Ultimate user profile manager will sync Windows, Mac, iOS, Android, SaaS, etc. Someone build it?

It's been a year-and-a-half since I wrote We need a new "user settings" framework before rich internet apps will replace Windows apps. I'd like to revisit the issue because it's never been solved.

It's been a year-and-a-half since I wrote We need a new “user settings” framework before rich internet apps will replace Windows apps. I'd like to revisit the issue because it's never been solved. So let's take this from an abstract concept to something that someone (not me) can go out and build.

What is this "user settings" framework / ultimate profile manager thing?

What I have in mind is something that's similar to the user settings portion of a Windows Profile, except this new idea would work for all applications on all platforms. In Windows, your profile maintains personal settings like what language you use, your custom spelling dictionary, your time and date formats, your avatar, your wallpaper, and a few hundred other little settings that make up your user environment. And if you use roaming profiles or a third-party user workspace tool (AppSense/Scense/triCerat/RES/Citrix/RTO/Liquidware) you can easily take these settings from device to device. (Well, from Windows device to Windows device.)

Now I want a way to extend that to other platforms that I use, like Mac, Linux, iOS, Android, etc. And I also want to extend it to the web applications / SaaS apps that I use.


As I wrote in 2009, I'm sick of having to right-click on the same f*ing word with wavy red underlines every few days because I'm typing it from a new device or from a new app. And I'm sick of having to switch my time zone settings on my laptop AND my iPad AND my SaaS apps AND my BrianMadden.com account when my phone (which I always have with me) knows exactly what timezone I'm in. (Why can't it just tell everything else where I am?)

This could also apply to just about every setting on every platform. (And if we do it right, we'll support the desktop's move to irrelevance that I wrote about last month.)

From an enterprise standpoint, we know that lots of companies are embracing web and SaaS apps. We can see things like VMware Horizon and Citrix OpenCloud Access and Centrix helping tie the provisioning and authentication of web apps back to Windows apps. But why stop there? Why not all settings?


Hey, if Microsoft can create a Windows XP-to-Windows 7 profile migration wizard (the "user state migration toolkit,") or if Apple can migrate Windows settings to Mac OS X settings, then why can't this just exist for all platforms and operate in a continuous fashion?

I'm thinking the easiest way to do this would be to create some kind of neutral (semantic?) standard for storing the various personalization elements as a web service somewhere, and then each platform or SaaS app could run some kind of agent that translated the centrally-stored settings into the native settings for each platform. Citrix has a Citrix Receiver for every platform, so I guess this would be a "personality receiver" or something.

At the most basic level, there could be a URL that was my personal storage location that was shared amongst all my apps, devices, and platforms. So I could point everything I own towards brianmadden.com/me.xml and rest assured that my changes were everywhere. My phone could update that location whenever I moved to a different timezone, and my other apps and platforms could query it every time they started or something. And of course not every device would leverage every setting. (My choices from touch would only be used for touch-based devies, etc.)

Who should build this?

My first guess is that this is something for Facebook or Google to build. Second is maybe Microsoft.

Of course any of the established Windows-based user settings management companies I mentioned earlier could add this feature in to their products. (In fact I can think of one company with about 70 million reasons to try something big and new like this!) Really the hardest part would be writing the plug-ins for each client platform that would map back to the central storage location.

Or is this something for a web standards body to create? (Or maybe The Coalition?)

The real question is how you'd separate the "standard" (or the "concept") from the actual ownership of the data itself. Do I want Microsoft Windows 8 to be able to access user profile data from anywhere via a specific XML format? Yes. Do I want Microsoft to own and host my data? No.

This is about user settings. It's not about data

Now before Tim Mangan freaks out on me for saying it's not about the data, I want to point out that yes, the data is important. But for this specific idea, I'm looking more for a standard way to interchange settings. The data is a different conversation--although what's actually cool is that the ability to share "data" between all these different platforms and devices has already more-or-less been solved by vendors like DropBox, Box.net, Windows Live Mesh with SkyDrive, SharePoint Workspace, etc.

In fact, ironically, maybe for this user settings thing we don't even need a central location or web service at all? Maybe we just need some standard way to write all the settings out to the file system so that our DropBox (or whatever) can handle the synchronization and replication. (Of course if we learned anything from Windows over the past fifteen years it's that we don't want a single huge file, we'd actually probably want a folder full of lots of little files--each containing a single configuration value.)

Tough questions we'd need to answer

I'm sure there are about a million questions we'd need to figure out before something like this could really take off (and I'm hoping you'll share your thoughts in the comments), but here are the things that come into my mind as I write this:

Who owns the data? If your company provides a mechanism to store your roam-able user settings, what happens to your settings when you leave? Can you take them with you? Does the company retain a copy?

Should this thing support a layered approach? Maybe there can be multiple "layers" of personalization settings, perhaps from different providers. (So you have a personal set of settings tied to your Facebook account, and your company has another set they maintain.) Then when using company devices, you'd merge the two sets together, perhaps with policy controls. This might answer the question about what happens when you leave, but it also really increases the complexity.

