If you're the least bit interested in writing new consumer apps for Windows, please do it! There's an incredible lack of new, interesting, exciting apps for Windows. Enough to make one want to switch to a Mac (plus a hundred other reasons...)

I'm stereotyping a great deal here, but it feels like Windows users with a bit of expertise inevitably:

1. don't appreciate nice software (it's a rare thing, most programs are considered good if they don't have malware in them)

2. don't pay for software because they're 'smart' enough to use BT

3. don't care a lick about decent UIs because there's inevitably a free/open source program with that skates by on virtue of being free

The OS, as of Windows 7, is damn nice (though the terminal is terrible). The UI of the apps is almost universally bad -- most apps disregard the system UI theme for absolutely no gain, and they end up looking garish. Many apps that target the 'power user' demographic fall into a pit of endless customization, which presumes that users want to tweak their software using components written by random people on the Internet. It's as if no app on Windows is content to just do its thing, it always has to be striving to do more in a thoroughly mediocre fashion.

Your points are valid, but I think the percentage of inexperienced Windows users far outnumbers the users with 'expertise'. Those type of people

1. Do appreciate applications that "just work" and are willing to spend time choosing the right apps that suits them.

2. Haven't learnt of BT yet or even if they have, they're willing to pay to save the hassle of checking for viruses, trojans etc.

3. Do appreciate user friendly applications, they'll choose a better looking application with lesser features over a featureful app with bad UI.

Going after this crowd can be lucrative indeed, just ask patio11 ;)

Yeah, it just seems like it is hard to write an app that less-experienced users will both appreciate and use...much less pay for. After all, they may think MSN Messenger feels like it was written for twelve year-olds, but do they care enough to switch? Do they even know how to look for a replacement? (IM might be a bad example as nobody wants to pay for a client on any platform, it seems.)

OS X appears to simply have more of a tradition of independently-written, beautiful software.

Well, in most cases they want a real good reason to switch and the software must market itself as a replacement. Case in point Firefox and Chrome gained a lot IE6 users by

1.Having multiple tabs which is the killer feature

2.Marketing themselves as a secure and less-crashy alternative.

This might seem like an unfair example(not everyone has Google's or Mozilla's muscle) but the basics are the same for any type of software and given sufficient effort and planning, they can compete in competitive markets with established players.

I like the idea of paid, high-quality Windows software, but I have to agree with you that it's a difficult thing to make work. Trillian makes a very high-quality, and very Windows-y (in a good way!) IM client, and tried to go the free/paid route, but recently ended up adding ads to their free version, presumably because the paid version didn't make enough money for them to be profitable.

I do however, expect that as Apple becomes less and less of a computer for geeks (of various types) and more and more of a mainstream competitor to Windows PCs that it will suffer from more of the problems you outline. I've noticed that my non-techy, but Mac-owning friends don't perceive the different between Mac and Windows software in the same way that I do, and many of them also do a lot more piracy than I do. (I don't really pirate software except for Photoshop, primarily out of laziness.)

That is discouraging to hear. I've given up on the idea of being able to make any sort of living writing nice, general desktop apps for most any platform. Seems like you have to target a niche with a quality product so they'll have a need and a lack of mediocre competitors. No offense, but FOSS has done considerable collateral damage via the second point. Usually people would rather put up with a crappy UI from 1990 if it's open source than pay $2.

The gamer demographic on Windows is also a bit notorious for piracy.

This definitely could sum up the majority of Windows users, but there are those who think differently. I for one am a current Windows 7 user who would easily pay for many popular Mac programs if they were available on Windows. Also, I'm very picky about design, though that's due to my Apple influence from iOS. And, for instance, if Things, Pixelmator, Billings, and several other apps were on Windows, I'd buy they and hold off switching to a Mac...

Dropbox and Spotify come to mind as high-profile startups whose product depends entirely on the capabilities afforded by being on the desktop vs. just doing a web app.

Dropbox is mixed web and desktop, as a large part of its functionality is exposed (only) through the web interface. Its desktop parts are probably born more out of necessity than desire: a web app simply cannot "sync a folder".

I have never once used Dropbox's web interface (I haven't even bothered to look at it). What part of the functionality is missing from the desktop application? The only functionality I need or use is dropping files into the Dropbox folder and knowing they're not going to be destroyed if a hard drive fails or the house burns down.

restoring a deleted file and versioning comes to mind, if i am not mistaken?

To my knowledge, sharing folders can only be done through the Web interface.

'Events', a list of creations, changes and deletions.

I'm currently working on a hybrid app that's a desktop app and web app. The desktop code is mostly done, and the web code is coming along nicely. It's very interesting working on both sides of the equation; both have their challenges. While I'm still keeping the idea close to my chest, it's an untapped market which could potentially be very big.

I think there are plenty of opportunities on the desktop-side of things; millions of computers are sold every month. While we may be entering the long-tail of dedicated computer sales, being replaced by iPads and other specialized devices eventually, it's going to take a while for that switch to fully happen. At least 5 if not 10 years or more; the keyboard/mouse interface still rules for content creation, and many people spend their days sitting in front of one.

