The random utterances of David Arno

This week, Adobe held a summit on their plans for handing Flex over to the Apache Software Foundation. A small sample of the Flex user base was invited to attend, but for the rest of us they made recordings of the first day available, and streamed the second day live, which was much better as it gave a sense of taking part, rather than just passively watching what had gone before. Others have summarised what was covered in those two days, so I won’t repeat them here.

The summit seems to have divided opinion: some feel that Adobe are abandoning Flex to die; others feel it is an exciting opportunity for the Flex community. To my mind there is no dichotomy here: Adobe are abandoning Flex and that creates an exciting opportunity for the Flex community. It also creates some challenges and there is a danger that many in the community seem to be in denial of those challenges.

Flex is dying. When I say that, I don’t mean that no one will be using it in one, five or probably even ten years time. Cobol is the cliché ancient, dead, language. Yet a quick google of the phrase “Cobol jobs” will show that there are still people being hired to maintain Cobol projects. So when I say that Flex is dying, what I mean is: it is heading the way of Cobol. It’s use will decline over time, few (if any) people not yet involved in producing Flex applications will learn it and so the community will itself slowly age. People will drift away from it. It will stagnate and will effectively have died. One clear indication that this will happen is the end of the design view in Flash Builder.

The common reaction I’ve heard to the news that the design view will go with the next release of Flash Builder is “well I don’t use it anyway, so it’s no big deal”. This misses the point: you don’t use it because you already know Flex well enough not to need it, or to even find it slows you down. This is not true of people new to Flex. The design view is a great introduction to the component set and how they work together. Without design view, a barrier to adoption by fresh blood has been erected. The Flash Player (on which Flex is 100% dependent) is losing favour with a growing number of people and it is no longer seen as “cool” by younger programmers. That crown sadly has gone to JavaScript. Add to this mix the fact that key people involved in Flex (eg Andrew Shorten and Deepa Subramaniam) are being moved over to HTML5/JavaScript roles and we clearly have a problem. Adobe might claim that it “does not currently have plans to develop any product targeted at enterprise scale user interface development” (I’m quoting Adam Flater here, not Adobe), but that is because they have decided there is no money to be made from enterprise-scale tooling. You can be sure that they are developing tools that will appeal to people who currently use Flex to create small-scale applications.

Flex is dying, I think this is an unavoidable conclusion. You could react to this news by donning your slippers and cardigan, lighting your pipe and settling down to a quiet life working with “Flobol” for the rest of your career. Alternatively, you could “jump ship” and start using other technologies. More radically though, you could help re-invent Flex to keep it appealing to the masses. To my mind, decoupling Flex from the Flash Player (and probably targeting HTML5), improving the AS3 language, smashing up the horrendous component inheritance chains and slimming down the Flex core (farming many components and data types off to separate libraries) are priority actions needed to save its life. Adobe are looking for volunteers to get involved in being committers to the Apache Flex repositories. So if you want to help save Flex from death by apathy and stagnation by helping rewrite it, you need to contact Deepa Subramaniam and request you be made a committer.