The next version of Windows really does need to take things to another level.

Let's face it, Windows XP is basically Windows 2000 with some fixes and cosmetic enhancements.

Longhorn needs to be much more than that.  Here are a few things I think it needs:

1) New display system. This is what Avalon is supposed to address. The new display system needs to let us always run our systems as the maximum resolution our monitor supports and have the DPI (dots per inch) be fluidly scaleable without impacting software compatibility.  I shouldn't have to run my laptop at 1024x768 in order to be able to read text if it supports 1600x1200.  I should be able to run at 1600x1200 and size everything on the fly to be bigger. 

2) Updated Searching.  Google Desktop search only exists because the Find Files feature of Windows is essentially useless.  I should be able to quickly find something on my system instantly.  WinFS won't be out as part of Longhorn so what will they be improving in the meantime?

3) Smoother Multitasking.  Windows still sucks at multitasking.  Even when running on an SMP box, if the OS is "busy" doing something, you still can't quickly do something else.  I eventually gave up on SMP since on Windows it's only good for CPU bound tasks and doesn't really affect multitasking efficiency very much (on MacOS X and OS/2, for instance, SMP basically made it so you could always be doing something in the UI, but on Windows, the UI is apparently not as multithreaded as it could be). In the meantime, I use Multiplicity to maximize my computing power. But I'd like the OS itself to let me always be able to do stuff -- even if the app is written badly.

4) More Componentized. Whether we'll get Microsoft to make it so that pieces of Windows can be replaced or inherited from remains to be seen. I would like to be able to easily add more views (no, Ishell stuff doesn't cut it) to foldrers. 

5) Stop bloating with needless bundling.  Every new version of Windows throws in some half-assed immitation of third party software.  While we can all appreciate having a "free" version of ZIP or uxtheme or movie maker, it damages third party software development. I'd rather think that when I BUY my copy of Windows that the work was put into features that only the OS vendor could do.  Especially since Microsoft rarely puts any effort to let third parties expand on what they bundle (like adding RAR support to the compressed folders for example). There are some features only the OS vendor can really do. I'd rather see resources put there.

6) Make Networking better. I don't know about you guys but the LAN support in Windows is still quite a pain.  As I type this, I am on a wireless LAN which has several computers on the same work group.  It often takes several seconds, if at all, to find all the machines on the network.  It would be nice if Microsoft re-thought how people use network resources and included ways of working with them in a more straight forward, ROBUST, centralized way.

7) Better use of memory. I have 2 gigabytes of memory on my main machine. I turn off the swap file.  And yet I still hear the hard drive chipmunks going away.  Why is that? And don't even get me started about the limited number of handles. Even on my 2 gigabyte machine, if programs use more than 24,000 or so handles, programs start crashing. The average person doesn't even know why their system becomes unstable because limited user handles on Windows XP has been largely ignored.

8) Fix Internet Explorer. CSS 2.0 compliance would be a nice start. How about making it much smarter about what it caches? I have lots of friends at Microsoft who admit to having switched to Firefox (or Opera).  That's sad.

9) Fix your third party licenses. One of the ugly secrets of the PC OEM market is that computer manufacturers can't install things on Windows that changes the first boot-up experience. At best, they can put a few things on the desktop.  But they can't, for example, include an alternative shell or have WindowBlinds running by default or change the boot screen or many ohter things.  In short, there's not much way for PC manufacturers to distinguish their computer from every other computer.  That means a LOT of lost innovation.

10) Fix Security.  Outlook Express is still a spyware/spammer's dream. We shouldn't have to "upgrade" to Outlook to have some basic protections. There should be more end user tools that make it very easy to monitor net traffic.  The "Network" tab in task manager is a nice start but it needs to go much furthre than that.  Worms and the like should be stopped at the OS level. SP2 was a nice start, but there's still so much more to do.