However, it's becoming very difficult to release a desktop-only application without some sort of web component. Everything from note taking software, to video games, to music software, is all expected to be connected online somehow.

I'd imagine a sticking point for some corporate software is security. Most companies are not comfortable having their data stored on someone else's servers, potentially unencrypted. Some even have special agreements with governments and businesses preventing them from doing so. A lot of these companies are being left out of the cloud.

Two possible solutions include providing server software for IT departments to deploy (been there, done that, it's not pretty or cheap), or to have some sort of encrypted solution syncing to the cloud. The trick is, data must be fully and securely encrypted "before" going to the cloud. If someone breaks into the cloud provider's storage, or intercepts the data in-transit, they must get gibberish. The cloud provider can't read the data either; only the business. Just an idea to think about. :)

When will this app see the light of day?

Definitely by the end of the year, if not sooner. I'm getting close to having a full functioning beta (likely in 3-4 months), and will likely do a private launch with some testers shortly after.

which idea for desktop apps you have?

How large and economically-successful a startup are you interested in?

I'm bootstrapping a native-client-software-based business. My business isn't successful yet, but we'll see how things play out in the next couple of years. I'm hoping that though web-based apps are all the rage, there still exists a large-enough market for desktop software. My economic goals are modest (though a wildly-successful outcome wouldn't be unwelcome).

For several years I've kept up with many, very-helpful startup-related blogs and sites. The vast majority of these recommend launching web-app-based startups. Even one startup owner who had hybrid approach phased out his desktop version (http://www.kalzumeus.com/2009/09/05/desktop-aps-versus-web-a...) So why am I developing desktop apps? The main reason is that I care passionately about user interfaces and the effect they have on my experience as a user. There doesn't seem to be a very consistent UI for the Web as a platform, not as consistent a UI as exists for Windows, Mac OS or GNU/Linux distributions. Web apps have their advantages, but UI-consistency isn't their strong suit (nor do I foresee this changing very soon).

A second reason for my focus on desktop apps is that I want to scratch my own itch and make programs that I want to use myself. I want to write high-quality software that 1) aids me with tasks related to various personal interests and 2) is desktop software, if possible. I want to exclusively produce "dogfood" (though not necessarily programmer tools). Having myself as a target user fuels my motivation to get up and work hard on making a streamlined, powerful, easy-to-use product, since I'll be using it. Yes, I'll still perform usability testing and try to understand my other target users. However, I believe it can be helpful to be a target user myself. I can't imagine how hard it must be to develop software for an unfamiliar field or making a program whose only value to me is as a source of income. (By the way, having oneself as a target user is generally considered by startup bloggers to be a mistake.)

A third reason for my desktop-app obsession lies in the opportunities I see to leverage the advantages of native client applications. Various articles lists the advantages of web apps. However, desktop apps have their own particular advantages, though many desktop applications fail to leverage them. I plan to keep identifying these advantages and capitalizing on them. At the same time, there's a lot for client apps to learn from web apps (and I hope to internalize those lessons as well).

As you can see, I'm a bit of a contrarian and somewhat stubborn about this particular issue. It may turn out to be my undoing. (I hope otherwise, of course.) Check up on me in a year or two... In the meantime, be careful about any advice you may draw from what I just wrote.

Check out my startup, Futureproof: http://iamfutureproof.com

Our goal is to build a large collection of small software products, including a mix of desktop and web apps, that help people form sophisticated relationships with technology. We also want to build "hardware" to support people in making healthy changes to technology habits.

Our first product, Awareness, is a C# app for Mac and Windows, already distributed through the App Store.

I'm actually looking to bring another programmer onboard, so send us your resume!

C# app for Mac? Is it through Mono framework or something else?

Yes, it uses Mono with MonoMac (http://www.mono-project.com/MonoMac). We built a platform-independent core library and application controller, then simple Windows.Forms and Cocoa views to support Windows and Mac OS.

That's really cool. I might get in touch.

Tower ( http://www.git-tower.com/ ) is a commercial Git GUI for Mac OSX that, by all accounts, is pretty excellent; don't know much about the team, but I believe it's their first product and doing OK.

I think you'll find more desktop-app-focused startups in Apple-land, largely thanks to the impetus to Obj-C & Cocoa programming that the iPhone has given.

Not sure if you'd consider them a start-up or not (http://www.skybound.ca/), but they have a pretty cool and fancy product. A few weeks ago, I read they are looking for solid dot net developers. You could check out. They are based in Canada.

http://yaletownsoftwaredesign.com is also Canada-based and writes some neat stuff for Windows. Not sure if they are hiring though.

What kind of enterprise app is better as a desktop app? Other than working if the Internet is down, what are the advantages of desktop over web?

There are things that are better done on the desktop. Take Spotify, for example. One of the key to their success is the fact that they keep bandwith costs low by using peer to peer technology.

I've done some research, and found that on average, 80% of the music I listen to on Spotify was coming from peers rather than from Spotify's servers. If I'm anything like an average user, that means that Spotify may be cutting its bandwith costs by something like 80%. That would be impossible to do as a web app, and web-based competitors face huge streaming costs indeed.

