The One Feature I'd Like From Adobe AIR

Comments (Archived):

1. 1. AIR doesn’t run on Linux yet (but Adobe promises it will)2. Why would you want separate applications linked as tabs based just on the fact that they are running on the same framework? Will you want all your cocoa apps as tabs on another mega-application too?

   1. Thanks for the clarification on linux. My badYes, this is what I’d like. It might not make sense to others but it makews sense to meFred

   2. Actually it does run on Linux, it’s in beta in Adobe Labs, and there’s already a Flex Builder for Linux too.(the Flex compiler and SDK has been available for a few months already)

   3. Adobe Labshttp://labs.adobe.com/downl…

2. Nice thought Fred, and along the lines of what I’ve been thinking about – am sure there will be an Air uber app framework play before too long, which will enable this.I’d actually like to see it go beyond this so that you can have an Air-based desktop, and where the individual apps are components within that desktop (whilst still retaining any alerting/notification functionality that you want). I’m a Vista user (for my sins), and I think Microsoft has wholly failed in their beyond-the-browser vision (from an executional perspective). However I think Adobe, with Air, is potentially in a great position to create a brilliant rich desktop environment which is all about the passive consumption of your favorite services (the desktop is not dead! It’s prime real-estate that just needs to be enabled properly). I can see this becoming the new startpage paradigm; leaving a browser tab open with your startpage on increasingly annoys me, and my preferred startpage, netvibes, isn’t doing anything to push itself out beyond the browser (it has got a nice mobile service though), which reduces its accessibility and utility – the desktop is a much more logical environment for this type of service/usage/experience. Definitely a bluring of web/desktop, but don’t see a problem with that. Then it just needs to mobilised too to create a truly seamless and integrated cross-platform/device experience. Can’t wait for mobile home screen to actually have some value beyond showing what the time is, what network I’m on, how many emails I’ve got unopened, and some basic calendar integration, etc (I want to passively consume my favorite services through that too). Bring it on, Air! And if this does become a reality Adobe’s position/potential on the web, and the desktop, will have been massively increased in terms of being a key layer/environment through which services are really enabled and made accessible on devices.http://www.twitter.com/chrsozhttp://www.boxedup.com

   1. Great commentI am going to reblog on my tumblog

   2. I blogged about this tangentially a while ago, the last paragraph of this article.http://www.kalio.net/blog/?…

3. Sounds like a great idea.After all applications start working this way you don’t actually need a desktop environment anymore. (your desktop, taskbar/dock are applications too)The obvious loser here is the Windows desktop, and I love it.

4. Web applications and sites are typically boundless (and if not, autocenter on a page) which makes a tabbed, single-sized container for them possible. AIR apps (and for that matter all desktop apps) typically have content that extends to a definite boundary (if they didn’t then they’d take up unused desktop real estate). How would a “tabbed” mega app work? Would it just be a giant empty window (harkening back to Windows MDI, *shudder*) large enough to contain the largest app in the group?I think the idea of desktop app context switching in place is interesting, but I’m not sure if a tabbed interface is the right way to approach it

   1. I think the age of the “mashtop” is coming.Tabs are just user interface elements, allow any app to be tabbed or clipped, et voilá.

5. As far as normal users are concerned, AIR Apps are no different than any other desktop app. Tying more than one desktop app together might better be handled by the operating system, no? Windows begins to do this with the taskbar, Mac with the dock, and Linux with whatever they call their app bar. Sounds like all you’d need added is the ability for multiple applications to share the same “physical” window (possibly with separate in-window navigation between the apps).

   1. I blogged about this tangentially a while ago here: See the last paragraph of this article.http://www.kalio.net/blog/?…Whilst the article was all about discussing the “social graph” problem, and what Web 3.0 really means, I tried to shed some light about how we’re needing a technology to enable “mashtops” to surface (as in: desktop mashups) bringing the flexibility and the ubiquity of the web to the desktop.This doesn’t necessarily mean the Web 3.0 *IS* the concept of mashtop, just like Web 2.0 is not exactly defined by Ajax mashups alone.But I can see how –almost inexorably… yet ironically– information at your fingertips is really becoming true.Finally.

6. As an air developer of my own application, I can say that my my windowing needs wouldn’t play well with other applications grouped in a “megatab”. Such a grouping would likely present the same restrictions a browser does.Maybe what you want is a specific group of clients all rolled into one… mega client. Twitter + Pownce + tumblelog + _______ which would allow you to post to any or all of the selected services? If the app was smart, it could import plugins for posting to ANY service.

