Introducing Atlas: A Visual Development Tool for Creating Web Apps
carsonified.com
carsonified.com
At the same time, this isn't exactly new. GWT did something similar with Java (although it wasn't necessarily as polished) and as others have stated there are similar things.
I guess the question is - do people really use this, how often, and do THEY make money?
It is as if some start-up business models are merely based on creating technology that either another company will want to buy outright because it runs parallel to their goals and portfolio - or because larger competition doesn't want to have to pay dearly for not buying them out while they could.
Acquisitions don't result in lasting innovation in my eyes - it increases proprietary code bases. Early exits are like leaving a cash poker game after your first big hand, the fun was just beginning.
I hope to see more and more "start-ups" use less and less VC funding (we do realize that it is the VCs who have the winning scenario here, RIGHT?), say "F U" to acquisitions, and stay around long enough to make a sustainable difference.
My pondering was merely a function of - is it enough to build a business?
For instance, if Google wanted to become your one-stop-shop for online desktop-grade web applications (aren't they kind of already? GWT?), developing a tool like this in house would be a fantastic idea as it would increase the productivity of the developers.
But that's Google. The exception not the rule.
How many desktop-rich web applications have "made it." Facebook, Google Applications, the newly exiting Mint come to mind. How many of you hackers here have created incredible technology with 3 hits a month and no paying customers?
I guess the point I'm trying to press here is how many tech companies can exist to feed the need of other tech companies (which presumably make money from other sources)?
Start ups like 37 Signals were smart as hell because they saw the demand for simple services (communication, version control, project management) from a large pool of willing customers ready to cut their costs (other tech start ups).
But where do THOSE start ups get their money? VCs? Acquisitions? (That money only counts if the founders reinvest it in the industry) This begs the question - are the large successes like Facebook/Google/Twitter/Mint/etc feeding a working-class of willing tech "entrepreneurs" for money that simply doesn't exist?
I encourage debate as this seems like a gaping hole in some Valley-Hacker mentalities.
maybe that depends on your definition of 'made it'. Personally I think any self-sustaining business is a sign of success - and using this definition there are many thousands of successful software developers out there making a comfortable living building rich web apps that you'll never hear of simply because they don't fit into the social media zeitgeist
It's the social psychology principle at play - what stories can we think of that are easy to recall?
That said - does anyone know of a desktop-rich web app that is in the black from charging to use it (That isn't already very popular)?
Well, considering they could probably charge $100-200 for such an IDE, and that there are now a very large (100K and growing) set of Cocoa developers (thanks to the iPhone, mostly) who can be immediately productive in such an environment, delivering apps on Web, Mac, and Windows, ...
Seek to 23:10 - http://www.youtube.com/watch?v=j02b8Fuz73A
(Man, look at that hair!)
I don't mean to downplay 280north's work - I'm very impressed with what they're doing. It just bothers me when people see a derivative work and think it to be original. (Ahem, Chris Hughes)
I've always viewed Cappuccino and Atlas as front-end only tools so this part of the presentation (Q&A) was surprising. Does anyone know where I can find more information about the server side of things?
beautiful, beautiful piece of art and engineering.
Id seen some of the RaphaelJS demos, so I knew this was coming.. but honestly was expecting this stuff to become a product in about 2 to 3 years time... and this guys integrated the whole shebang.
I just want to see the same thing happen with data - people will realize that most data is a graph and not a table, when and only when we can demo something that's fluid, has the same qualities of immediacy and usability and solves a real problem [for which the spreadsheet is the current workaround]...
Each component of this might have been done before, but the combination is a new new thing we all knew was just around the corner.
I spoke to Francisco afterwards and he said that the resulting code would be able to run across both Mac & Win32 platforms, which I thought was nice.
The interface files that Atlas creates are essentially freeze-dried archives of the interface objects. IMHO this is a better solution than code generators.
I unfortunately don't have a chance to watch the video right now but I'm going to go ahead and say congrats to the 280North guys anyway because their stuff always amazes.
But it does seem like a huge step toward liberating developers from writing front-end code. The biggest possible downside I see to it that it could also liberate us from writing truly beautiful apps and we'll have a million apps that look like the standard interface.
It would be awesome to see some demos of apps not looking like 280 Slides.
Atlas is going to make it pretty easy to customize the standard controls, so I wouldn't be all that concerned about cookie cutter interfaces.
It's also worth pointing out that most people don't have the ability or resources to create beautiful interfaces. Those people will get something that looks good by default.
Edit: Not as impressive, but I believe it's the same concept - http://media.tibco.com/gi/2_create_gui.swf
edit: I see you guys just changed the 280atlas site.
And they just announced that Atlas will be available November 15 in beta for $20.
Picurio (YC W09) has a browser plugin to allow uploading of photos directly from your digital camera. The GUI frontend is built with Cappuccino and integrates with this plugin. http://picurio.com/
Also, I see from the comments on carsonified that they're going to charge $20 for the IDE. If anyone watches the video, I'd be interested to have that verified.
Do not ask people who do not pay money for software how much money to charge for software.
If they're trying to undercut their competitors, I think it may be an error.. Charging $20, rather than $400 will make people think it's cut-rate, and not up to the challenge. No one who's making a professional/moneymaking website will be stopped by a $400 charge.