Flash has a built in p2p facility

http://www.flashrealtime.com/basics-of-p2p-in-flash/

So does Java :)

But seriously, I don't think Spotify could be the great app it is if it were a flash app. Can you imagine yourself running a tab in your browser with an app like Spotify inside and your fans running full-speed on?

IIRC, Chatroullete was streaming the both audio and video p2p over flash.

The desktop still offers far more support to developers than webapps do. Compare using Qt and Gtk to the mess and hacks required to build web user interfaces. The desktop gives access to hardware: web cams, microphones, printers, scanners, cameras, storage, 3D graphics and proper networking(http is a terrible protocol for a lot of things)

The web as a platform still has trouble keeping up with capabilities of desktop apps from 10 years ago. But due to having pretty much no reasonable competitor in the deployment market and deployment being so essential to a company's success the web will likely continue to be a dominant platform for startups and should eventually catch up to desktop platforms(probably 5 years or so).

>> What kind of enterprise app is better as a desktop app? Other than working if the Internet is down, what are the advantages of desktop over web?

Compute-intensive and graphics-intensive applications

Anywhere where you need local integration that is tighter (or easier) than what you'll get on [preferred web platform].

E.g. for a Windows app: reading files, connecting to Outlook, adding things to the registry, Windows Explorer context menu support, draw stuff on the desktop, plugin for Visual Studio, easier task-tray integration, better global shortcut key support, Win32 API trickery if you need to make controls do what they don't currently do.

Well some apps are only possible if they are on the computer. For my case I created a tool for letting IT agencies keep their software up-to-date on all of their workstations... this doesn't really make sense to be on the Web.

Where security is an issue. For example writing apps for small clinics/patient information.

It might not be considered a startup (anymore) but maybe Sofa (http://madebysofa.com)

Ninite http://ninite.com (YC W08 too!)

My friends and I are currently working on our first product retickr right now. Our Mac client will launch this summer and we are hoping to bring a Windows client to market this fall. Initially retickr will be "another" news aggregator but we are working closely with businesses and niche users to provide a tailored solution. http://www.retickr.com

Wunderlist, Hype, Spotify, Sparrow. Just a quick list of desktop apps. But I agree, most startups are focussing on the web (and rightly so.)

The Wunderlist desktop app is also web-based. It's using Titanium Desktop as a wrapper.

Dropbox too

Not really a start up per se, though I am working on it full time:

http://shavevideo.com/

Just got a great review in MacLife:

http://www.maclife.com/article/reviews/shave_video_review

DeskMetrics (http://deskmetrics.com/, http://www.crunchbase.com/company/deskmetrics) comes to mind, they're doing analytics on the desktop.

(I've got no relationship with them)

Thanks for mentioning us. Just noticing: we are a "hibryd" startup: we have a focus on Desktop market, but we have to pay attention on the "Web side of the force" too. :-)

Sometimes you have to stay on the desktop out of necessity. Can't intercept low-level keyboard events or manipulate the file system, for example, from the web.

Not a 'startup' per say but I'm in that boat. If I wrote this as a web app, it would work only on my website. As a desktop app it works in every open text-field, regardless of the program. http://onehandkeyboard.org

Also check out the the Business of Software forum, lots of desktop based app developers there. http://discuss.joelonsoftware.com/default.asp?biz

HTML5 game creator with C++/MFC desktop editor application: http://www.scirra.com/forum/viewtopic.php?f=35&t=8468

Evernote's Desktop App is a pretty big part of their product. Here's their open source page which reveals some of their technology choices:

http://www.evernote.com/about/opensource/

I think they got the Webkit license wrong. Just like the LibXML License, it says:

> Copyright (C) 1998-2003 Daniel Veillard. All Rights Reserved.

We're building our MVP in wpf, with the intent of supporting silverlight at some point. Being B2B, we think its less of an issue than a B2C company. Our platform requires access to local databases, which isn't possible directly through a web app. Wpf and silverlight give us the ability for an enriched user experience while still being able to support web distribution through xbap and xap.

Soluto is (http://www.soluto.com/). They won TC Disrupt last year.

The only real desktop "Startup" in recent times is doubleTwist. Most other are either client for a web service, or more of a lifestyle business than a startup.

On a sidenote: That doesn't really count either, but MacHeist has done fantastically well selling Mac desktop software in bundles online.

It's not about the platform, it's about the idea which implementation other people will be willing to buy.

http://udeployer.com/ here :). I'm trying to fix Windows software deployment by providing an apt-get like tool for IT agencies but more targeted to deploying to multiple remote computers instead of only your local machine.

Xobni is an option though I wouldn't consider them a startup anymore.

We do WinForms apps in .NET over here at www.dextronet.com and .NET components, namely Better ListView, written in C#, at www.componentowl.com

Some startups have a desktop component (Yammer), and it's often not very good.

Yammer's desktop client sucks. Perhaps it is because writing for the desktop is considered so passe that the art is dying among younger devs.

Yammer's desktop client is written in Adobe AIR which is the normal web stack. It runs in an embedded WebKit browser.

try http://www.appcelerator.com/products/titanium-desktop-applic...

Evernote