   1. What you suggest is the same monolithic approach to windowed applications of the last 20 years or so. (only somehow mixed together)One such type of “pre-mashed” application is Flock, I’ve used it and I don’t think it’s all too great.What the web has demonstrated is that *users* want mashups, exchanging data between applications and using them in ways not anticipated (or even intended) by their designers.It doesn’t matter what *I* as a developer want.Of course I would love to continue developing local, monolithic binary apps for the IBM PC circa 1985 and charging tons for it.But that’s not how the world works anymore.The world has moved on.

      1. I’m not recommending a monolithic approach. One app will never do all things well. I am simply suggesting implementing similar APIs in one app so that a content item you create can be published across your web space, with a little effort as possible, with as few windows as possible, which seems to be the root of Freds problem.[snarky]I’m sure what mashups has to do with that but hey… run with it, it’s what the users want, and the world has moved on and such. [/snarky]; )

         1. Your idea sounds exactly like the Firefox plug-in architecture, and Fred uses that too.Have you used plugins like Delicious’ for example ?The problem is not really about apis, but about the presentation layer.How do you combine XUL apps with Java Swing apps ? You can’t.Both can use a remote service api, but you can’t build a new visualisation perspective, or interchange data between those two apps.What I’m talking about is different, a common runtime system that allows applications to run side by side and communicate among themselves.I don’t think AIR is the end of the road by any means, but it’s a good start.

   2. Yes. Something like that. A plugin architecture would be great. Bottom line is FF has allowed me to control my interaction with web apps and a desktop client for web apps should do the sameFred

7. I think this speaks to the uselessness of the OS task manager rather than the fact that a bunch of sortof-unrelated apps all happen to be written in the same framework.Back when I used Xwindows, I used a window manager (pwm) that let me put any windows I wanted together into the same frame – even from different applications. Example screenshot at http://modeemi.fi/~tuomov/i… – you can see multiple apps in each tab.I think one of the core ways we deal with information overload is to group things into larger functional chunks, and Windows doesn’t let you do that for your running applications. Firefox does for all your web sites/applications, but it’s not enough (See Mozilla’s Prism at http://labs.mozilla.com/200… which turns webapps into top-level applications.)I’ve vented similar frustration over this kind of stuff: http://joshua.schachter.org…My personal solution? Two monitors; one for working and one for IM/email/chatter/etc.

   1. Awesome comment and link. I am not sure my suggestion is the ultimate solution but I know I want less apps that do more thingsFred

8. Seems counter-intuitive. The common complaint about webapp-in-a-browser is that certain apps occupy enough of your attention to be treated as a separate task. You avoid clicking twice to reach the app because it is fundamentally higher priority than tabbed content. I think a more meaningful way of organizing lightweight, high-touch apps such as Twhirl is via Apple’s Dashboard feature.

9. The problem with AIR is that unlike Firefox, which allows for tabbed browsing, there is no way to launch one AIR client and get multiple services. I’d like to have Twhirl, AlertThingy, and Shifd (another cool AIR client) all running at the same time in different tabs.Why on earth are we not concentrating on integrating more application space inside the “browser” (or some metamorphosis of it)? It seems to me you just want a richer Web User Xp. (which we already have a base for). The ‘winner’ of the AIR/Silverlight battle isn’t going to be the overriding service that sits on your desktop, it is going to be the one that runs in a webbrower when your browser becomes your desktop.

10. The other part of the AIR story is code reuse across multiple environments (desktop and online). We launched uvlayer, an enhanced online video experience that utilizes the YouTube API and integrates with Facebook, AIM, and Gtalk, initially as an AIR application for the desktop, then we were able to take roughly 80% of what we had done in AIR and put it online using the same code for a in-browser Flash experience at http://www.uvlayer.com (still in Beta). As you can imagine we are hoping to take this experience wherever Adobe can get Flash to run in the near future, mobile devices and set-top boxes, however now we can let the users decide where and how they want to experience uvlayer.As software creators, it is like being a painter and not having to create the paint or colors we use, we just need to create the masterpiece or user experience and leave the underlying tech to Adobe, I should add we had built a proprietary C++ runtime before AIR came out and we were more than happy to switch it out, a decision that allowed us immediately to run on both Windows and MAC and now Linux.

    1. I think that’s an interesting idea. You might like Fluid if you’re on a Mac: http://fluidapp.com