Friday, November 13, 2009
Mobile cloud does not mean network computing
I found an interesting comment the other day in an article:
Nokia has batted back the concept by stating that mobile devices will not become ‘dumb clients' relying on the Cloud for the majority of intelligence.I could not agree more. I do not see a future where mobile devices are dumb, and source everything from the network. I see a future where mobile devices have local data, where the data is updated when the device is idle (from the cloud), where the applications are installed on the device. A network where the data is dispersed on every device and aggregated in the cloud.The company's chief development officer, Mary McDowell, has stepped forward by accepting that the Cloud will grow, "we don't think the cloud is the total answer. Mobile devices are becoming more personalised and increasingly part of an individual's life. We think it will not be either/or," she said. "There will be a lot of intelligence in the Cloud and in the device, and the ability to exchange data with the Cloud will not pave the way for thinner devices, but increasingly powerful ones."
Would it make more sense to just have all the data in the cloud, and have dumb terminals access it?
Yes, if you do not consider the following:
- COST. That for me nails the argument. The cost of network is going down ever so slowly, if it ever does. And the network are ever more saturated (think about it, they said we had enough capacity with 3G, now even 4G sounds limiting). On the other side, the cost of storage is going down dramatically. I can store on my cell phone what I could store on my PC a few years ago, for a fraction of the cost. Who can point to one single projection where network bandwidth will be sufficient and cheap, while storage on devices will cost a ton? Exactly...
- USABILITY. It starts with bandwidth, but it does not end there. Data on the device and apps on the device mean immediate access to what you are looking for. I wrote about it many times in the past: the scenario to keep in mind is the user with an open umbrella in pouring rain.
- OFF-LINE usage. It is a scenario that might become less frequent with wireless reception getting on planes, trains, tunnels, rural areas and so on. But it is not going to disappear completely. And, believe me, in that particular situation you will need that information on your device badly...
- OWNERSHIP of data. This is a silly need, because you still "own" the data even if it is not on your device, but it is stored in the cloud. However, the perception of it being far away, even if it is yours, will come in play. One article of a cloud player who lost all of the users data, and people will appreciate having their own data on their own device as well. It is mine, I want to keep it with me. All the time.
Posted by Fabrizio Capobianco at 17:56
3 Comments:
CEO said...
Agreed. Related to ownership vs. where the data resides is the level of *legal* protection or laws applicable to "your data". Once is other people's servers, law is weak (in the USA). That might change in the future.
Comment Posted at 16:56
jackr said...
I'm with you on everything except your "ownership" paragraph. Your reasoning applies to ownership only in the sense of "easy, reliable access by me to what I own." This certainly includes protection against the risk of a cloud data holder going off-line, which you mention; it also includes protection against the risk of a cloud data holder blocking your access to information, or preventing you from knowing they hold the information.
But there is another sense of "ownership," which is at least as important: "protection against the data I own being used in some way I do not approve." This includes marketing of contact info and aggregation of use patterns. And it requires protection against some very heavy-weight risks, such as cloud data holders using the power of of bankruptcy courts to break even explicitly agreed contracts relating to privacy and ownership, a thing which has already happened.
In personal data ownership, as in copyright, digital data presents problems fundamentally different from classic physical-object models: data's nature is to be copied cheaply and widely; ownership notions must take this into account.
Comment Posted at 11:08
Antoine said...
I've been staring at this piece for over an hour, knowing that I've wanted to comment on it, but not having the words. I walked away to post a few things and read a few others and came back and saw this:
"...I see a future where mobile devices have local data, where the data is updated when the device is idle (from the cloud), where the applications are installed on the device. A network where the data is dispersed on every device and aggregated in the cloud..."
Exactly!
This is how mobile's can and should get by without relying on the cloud, but in effect seeding it.
If done in this manner, social networks, smart devices, and even traditional communications start to take on a new life.
Now, this is possible when the device is a server of some sort. You've got initiatives such as Nokia's Mobile Web Server, and then there's the Open Mobile Web Server Standard (mobile web server on a SIM card) that's under some development. Both of these would enable this functionality from the side of the mobile device. The harder part is taking the traditional ideas behind cloud-apps and websites, and changing that to being an aggregration of authenticated users (think authenticaing an application to use/access Flickr or Twitter).
It would be up to the broker (carrier, mobile device maker, service provider) to make sure that you can connect, but the information and what information is connected is up to the user.
This is what Nokia's 2015 video was trying to get folks to see - and dang nab it, I've got the MWS on my device right now but am not using it in this way because I'm missing the service end of things.
You said it though. I commend you, great comment and spark there.
Post a Comment