The install process kicks AIRs butt. The sample app tweetanium downloaded as a DMG. I dropped it in my apps folder, opened it, clicked the customary OSx "you got this from the internet", warning and boom I was twitting.

Comepare this to AIR: •flash is required •specific flash version required •install air runtime •install air app •present REALLY scary warning to user (causes massive dropoffs in the installation process) •require 200-400$ cert from dev to make that warning 5% less scary.

I couldn't agree more. I've really wanted to jump on the "air" bandwagon but for the reasons you mentioned it's just too much of a risk. Why can't Adobe offer the developer the option to wrap an air application into a standalone piece of software? There was some work done by someone to do this and they created the "Shu Player" but Adobe cried foul. (http://shu-player.com/air-runtime-notice) Can anyone explain why Adobe is so resistant to improving the distribution process for air applications?

I couldn't tell you about the runtime install, but the scare screens are because Adobe, like Macromedia before them, are EXTREMELY paranoid about security. They don't want someone downloading an AIR app, saying "this looks legit," and then it deletes C:/ or steals their identity or something. The "we have no idea who made this and have no reason to trust them" screen makes sense in that context.

Of course, the exact same warnings apply to every single app anyone has ever downloaded off the internet, so I can't really agree with AIR's stance except in the purest abstract.

That the users even know something is an AIR app is an indication as to how broken that install process is. Compare with VB6 / Delphi / Qt / MFC - users who aren't themselves developers never had a clue.

Just this week, the Paint.NET folks were blogging about not being able to move to the latest version of Windows Installer because XP installs without SP3 won't have the runtime by default. For a downloadable application, extra megabytes are a big blow and extra required steps from the user are an absolute killer. Why would you take the upfront hit to people trying your software?

Java and .NET are still far less prevalent in the apps-for-sale market than they are in the inhouse-corporate market, and I think this is largely the reason why.

Exactly... why they feel the need to go so much further than Apple or even microsoft go with desktop apps is beyond me.

The sandbox security restrictions are HUGE as it is and add days of programing time already, why halve our install base with scare tactics?

We agree! The sandbox is ridiculous for a 'desktop' application. You can delete a users entire hard drive but cant launch an application! Shu (http://shu-player.com) does deal with the shortcomings of AIR and we also recently released a version that is compatible with Adobe's license agreement.

In general, beyond AppStore, there is no good download platform on the leading OSes. With this product, you still have to know how to drag the app into the App's folder and then you need to know how to dig around your Applications to find it.

If you can get past the AIR app install, it automatically launches it for you.

I don't know which would be easier for grandma to do.

I'd go with one step, dragging the app over to applications, (I admit it's a bit wonkey) over the 5 separate windows AIR manages to spawn.

Good Ajaxian write up here http://ajaxian.com/archives/appcelerator-titanium-breathing-...

God knows how they're going to make a business out of this. They raised 4.1M, so there's definitely an expectation of revenue.

Adobe has more resources and a significant head start. My money is on AIR.

(Full disclosure: I'm biased- Appcelerator recently laid off 3 of my good friends)

they can make money the same way adobe makes money

flex is currently free, soon to be open source I believe, they make money by selling you the IDE and tools to make development easier

I really don't see how Appcelerator can provide a better IDE (and charge for it), seeing as how it is all written in Javascript.

I can't think of any tools, short of some kind of WYSIWYG tool, which I don't think they have the resources to create.

They already tried consulting, using their product with all of their clients. That has failed.

I think they're banking on support, but haven't all successful support-backed open source companies started out as projects first, and THEN were turned into companies? Is it feasible to put the cart before the horse when trying to make money from an open source framework? (Adobe, Microsoft, et al can do this because they have massive amounts of money- far more than 4.1M)

I'm not trying to bash them or be antagonistic- I honestly want to know how something like this can make money.

agree... After watching their screencast I was asking myself those same questions. I don't see them making a dime, but maybe they have some trick cards to play.

Two questions: how is it different/better than Mozilla platform? And why is their demo app looks butt-ugly and doesn't resemble native desktop software at all?

We've gone this route on a couple of projects: and running offline web pages inside of an embedded browser window is not a "rich desktop application" - it's a web site running off your hard drive, there is a big difference.

I think the days of "native desktop" widgets are waning. One the web, there is no "standard" and people are getting used to it. I'm a fan of "native", don't get me wrong. But I'm seeing more of this every day.

No Linux client yet.

And AIR Linux 64-bit can't use the keyring properly. If only I could get one of the shoes twitter clients to work out of the box.

It's nice that they bundle the runtime. With fat hard-drives and blazing fast connections, it would be nice to put an end to the days where installing has been a serious headache for end-users.

yet-another "rich app platform" no one will ever use and no will ever care about

The beauty of this is that end-users have no extra install path. They download an app, and use it. For the developer, it's a matter of taking the same skills that make rich web apps, and make them run client side.