Do we need to build this from scratch? Look at things like OpenID. Wouldn't it be great if they could sort of just

Is this for us to solve? I mean "us" in the context of us in the corporate IT world. But I wonder if something like this is already in the works from the more consumer Web 2.0/3.0 companies? (Maybe I should have attended SXSW this week). Then again, we corporate-focused people will do a better job of building something that will integrate with companies, security policies, compliance, etc. (Although we corporate folks aren't always the best innovators. :)

Do we allow extensibility? How tight should the standard be? Do we let vendors and individual users extend this framework with their own name/value settings pairs? At first you might think this makes sense, so new apps/devices/capabilities can be added. But of course if we do that vendors (probably Microsoft) will just f*** it up with all their own customized crap. (Just do open a .docx file in a text editor to see how far from XML Microsoft can take it. Yeah it's technically XML compliant, but it's not quite the easy-to-use interchangeable open format we were hoping for.)

By the way, if anyone wants to go out and start this company, I'll trade you the LufLogix.com domain name for a 1% stake.

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

I think you hit the nail on the head with why people are getting so excited about user virtualization these days. Window's profile management, while certainly a critical element of rolling out VDI and terminal services at scale, is not the "big idea." This is!

I'll chime in on some of the questions you posed:

Who owns the data? This is a great opportunity for a trusted intermediary to step in to be to user identify and data what VeriSign (or now Symantec, I guess) is to online transactions. For this to take-off, there needs to be two-way protection and trust. Even if my company told me my personal settings are portable when I leave, I am not sure I would trust it. Also, I think companies would value a line of demarcation between them and my personal data. They don't want to be on the hook if I have something illegal or embarassing sitting in my personal data. (It's always the quiet ones you need to look out for...).

I also think this intermediary needs to be a company that specializes in this and has a tranparent business agenda. I would trust Dropbox or say....AppSense. I would not be comfortable with Facebook or Google running it, even though I like and use both. Advertisers are their customers - not my company or me.

Should this thing support a layered approach? Yes, I think it does need to be a layered approach with seperation of data and policies to work for everyone for the reasons I described above.

Do we need to build it from scratch? Yes and no. I think it needs to be architected from scratch, but there may elements of OpenID, OATH, etc. that may fit into the puzzle. OpenID is a great architectural concept with the worst possible front-end implementations. The "pure" implementations are too esoteric for regular users to understand and trust. And then more mainstream implementations like Google don't really achieve the true intent and potential of OpenID.

Is this for us to solve? I really think it is. Wearing my AppSense company bias on my sleeve a bit, it all starts with Windows today. Starting inside in the enterprise with technology that IT folks know and growing outward into the cloud and personal data space is going to be much more effective than driving Web 2.0 stuff in. Some forward-looking IT folks would get onboard I am sure, but I think enough would throw the sandbags in front the door with something that originates outside the enterprise that the whole thing would be stalled. Great opportunity for us user virtualization vendors that we would be crazy not to be thinking about. A lot.

Do we allow extensibility? I think we need to. You are correct that some vendors will "crap it up," but there could just as easily be other vendors that are able to innovate faster than the standard and perhaps even contribute some of this innovation back to the standard to accelerate things.

A question of my own: Wouldn't this be a great judo move for one of the mobile / tablet device vendors trying to create any point of differentiation from Apple? "Yeah, we aren't quite as cool and cost a bit more, but we a framework for governing enterprise and personal computing on a single set of devices." RIM, HP, Samsung, Motorola....call me.

Doug Lane



@Brian - I think AppSense beat you to the idea.



@Icelus Oh cool! Yeah this is it! (Probably he read my original 2009 post. :)


Let me muddy the waters a little here...

Especially as we move to cloud based applications, we have a severe scale issue associated with "my stuff".  Not a capacity issue, this is an architectural one that becomes evident as we start to use more and more services.

Essentially, today, about the only linkage between multiple services I use happen through My PC.  the industry will need to develop frameworks to go cloud to cloud - which means all those pesty issues like privacy and such need to be addressed.  

A simple example, If I'm on Facebook, and want to see the LinkedIn info for an email address or phone number for someone I'm connected to in LinkedIn that posted on Facebook, that's a manual step procesing through My PC.  As we add more and more of these services we will want them working together with appropriate communication flowing directly between them to better serve us.

I'm not sure even Microsoft is big enough for that job.


This would be a very interesting space, but would require the right kind of leadership that get's the current world and the new world. I don't believe the Appsense's of the world have that DNA today.



I have discussed this exactly same idea in one of my articles back in July, 2010. In this article I describe my idea of a VDI cloud stack where diferent entities would be responsible for management the stack. It definetely can be done and would be way too cool!

Read on at The future of Desktop Virtualization (VDI)?


Let me know your thoughts.