There's lost of little things that are being addressed that I'm very excited about.  Avalon is the big thing for me. XAML in particular is interesting but I fear it may lead to a ton of wacky looking "apps". I don't want my apps to be as poorly designed UI as the typical website. Let me put it this way, the people who make Office are VERY different from the people who made http://www.microsoft.com.  I want the former writings the stand alone apps I use, not the latter.

So what would you like to see? What features in Longhorn make you excited?

There are some people who think that the moment someone starts a business that they instantly become greedy, money grubbing corporate robots.  But the reality is, most companies are privately held and those companies do things based on what motivated the stockholders.  In a private company wealth may only be a secondary motivator.

I can say as the principle shareholder in Stardock that accumulating wealth has never been my motivator. I want to do cool stuff. And I consider it my "talent" to be able to generally take things I enjoy doing and be able to do them as part of my "job".  Some things are done strictly because we think they're neat to do even though there's no remote business justification for it (JoeUser.com for instance).  We recently started doing PowerUser.tv too. There's no money to be made in podcasting (unless you're a mega player). But it's fun.

WinCustomize falls into a middle category.  Stardock makes software such as WindowBlinds, IconPackager, DesktopX, and so forth.  People who buy that software or are thinking of buying that software will want to see what can be done with it.  So in that sense, having some sort of gallery of content is necessary.  Statistically, most people hear about the software somewhere else or see a screenshot, download the trial version, and THEN start downloading skins after they've purchased it. It's not the other way around -- as a percent, few people discover our software through the skins first.

WinCustomize is a lot more than a skin gallery though.  It's a whole web community.  It's the largest - by far - site dedicated to downloading content to enhance your Windows desktop experience.  With over 22 million monthly visitors, it's one of the largest sites on the net period.  And it's safe to say that of the 22 million visitors, only a tiny tiny % actually have purchased (or will purchase) a Stardock product.

The growth of WinCustomize has been steady.  In December 2004, the site was getting around 14 million visitors per month. In September 2005, it got just about 22 million.  That's a very significant increase for such a short period of time.  Yet subscriptions to the site have declined.  That is, people who pay $20 to support the continued existence of the site and receive a number of (I think) pretty cool services. 

Some people have argued that WinCustomize has no business even trying to get people to buy subscriptions. Since Stardock "owns" it then Stardock should pay for everything because it's a "marketing" expense.  Running a site that gets 22+ million visitors per month is not like running some home page.  It takes rooms of servers and hundreds of megabits PER SECOND.  It also requires an IT staff, database developers and web developers.  You're talking hundreds of thousands of dollars a year in expenses.

As a marketing expense, WinCustomize's value is pretty limited.  After all, DesktopX has been on WinCustomize since its founding and yet the wider world seems completely unaware of its existence.  You hear a lot about Konfabulator.  But little about DesktopX.  Why is that?  There's lots of reasons. But one of the root reasons is a lack of marketing.  A lack of ability to get the word out on DesktopX to get users and developers to try it out.  More developers doing things with it would mean more and better content which in turn would encourage more users to use it.  But in the end, Konfabulator ended up as the program people think of for adding mini-applets to their desktop.  And those who have heard of DesktopX are often unaware that it does a lot more such as build entire desktops and allow for desktop objects.  So how valuable is WC for marketing? Not much.  To be generous, I'd say $50k.  And that's just one example.

But WinCustomize currently costs over $300k per year to exist.  And in a year, it'll likely cost closer to $400k.  Advertising and subscriptions bring the cost down by around $100k. That leaves $200k left.  It WC worth $200k in marketing? Not on your life.  If Stardock had $200k in extra marketing dollars to spend each year, think of all the advertisements, high end PR firms, and community outreach programs it could have invested in?  

The point being, while WinCustomize has some value as a marketing resource, its value is pretty limited. The simple fact is that WinCustomize exists because of what I mentioned at the start -- Stardock, as a private company, can do things simply because it thinks it would be cool.  It cares about the skinning community and it thinks having a site like this is neat and worthwhile -- to a point.  Individuals have hobbies. Private companies have hobbies too (that's a big difference between private companies and publicly held companies).  But just as with personal hobbies, at some point they get too expensive to keep going.

