 |
|||||||||
|
|||||||||
|
|
In a previous blog post, I mentioned that Live Mesh could possibly be the answer to the big problem about state sharing in peer applications. That problem is that in order for an application to propagate its changes to another application, both of those applications need to be online at the same time. Live Mesh solves that problem by allowing an application to synchronize with "the cloud" and then the next time another instance of that same application starts up, it too can synchronize with "the cloud" to obtain the latest changes.
Currently, the beta implementation of Live Mesh includes the ability to share folders, share files within those folders, and you get an RSS-like newsfeed that tells you when people come and go in your mesh, when devices are added and removed from your mesh, and when changes to files take place in your shared mesh folders. The similarity with RSS is not trivial - the entire Live Mesh system is based around bi-directional change notification using APP (Atom Publishing Protocol) and other technologies like FeedSync (and RSS bi-directional change notification spec).
So how does this relate to MobileMe? MobileMe is, according to Phil Schiller's keynote, "Exchange for the rest of us". What this means is that using MobileMe, you will receive push contacts, push e-mail, and push calendar notifications. This will work with any MobileMe-aware application, including Outlook on the PC and iCal, Mail, and Address Book on the Mac and iPhone. This also includes the old iDisk functionality which allows you to share files among all of your devices using file synchronization technologies. iDisk works, but don't ever try to code directly on an iDisk folder with Xcode unless you have a back-up. Hopefully this peculiarity has been fixed in MobileMe.
So, what's funny here is that MobileMe will hit the market on PCs and Macs before Microsoft's Live Mesh is even out of beta. Current implementations of Live Mesh don't even include Mac clients, though they claim to have Mac support in the future. Fast forward into the future when both MobileMe and Live Mesh are released. What will be the deciding factor between whether and end user will want MobileMe or whether they will want Live Mesh?
Bottom line is users will use whatever is least intrusive. This means whatever is the the most integrated solution will be the one users prefer. I don't know which one's going to win this war, only time will tell because MobileMe isn't finished announcing features and neither is Live Mesh. However, given Apple's track record, I will put good money on MobileMe being dead simple to use. The deciding factor for me as a developer will be the SDK. If I get a chance to look at the MOE (Mesh Operating Environment) SDK and I get a chance to compare it to the MobileMe SDK, then I'll be able to tell which one I might want to adopt for my own applications.
The ability for my application to have access to "cloud" services like MobileMe or like Live Mesh means that my applications are going to have an unprecedented level of functionality, robustness, and flexibility. Bottom line is regardless of whether MobileMe or Live Mesh is the victor, my applications will be better for having these services available to me. Who knows, I might be able to write an app that can sync data with either one, depending on the service to which my users are subscribing.
Either way, I can't wait for the next beta of Live Mesh and to get my hands on the MOE SDK. Nor can I wait to get my hands on MobileMe and, if there is one, a MobileMe SDK.
Before you even ask, yes I do know that MobileMe is a repackaging and
re-envisioning of the original .Mac service. I think MobileMe has more
cohesion and a clearer vision, and certainly better support for the iPhone.
Time will tell whether MobileMe will take off and whether Live Mesh will,
and if they will be competitors... lots of questions still remain to be
answered, many of which will depend on the quality of the SDKs.
Your not saying you were working DIRECTLY to .Mac? AFAIK, working directly
to a server (especially with the latency of iDisk) is a recipe for
corruption in may apps.
I was actually saving my files to the iDisk instead of saving them locally
and backing them up to iDisk. Stupidity with iDisk? maybe...but it works
fine under Live Mesh.
Personally, I hope we're moving out of the 'all or nothing' era, where a
single solution (Windows/IE/Exchange) dominated, back to the more
competitive situation that existed in the 80s. So, good luck to both.
I think Mesh is a lot more than MobileMe. I may be wrong, but it sounds
like Mesh will enable collaborative editing shared data. This has been
demonstrated on Ch 9 already with the Family Tree demo. At the moment I
don't see MobileMe allowing custom schemas for third party apps. Given
that Windows 7 and Office v. next are likely to leverage this I think this
is where Mesh has an advantage.
If it doesn't violate any NDAs that you might be under I was wondering if
you have seen and could comment on this take of SproutCore?
Once SproutCore.com comes back up, I'll do some poking and maybe blog
something up.
