Some thoughts on Hyperdesk

Published on Friday, May 2, 2008 By Brad Wardell In OS Customization

When it comes to skinning the OS, I'm a fanatic. I've been into changing the look and feel of the OS for nearly 15 years now.  From Object Desktop on OS/2 to working on the very first GUI skinning program on Windows, skinning is a passion of mine.

Over the years, I've seen a lot of programs, companies, and ideas come and go.  On the software side, there's been great programs like eFX and Chroma (to name two) along with the ubiquitous WindowBlinds.  On the content side, there's been teams of artists who have created some cool stuff like Pixtudio, The Skins Factory, SkinPlant, LIghtStar, and so forth.

When Windows XP came out, all of the third-party skinning engines went by the wayside except for WindowBlinds. WindowBlinds had an advantage in that Windows XP came with a skinning engine called uxtheme that was essentially a derivative of WindowBlinds 2 (i.e. they function the same way, have very similar formats, etc.). So rather than being hurt by XP, WindowBlinds got a free boost.

So back in 2001, the skinning world kind of branched into two groups. You had those who used WindowBlinds to skin Windows XP and you had those who used uxtheme to skin Windows XP.  Each solution had its own pros and cons. 

WindowBlinds costs money ($19.95) but had a much wider variety of skins and was a superset of features of uxtheme. 

On the other hand, uxtheme was free but to use it, users had to apply a patch to crack the theme protection. Microsoft didn't want people making skins for uxtheme so they added digital signing protection to their skins. To get around that, hackers patched out that check and allowed the creation of third-party skins for it called msstyles.

So the division in groups largely rested on whether the 20 bucks for WindowBlinds was a sticking points. The people who went with WindowBlinds would say that they have a lot more skins, could use any msstyles skin converted to WindowBlinds and had a lot more features. Users of uxtheme rationalized their choice by pointing out it was free and it wasn't hard to patch the file.

And so for 7 years the two skinning communities developed on a parallel course. WindowBlinds, being a commercial product, continued to be developed over that time. uxtheme, of course, remained unchanged.  Some companies began making skins for money. The first were Pixtudio, The Skins Factory, and Skinplant. They all made skins for WindowBlinds.  But others made pro skins for uxtheme such as LightStar. 

In 2006, Stardock, the makers of WindowBlinds had a new idea for skinning called MyColors.  The idea was that Stardock would create a new group called Stardock Design and go out and partner up with major brands starting with sports teams and universities to sell branded themes. These themes would have the necessary software embedded in them. To make this happen, Stardock Design would also need artists.

The existing studios such as Pixtudio, The Skins Factory, and Skinplant simply couldn't produce enough content per year.  Instead, Stardock Design went out and brought in house the skinners who made up these studios. All of Pixtudio and Skinplant work at Stardock. 

This is where things get interesting...

The owner of The Skins Factory saw Stardock Design as a competitor and relations between the two soured. Moreover, The Skins Factory made it clear that it was going to work to create its own skinning solution rather than rely on Stardock's solution so that it wasn't dependent on what it perceived as a competitor.

Since The Skins Factory was (and still is) a one or two man shop that contracts freelance artists, the steady work at Stardock Design in creating content for major brands led some of the free lancers to join Stardock full time which further soured things with The Skins Factory.

Thus "HyperDesk" was born. Hyperdesk was discussed on WinCustomize considerably so I won't go into all the technical details here. Since it isn't released at the time of this writing, I have to go by informed speculation. It is essentially a uxtheme patch combined with a theme manager that will apply icons, wallpaper, along with specific support for applying skins to specific applications (like Winamp or Media Player).  The problems with patching system files are well known, particularly the long term viability of patching uxtheme.dll from a consumer point of view.

Now for the commentary:

Hyperdesk and MyColors are similar in that the idea is to let people buy a "theme". You want your desktop to look have a complete look? Don't want to mess with a bunch of programs to do it? Then just buy the theme and with the press of a button Hyperdesk or MyColors will apply it.

Hyperdesk seems to go with the longstanding Skins Factory tradition of quality over quantity. Ironically, because they're stuck with uxtheme, they're very limited on the GUI skinning (it's no coincidence that their teasers are limited to shots of the media player as the few hints of the actual skins are pretty typical 8+ year old msstyles based tech which is pretty bland now). They will be hard pressed to approach the quality in their existing portfolio. That's because their existing portfolio was based on the technology that serves as the precursor of MyColors.