To try to change things, Stardock has brought in two new people right away to help create more incentives for people to subscribe to WinCustomize -- more content for subscribers.  But odds are, things are going to start getting tighter for those who simply visit the site month after month who haven't contributed anything (on WC, skinners, even if they've not subscribed or bought anything are given access so that they are treated as subscribers).

Some things that come to mind:

• People with no accounts will only be able to download things that have been uploaded in the past 48 hours and only 5 megabytes of content.
• People without account will likely see a lot more ads. See www.gamespot.com for a good model.
• People with accounts will be able to download up to 50 megabytes. After that they must become an SD customer or a subscriber.
• People who have purchased a SD product (Object Desktop, CursorXP, etc.) will have unlimited downloads but there will be some skins that will only be available to subscribers.
• SD customers will see fewer ads but they will see some ads.

These are just a few ideas floating around -- ideas submitted by users incidentally -- that we're seriously considering.  Since WC isn't paid for by ads, the # of monthly visitors is not as much of a goal as one might think.  If we could cut our traffic in half, we could probably save around $50k per year in expenses.  Another $50k would likely result from increase SD product purchases and/or subscriptions.  That would be a net difference of $100k.

There are plenty of other good skin sites out there people can visit. Here are a few: http://www.skinbase.org, http://lotsofskins.com, http://skins.deviantart.com/, http://www.skinz.org, http://www.customize.org.

Stardock - the company - needs sites for people to be able to download skins and themes for its software.  But it doesn't have to be WinCustomize.com. Ultimately, if the wider userbase doesn't or can't support WinCustomize as it exists today, then WinCustomize will evolve to be a site that caters more to those who can support it.  The current model for it is unsustainable.

Screenshot - Download

It all started with another brain-freeze today. I just couldn't find a section that contained the File Dialog places bar images. You know the bar on the left on the File->Save and File->Open dialogs? I bet you've been cursing more than once searching for a section like that. SkinStudio will help you with searching for a section if you know it's UIS name or its part AND if such section actually exist. But in my case it didn't.

Wouldn't it be great if I could just lookup the section by it's name or a part of it?

So I've added a tool button like you can see on the image below.

Press this button (or press Ctrl+F while the tree view is focused) and you're getting the familiar "Find" dialog.

Enter the section name or a part of it and press the "Fine Next" button.

There you are... Where have you been? I've been so worried!

It will also work if there is more than one section matching, in which case SkinStudio will iterate through the sections when you press the "Find Next" button consecutively.

The feature will be introduced in SkinStudio 5 Beta, soon to be made available on Object Desktop.

It seems every time a larger company incorporates a feature that a smaller company has that a bunch of people start screaming bloody murder.  Shrieks of "copying" and "stealing" come up.  What's the point?  Last year the folks who made Konfabulator made a be stink when Apple incorporated Dashboard into the OS.  I totally agreed with Konfabulator's developers that it was pretty lame on Apple's part but in the end, complaining just looks bad I think to most people.  The fact is, if you make something good, other people will want to make something similar. That's life.

And I say that with some experience...

Object Desktop in 1995/1997

This is a screenshot of Object Desktop taken in 1997.  Mind you the product originally was released in 1995 and was in public beta testing in 1994.  It had features that were quite new (Windows 3.1 was the OS in 1994).  It provided featurs such as a skinnable UI for the entire OS (what we call WindowBlinds today), treated .ZIP files as compressed folders, allowed users to package up desktops (Object Package), Had a big sidebar on the right side with virtual desktops, system info, a clock, etc.

So when Microsoft and other companies started treating ZIP files like compressed folders, we didn't make a peep of complaint even though we were the first to do it.  Why? Because we didn't think the idea was non-obvious. 

