However, maybe I'm a bit of a Luddite as I really don't like all this
'cloudy' thing
I believe people around here is afraid of the cloud because they have
seen mysqueezebox.com and know what happens when it goes down.
There are two scenarios:
1. You want to listen to local music
2. You want to listen to online streaming services
People understand that scenario 2 isn't possible without a reliable
internet connection but they want to be able to accomplish scenario 1
without a reliable internet connection. If they loose the internet
completely, most will accept that you don't get full functionality in
scenario 1, but it's important that you at least are able to listen to
your local music. As an example, if they can't share the local music
they are listening to instantly on Facebook while the internet is down
they will generally accept it.
Logitech Squeezebox solved scenario 1 pretty nicely through LMS and so
does Sonos (who only relies on a network mounted drive with local
music).
With UE Smart Radio Logitech showed how you can make a system extremely
dependent on internet both for scenario 1 and 2. It's obviously a bad
idea to do this, but they had to do something quick and they didn't have
the time to do the necessary adjustments to their architecture, so the
UE Smart Radio architecture is what we got.
Now, people who have sniffed the network or looked in the server log
files from LMS with network debugging enabled knows how much traffic
there is continuously between the player and server when using the
system. What Logitech did with mysqueezebox.com is essentially to move
the server part to the cloud which means that all this network traffic
have to be handled by a central server with the result that you are
extremely dependent that the central server must be available for
scenario 2, so you aren't just dependent on an internet connection you
are also extremely dependent also on the cloud server itself. This is
what you get with a slim device when you move the server part to the
cloud, basically a system which is extremely dependent on a cloud server
being up.
As long as you can keep the cloud server up, it's not a problem, but
let's just say that it's a lot easier to operate a cloud server which
responds to 2 messages per minute from each device compared to one that
have to respond to 200 messages per minute from each device. I'm not
sure about the exact numbers here but generally a solution with slim
devices requires a lot more network traffic than a solution with
thick/rich devices.
I'm not going to go into any details how we have solved it but let's
just say that we have learned a bit from Logitech's mistakes and we
don't plan to repeat them, the market looks different today than it did
10 years ago when the Squeezebox architecture was designed, due to this
our system is not a clone of the current Squeezebox architecture.
Regarding having cloud server at all, it's needed for two reasons:
1.
People wants a modern system which easily can be updated with new
features, as bluegaspode mentions this is very hard to accomplish
through software installed on the local network or by updating the
firmware of the devices. Firmware updates is fairly easy to handle as
long as we are talking about a single hardware vendor (look at Sonos or
Apple) but it's a lot bigger challenge when involving multiple hardware
vendors (look at Android as an example). Having the part of the logic
that changes often in the cloud helps a lot in the multiple hardware
vendor scenario. Making our own hardware would also have been an easy
solution to this problem, but we believe it's very hard for a single
hardware vendor to fulfill all possible usage scenarios so we decided to
go for the multi vendor solution instead and focus on making a software
platform.
2.
People want easy access to the latest online services, this includes
everything from the latest social network services to streaming
services. Here open source is an issue, because if you call a premium
streaming provider (for example Rhapsody) and ask if you can get access
to their API to be used in your open source software they are just going
to hang up. If your software is closed source, the first question they
are going to ask is how many customers you have, if your answer is that
you think you can distribute your closed software to 5000 people who are
willing to install it on a local computer, the next question they are
going to ask is if you have some stats that shows that these 5000 people
actually use your software. Try to answer those questions without some
kind of cloud server, then realize that the minimum volumes you are
going to have to show them to make them even interested to talk to you
have to be significant higher than 5000 users, which also means that you
have to be able reach people who aren't happy to install custom software
on a computer.
Independent if we like it or not, a cloud server solves many of these
problems because it makes it easy to upgrade the software with new
features and it makes it easy to get confirmed stats about number of
users using the system which is needed in discussions with some
streaming service providers. It's also important to notice that we have
people in the team who have extensive experience of negotiations with
streaming service providers, so these are not just guesses, it's pure
facts based on previous experience.
If we thought it would be possible to fulfill the user experience,
satisfy hardware vendors and get necessary agreements with streaming
providers with a solution purely based on server software installed on a
computer, we would have done it that way, but this isn't how the market
works.
------------------------------------------------------------------------
erland's Profile: http://forums.slimdevices.com/member.php?userid=3124
View this thread: http://forums.slimdevices.com/showthread.php?t=98467