3 ms·
Wouldn't the more obvious conclusion be that Adobe believed 10 years ago that the way to make money was to cater to those that don't want to program and just wa
by citricsquid 14y ago
Wouldn't the more obvious conclusion be that Adobe believed 10 years ago that the way to make money was to cater to those that don't want to program and just want something that "works" and now they've decided that there is a large enough user base of people that are willing to pay for tools to help with real programming?
I don't think these products would have been at all successful 5 or 10 years ago, the web "wasn't ready".
> Adobe has been steadily building steam in this area for years, but they have never marketed their efforts so well as they are today.
If Adobe have been building these products for years that would mean they haven't been "not understanding developers": they just didn't guess 5 years in advance this is where the web would be. Silly contradiction.
- Wintamute 14y agoUp until a few years ago I was a full time Flash developer, working in ActionScript3 which is a strictly typed, fully OO language which conforms completely with the 4th edition of ECMAScript. I used a great IDE (FlashBuilder) that had full code introspection, static analysis, amazing auto-complete and all sorts of great features built in. Unit testing was not uncommon, as was continuous integration, automated build processes and version control. I, and many others, were doing "real" programming on the Flash platform and felt totally understood as a developer by Adobe. I'm not going to defend Flash's continued relevance on the web because like most people I think it's had its day but from about 2003 to 2010 the Flash developer community was massive and thriving, people were doing serious programming and Adobe was doing a pretty good job of understanding them and supporting them. It's this wealth of experience that I hope Adobe brings to bear on its HTML5 developer tools. I feel like the relevance of Flash (and Adobe) is really quite poorly understood. A lot of the expertise and good practises from the more serious elements of the Flash community flooded into the JavaScript community and I feel this is one of the reasons JavaScript has developed so quickly. What's more, a lot of the "web2.0" style dynamic and interactive elements of websites that we take for granted now are watered down (and much better) versions of ideas that were conceived during those frenzied years of UI experimentation in the Flash community. Granted those years of Flash threw up some UI abominations, but it was also a melting pot of ideas and creativity, the like of which we don't really see anymore, which is a shame in a way. Even though Flash was my livelihood I was happy to move on because I could see it was for the best, but if you ignore Flash your understanding of the last 10 years and the current context of web development is impoverished.
- frozenport 14y agoYou made me nostalgic for my childhood. It is important to note that for half of the glory days it was Macromedia Dreamweaver. Until FlashBuilder, we were told with straight faces to use the Adobe Flash Professional. Remember, FlashDevelop? I think adobe figured it out maybe 2 years ago.
- ktizo 14y agoThe way things went with flash after the macromedia acquisition was part of what got me into using haxe.
- Wintamute 14y agoYeah, FlashDevelop was (and still is) ace! I think Adobe were beginning to figure it when they started to let you do AS3 only projects in FlexBuilder, whenever that was, maybe a little longer than 2 years ago. FlashBuilder is a brilliant IDE IMO, there's certainly not equivalent in the JS world yet.
- untog 14y agoI needed to make a plugin for our video player recently, so I downloaded FlashDevelop and gave it a go. I was astounded at how good it was for an entirely free piece of software. It's no wonder Adobe gave up on Flash, they certainly got no money from me while I developed with it.
- seanalltogether 14y agoIn my opinion Flex is still the best UI framework I have ever worked with, and I've worked with many. 1. Source code for the framework was available in the default flex builder installation. Wanna see how adobe designed the default table renderer, drop right in and look. You didn't have to search online for how to use esoteric features, you could just look at the super class to see the code path it takes. This is my biggest complaint working with Apple and MS frameworks. 2. Proper handling of component invalidation. Many other frameworks still have problems with calling methods out of order during initialization. Flex collects all changes to a component before processing them in the invalidation loop. 3. Data binding let you do some really cool things by automatically wrapping them in anonymous functions to process like <mx:Button id="myButton" enabled="{isButtonEnabled(someString, items.length)}"/> 4. Lots of support for component customization, especially in the skinning department. Going from flex back to mac cocoa was like going from paintbrushes back to using twigs to draw. However, the biggest downfall for flash was always the performance, they were stuck in a corner because the flash engine was designed first and foremost for processing frames of animation, on the cpu, in another programs window, with no control over their own run loop.