Similarly, even though (to our knowledge) no product or OS had ever done a sidebar quite like this (NeXTStep's dock is pretty dissimilar), we haven't made a peep about other sidebars and such when they came along.

Even in 2001, when Microsoft incorporated GUI skinning into the OS using the exact same mechanism as WindowBlinds (uxtheme.dll in Windows XP might as well be WindowBlinds LE or something) we didn't complain or make a huge uproar. 

When in 2003 Konfabulator came out with a program that essentially did a subset of features of what DesktopX had been doing for 3 years, we didn't accuse them of stealing their ideas from us.  The only thing we objected to was their claims that they had somehow invented the concept.  It should be noted that we never claimed to have invented the concept.  We were inspired by IBM's work on an OpenDoc based desktop back in 1996.

And when we learned that Microsoft was going to be doing gadgets some time ago, we could have made a huge stink about that too.  Gone to CNET and Wired and other media outlets and talked about how Microsoft was "stealing" from us.  But we didn't.

Why don't we? Because nobody likes a whiner.  And in our experience, each change brings a new opportunity. Software companies must endeavor to adapt and change to new circumstances.  The only metric that should be used in my mind is whether a change is good for users/customers or not. 

To be sure, it's annoying when users will say "Oh, Stardock's screwed." Or even more annoying when someone else gets credit for what we were doing first simply because they squeaked louder.  But the reality is, in the long-term, changes bring new opportunities.

For example, looking at that Object Desktop screenshot from 1997 (which has the same features as were present in 1995) you can see a proto-Sidebar, Compressed Folders, ObjectDock, widgets/gadgets, WindowFX, WindowBlinds.  But changes have spark new innovation.

Most people don't realize this but WindowBlinds didn't really take off until after Windows XP shipped.  When Microsoft embraced GUI skinning, it opened the door to far greater opportunities.  It legitimized what we were doing.   Similarly, we had long discovered that users were generally not willing to pay money to run someone else's widget, we had already begun to migrate towards stand-alone widgets (gadgets).  But for us, the challenge was, how do we spread the word about why they'd want to use gadgets? Voila, Microsoft comes in and is going to do much of that evangelizing for us.  People don't care how their little RSS reader is made.   Does anyone care what compiler we use to make TweakVista? No.  Will people care what program is used to create gadgets? No. They will only care about what the gadget does.

Object Desktop in 2000.

This pattern will keep repeating itself as long as we continue to make software that extends and enhances the operating system.  And those that bitch and moan the loudest about being copied from will probably continue to receive unfair credit for "innovation" over us. But in the long term, I think people are more comfortable dealing with companies that look challenges as potential opportunities rather than looking for pity. 

Here are some screenshots of Microsoft Office 12.  Note the ribbons at the top. It's a very differnet design.

Okay, only 1.  I knew I should have taken some while I was there.

Windows Vista is going to provide an immense amount of opportunity for developers. It's going to be a real shakeup for existing developers too.  The reason I say that is that Windows Vista might as well be simply called VistaOS or something because it's a totally different ball game now.

Windows NT 3.1 was released in 1994. Microsoft gave it a new shell and moved the video drivers into Ring 0 and called it Windows NT 4.0 in 1996.  In 2000, Microsoft further refined it, made it more user friendly and called it Windows 2000.  A year later, they skinned it, changed the icons and called it Windows XP.

Building for the future..

Over the past 4 years, Microsoft has looked at where technology is going.  They needed to start replacing basic parts of the OS that were just antiquated at this point.  GDI? A 2D drawing system whose roots were based on the IBM VGA graphics specification (16 colors).  They've patched it up a bit with GDI+ but really, it has no concept of 3D, alpha blending, shaders, etc.  The networking in Windows XP is still based on a hodge podge of networking APIs that were built before over-wedgified losers made us learn words like "worms" and "malware" and "phishing".  So Microsoft developed a managed API called .NET and a new communications foundation code-named Indigo.

