Open Source's New Frontiers (Business Week) [LWN.net]

Open Source's New Frontiers (Business Week)

Posted Feb 8, 2006 16:13 UTC (Wed) by sepreece (subscriber, #19270)
In reply to: Open Source's New Frontiers (Business Week) by njhurst
Parent article: Open Source's New Frontiers (Business Week)

Most cell phones are sold to people who use them as phones. Few people, even people with smartphones, add software to them. For the vast majority of the market, adding ringtones is as far as they go or want to go.

Very few people add or modify the software in their MP3 players, wireless routers, DVRs, etc. Most consumers just use them as the consumer devices they were built to be.

Consumer device makers are generally aimed at that vast majority. Taking on the added cost of creating SDKs and managing customer support for the relatively few who want to change the software is not an attractive business plan. As to liability, consumer device makers do regularly get sued because their devices don't behave as expected. If users can modify the software, there is the potential for additional liability for misuse of the device (if, for instance, someone turns her cellphone into a cellphone-jamming device).

(Log in to post comments)

Open Source's New Frontiers (Business Week)

Posted Feb 8, 2006 17:04 UTC (Wed) by mikec (guest, #30884) [Link]

This is indeed the very boring future engineers software and hardware alike face...

Everything will tend to the "appliance" model as that is what the "typical" consumer is capable of using...

The use model will evolve (or devolve) to the point that the choice of OS and processor will be uninteresting to say the least...

I am growing at once excited to have this stuff "just work", and frightened that we are going to return to the point that "consumers" don't need and cannot get compute power since GHz processors make no sense for the vast majority of consumers and consumer applications.

It certainly makes no sense to have a TV that is even capable of reciving a virus, much less running a dual core processor to do what can be done in a $0.20 codec chip (in bulk).

The remaining hold-outs justifying the provision of commoditiy/consumer compute power are:
1. games, but dedicated games consoles can and perhaps should make some interesting tradeoffs here that limit their usefulness for "generic compute horsepower"
2. video editing, but see above - cheap codec chips remove the need for CPU horsepower.

So, our Faustian bargain could be rock-solid, unhackable, cheap, cool appliances that wipe our figurative backsides and just work without hassle, but a return to the old days of Big-Iron and dumb terminals...

I for one, am making use of my basement cluster and bemoan the loss on many fronts...