
>>>How would such hardware differ from soekris and openbrick?
>>>
>>>--francois
>>>
>>>On Wed, Jan 28, 2004 at 10:03:36AM -0700, Scott Serr wrote:
>>>      
>>>
>>>>To clarify what I am thinking a little...  What if XYZ company could
>>>>supply you with hardware that is almost like the SOHO router that you
>>>>are porting to, but XYZ company adds widget A to make it hackable.  And
>>>>they try to have a long run without EOLing so fast.
>>>>
>>>>Basically XYZ would obtain or build the super hackable SOHO you want,
>>>>give it to you at a decent price and in a limited way support the
>>>>hacking and Linux porting that you guys do.
>>>>
>>>>-Scott
>>>>        
>>>>
>>_______________________________________________
>>LinuxAP-dev mailing list
>>[email protected]
>>http://linuxAP.ksmith.com/mailman/listinfo/linuxap-dev
>>
>>    
>>
>
>
>_______________________________________________
>LinuxAP-dev mailing list
>[email protected]
>http://linuxAP.ksmith.com/mailman/listinfo/linuxap-dev
>  
>