You put all that together and you have a pretty damn impressive OS.  On top of that, Microsoft is doing their usual bundle mania which tends to get on my nerves but monopolies do what monopolies can.

To support the new technologies, Microsoft is also creating some pretty impressive development tools.  Sparkle looks simply ingenius.  Let designers do their thing with Sparkle and hand over the XAML it creates to the developer using Visual Studio and they can do their logic behind it.  Microsot is really pushing Windows developers to stop making software that looks like..well crap.  Sparkle and the rest of the Expression Suite seems designed to do that.

Developer Opportunities

As a software developer -- a commercial software developer at that. My job is to figure out how best to take advantage of these new technologies.

In my mind, getting the most out of Microsoft's strategy involves leveraging two things:

Number one, it means making products that allow existing Windows XP users to be able to have some of the more obvious benefits of Windows Vista without being on Windows Vista.  There's plenty of reasons to upgrade to Windows Vista when it comes out, but users who elect to stay with Windows XP (and indeed, it'll be quite awhile before Windows Vista is out) should have some cool stuff to use.

To that end, we're working on WindowBlinds 5, WindowFX 2.5, and DesktopX 3 Pro:

These technologies will show up in the next month in the overall suite call Object Desktop.  Specifically, the Aero glass type technology will be available to WindowBlinds 5 users (beta due out this month).  We've got a bunch of skins in development that take advantage of the new glass technology in WB.  WindowFX is going to provide improved shadow support windows, an alternative Alt-Tab mechanism and window previews for minimized programs.  And DesktopX is going to evolve to more closely align itself with Microsoft's gadget mechanisms.  Gadgets created with DesktopX Pro will hence work on Windows 98, ME, 2000, and XP.

Secondly, it means taking advantage of the new technologies in Windows Vista to create new kinds of software.  In our view,  the Windows Presentation Foundation (Avalon) is particularly promising for creating several new programs to put into Object Desktop. And best of all, Microsoft is committed to bringing WPF back to Windows XP so those users won't be left out.   Stardock is already working on a couple of new programs to be made part of Object Desktop that will heavily leverage WPF and WCF (Indigo).

Which leaves us with the marketing question? When will Windows Vista be released?  Our guess: October 2006.

Vista looks good but they still haven't gotten rid of the ugly icons in some places.  PLEASE change the icons throughout the system. 

An alternative way of doing Alt-Tab. It looks good but I didn't find it that practical.  The primary Alt-Tab method is fancier than the Windows XP alt-tab but noticeably slower.  I like slick but I don't want it to noticeably slow me down and I do a lot with Alt-Tab.

The sidebar and task preview (something we are working on for WindowFX too).  The sidebar now is merely a container for gadgets which is absolutely sweet in theory.  I like the concept a lot.  However, right now, the gadgets are apparently implemented in DHTML.  I relaly don't see that as a practical thing.  I'd like to see them switch to have gadgets be Avalon based with the noborders switch on.  I think gadgets would be a great way for Microsoft to get people to start using XAML. Note: This is not necessarily what the Sidebar is going to look like in the final version. This was a beta version that was shown at PDC.

Gadgets can be pealed off and live on the desktop as well.

New file save dialog. I like the design but if you look closely, Windows classic can still peer through so there's still some fit and finish to be done there.

And for fun, here's a screenshot of my Windows XP desktop running Object Desktop with the WindowBlinds 5 beta (public beta in a week or two):

WindowBlinds 5 on XP can do semi-transparent borders and Start bars and runs fast on fairly low end hardware.  On Windows Vista, WindowBlinds 5 will make use of the new Windows Vista compositer to allow for much more advanced skinning than we've seen previously.  Much of the next-gen skinning will be back ported to Windows XP (though users with install the Avalon libraries on XP will likely benefit). 

Overall, Vista's a really good experience.  But I do hope there's an overall fit and finish guy over there so that things like ancient icons and inconsistent fonts get fixed.  Some of these icons have managed to survive since Windows NT 3.1 (1994).