This makes me suspect that perhaps they're after the Wix/weebly market. Drag and Drop a website, rather than a professional-level tool. These tools are consumer facing, and don't generally resemble the Atlas versions I've seen much.
Don't get me wrong, I'm excited about Atlas, and look forward to buying a copy, but I'd rather do so knowing that that company is making enough to hire competetant engineers to keep expanding it! ;)
Internet / web application platform is not likely to go away.
Lightly developed on NeXT Obj-C, spring controls ~20 years ago. Glad seeing good NS ideas being adapted and extended.
He equates desktops to VHS and the web as DVD, the only problem is there is a clear superiority of DVD over VHS (greater resolution, more features, compact form factor) where there is no clear superiority of the web over desktop in terms of usability and features.
If someone can name me one app that is better than the desktop equivalent, please do so.
Flickr is no where near as useful for image cataloging and management than many desktop equivalents. It wins in sharing for sure, but not actual functionality.
Outlook (and I'm sure the xobni guys would agree) trumps it's online version by a mile. Gmail still lags here as well.
Video editing ... Jumpcut? Yeah right.
Audio editing ... show me one that supports high end audio effects or plug-ins.
Image editing ... Aviary < Photoshop. Aviary < most freeware image editing tools.
Google docs - ever press the tab key when editing a doc? Dragged and dropped an image into the document?
I could go on and on, but you get the point.
If you are going to bring a metaphor and present it as the truth, there had better be some truth to it.
Atlas is cool, but derivative of all that came before it: Delphi, Interface Builder, Visual Studio, Dreamweaver, Visual Basic, RealBasic, etc. Sure it runs in the browser, sure you can target the desktop. I dunno, color me unimpressed.
But, listen, if you're going to downvote me at least come correct.
You said that outlook is better then gmail. While you might be able to make an argument when you compare it side by side one one machine. You would be mistaken in making that comparison. Gmail kicks ass for me not just because I feel it can replace outlook, but I can have my gmail, on my iphone, I can access it from any computer that has a web browser, I can access it from the linux command line using lynx, or links. Can outlook do that.
You need to realize that when they say desktop apps aren't as good as web apps, I think what they are saying is that desktop apps aren't mayby the best way for most people to use apps.
I get Gmail having more utility than say Outlook for certain things. But I can access my exchange server (I use google apps for business, but used to use exchange) in all the same places I could access Gmail which renders the point slightly moot.
I have no issue with people disagreeing, but at least argue the disagreement if you feel strongly enough to vote me down.
But that is where my fascination with 280 ends, they are taking a great language, Javascript, and turning it into something its not. Take a look at Objective-J its not javascript. It doesn't really even look like javascript.
As a JS developer, I am not certain I will be willing to give up JS, just so I can use these visual programming tools.
I am going to spend the 20 bucks though, I really want to see how much of the Objective-J I can strip out just use JS.
To an objective-C programmer, it looks like home :)
And in both cases, you can drop down to the native language whenever you want. In objective-C you can write straight C code, and in Objective-J you can write straight javascript if you want.
Code importing is clearly a feature we can all get behind. Classes are clearly something that almost every JavaScript library uses. We make those things language features instead of library features.
The unique feature we add, and the reason we really have to introduce a custom syntax at all, is dynamic message passing. Message passing enables all kinds of cool new features, because method implementations aren't bound until the message is received.
If you really wanted you could do this yourself:
objj_msgSend(anObject, "message:", aParam);
[anObject message:aParam];
Besides if they feel they have something JS developers want why not take it through the standards bodies.
Standards bodies are hopelessly slow. If I wanted to wait a decade to get a new feature, I'd propose it to a standards body. I actually want to write programs.
Part of the power of Objective-J is taking control of the language away from standards bodies. Objective-J can evolve at any pace we like, and anyone can fork it and make their own changes.
JavaScript does have a long history, but it's an awful one. It's a history of being a second class language that was slow and that was not used to produce anything significant. Only recently has that changed. And there is now a growing culture of implementing languages on top of JavaScript. Python, Ruby, Smalltak, and others have all been written on top of JS. GWT compiles from Java to JS. There are similar compilers for Lisp, PHP, C# and others. It seems narrow minded to assume that only the flavor of JS that has existed for over a decade should be allowed to exist. That's why we build abstractions.
We are not rewriting history nor are we moving far away from what JavaScript is. Most of the code you write is pure JavaScript anyway. You can do as much in pure JS as you like. We are making the programming environment easier to use, should you choose to take advantage of that.
and its great to see that people are working on removing the hurdles that it takes to create web/html application.
but i wonder how this whole web app stuff breaks up once we move away from WIMP to more task oriented applications with NUIs.
From what Francisco said, it sounds like they may still plan a web-based version.
Often the problem with these types of high level abstractions is that there is so much magic going on behind the scenes that customising how things work becomes difficult and you end up getting locked into the 'Atlas' way of doing things.
I can't wait to get a chance to start playing with this though :)
Other than that, it's just like Objective-C + Interface Builder (or better; more dynamic, more portable, no compile/link/go slowdown).
So you can expect generally the same amount of work to build stuff as you would expend building with Cocoa on the Mac/iPhone.