MyColors, by contrast, is powered by something that can offer better quality but the need for quantity to get serious distribution attention has spread Stardock Design thin.  This means that there just isn't the marketing bandwidth to spend the time to create an incredible web presentation.

The presentation of say the Hyperdesk Sony Ericson skin absolutely blows away the presentation for MyColors themes.

For example:

MyColors Mustang vs. Hyperdesk Ericson

But you'll notice that they aren't showing very much. They show one icon (which ironically was made by someone who works at Stardock) and just bits of GUI. But the presentation if first rate.  The MyColors Mustang page, by contrast, looks like an page or something.

Because Hyerdesk hasn't been released yet, I can't comment on the actual quality of the Ericson suite.  I can point out some facts that The Skins Factory has made public.  First, it won't support Vista which, in 2008, is pretty catastrophic.  The number of people so into skinning that they'd pay money for premium themes but have stuck with XP is not very large. Secondly, unless Hyperdesk gets a miracle soon, there won't be any significant distribution channels. The uxtheme patching means he won't be getting it preloaded (not to mention the lack of Vista support) and he has no native channels to get started in which has relegated to him to having to post "teasers" in forums.  In fact, this article is probably the most significant publicizing Hyperdesk has received so far.

The thing about selling these all-in-one themes is that it's based on conversation rates. That is, N% (where N is typically less than 2%) of people exposed to it in a significant way will actually buy something.  So if you get 1,000 people to download and try Hyperdesk (or MyColors) you might get 10 to 20 of them to pay for it. 

A decent theme made by non-slave labor costs tens of thousands of dollars to make.  Let's say $10,000 to break even on a minimal theme. If you sell your theme for say $12.95 you're probably netting around $10. This is assuming it's an inspirational theme (i.e. has no royalties attached to it).  To break even, you have to sell 1,000 of them. That means you have to have a channel that can get 100,000 people exposed to it. That's a lot of people - just to break even. And I'm being cheap on the cost.  It costs Stardock Design more than that to produce a suite once you count associated costs.

Worse for Hyperdesk, even if they somehow get a uxtheme patch solution for Vista, the msstyles format on Vista is completely different and doesn't have a nice editor like XP msstyles did.  At the very least, it would require creating a whole new msstyles for Vista which would increase the cost.  So now you're probably closer to $20,000 to produce a single theme -- and this is if it's purely original work. If you were doing, say a Disney based theme, you would have royalties and upfront payments and approvals involved which drive up the costs even further.

The Business Model

Now everything I mentioned here were things we considered when doing MyColors.  The only way MyColors succeeds is if it gets massive distribution. That means getting preloaded or doing special distribution deals with major brands. That's why MyColors uses WindowBlinds OEM technology (its skin format is a bit more restrictive to ensure maximum compatibility). It's also why it includes gadgets. Gadgets can be branded and remote control any media player instead of having to make a skin for a specific media player. 

Over the past year, Stardock has signed on several major distributors and PC makers to begin phasing in MyColors distribution. So by end of this year, MyColors will be on millions of computers.  But this was only possible because a) MyColors doesn't tamper with system files and MyColors has a library of hundreds of themes. Those were the two pre-requisites because most distributors aren't that interested in distributing something that brands them as much as finding ways to generate measurable increases in revenue using their massive distribution. 

And that combination is what I think will be the death blow to Hyperdesk.  You can argue I'm biased or whatever but you can look at the facts for yourself:

  • Companies with big distribution channels aren't that interested in increasing their "brand awareness" with a skin because they already have massive brand awareness because they have big distribution channels.
  • The above companies generate additional income by selling things through their massive distribution channels
  • These companies will only include things likely to make a lot of money which in this case means a large library of content
  • These companies will not tolerate support issues from what they bundle. Hence, something that patches system files is DOA to them. People will flame WindowBlinds but the reality is, it has a long successful history of enterprise-level robustness when the content for it is provided by professionals.
  • These companies will want to support the current version of Windows (obviously).

Without the above criteria satisfied, Hyperdesk can't get massive distribution. And without that, he's stuck selling to the hobbyist community - except Hyperdesk doesn't have a community to sell into. Posting on a personal page on deviantART isn't going to cut it and deviantART isn't going to get into the business of trying to sell a handful of premium themes on their site any time soon.

Now, does that mean you, the reader, shouldn't buy a Hyperdesk theme? Oh no. If it's good stuff, you should buy it. I will probably seriously consider buying Hyperdesk themes if I like the themes. I can, after all, always convert the msstyles to WindowBlinds to avoid patching anything and then I get to run it on Windows Vista.