Flex Community Feedback Forum – Attend or Open Source May End?

That’s the impression I got from reading this post.  I guess I’d be bitter too if I worked my tail off making something open source worthy, weeded through the thousands of bug reports on Jira to ensure I captured all the contributions, only to read not-so positive reviews about Adobe’s awesome efforts online.

If you read the various blog postings by both news sites and individuals you get the impression that people want to contribute to the Flex SDK, but feel they can only contribute bug fixes (and even that is perceived as being challenging), not actual features, nor being capable of being involved in planning sessions (aka, Fx prefix == wtf).  Sounds like a disconnect somewhere.

Adobe (aka Matt Chotin & Friends) have admitted it’s hard work and they don’t appear to be seeing the expected fruits of their labors.  My guess based on my readings, this means they expected more contributors who were non-Adobe employee’s.  This would justify the time the developers spend combing Jira for fixes vs. contributing to new features, and thus the intrinsic value the Flex SDK provides, thus selling more Flex Builder licenses.  Keep in mind, that’s a guess.  When the SDK was announced as going open source, no one told me how they were going to make money doing that; yes, I asked.  I just assumed it was to drive interest in Flex from the Java crowd as well as the open source zealots that only use open source software, yet are perceived as influential and have monetary value.

The tone of the post implies that if Adobe doesn’t start to see “results” from their efforts, those efforts will have to be scaled back.  To me, I’m liable to agree if I don’t start to hear them talk about improvements being submitted by the community more.  I’d prefer they spend more time creating better workflows; Catalyst v1 I can already tell isn’t going to cut it (yes, it’s still better than not having Catalyst).  I’m paying money for Flex Builder (although I attempt to use FlashDevelop as much as possible), and thus expect to see something that makes my job easier.  Flex 2 made my apps work.  Flex 3 gave me a faster compiler which was good enough.  The Open Source only seemed to give Adobe more interest in Flex, and thus hopefully a better bottom line, and thus more interest in improving the Flash Platform as a whole.  Did it really though?  I don’t know how much time some of the talented Flex engineers are munching on Jira bugs vs. coding new, cool, useful stuff in Gumbo.

I was probably the only one who didn’t care about the open source announcement.  I recognized its potential market value, but only to a small degree as marketing to folks who don’t want to pay for something seems backwards to normal business practices. Yet, I’ve seen the Spring & Hibernate revolution in action, and clearly businesses ARE contributing there because they have a financially vested interest (…they ARE contributing… right?). Then again, I’m still an OSS n00b.

Regardless, the blog world lit up, and that’s great and ultimately good for the Flash Platform as a whole.  But like Simeon said in his blog post, I have worked with a few gigs where I just hacked something in the SDK to work.  If I couldn’t find a solution on Google, I’d just hack it, and move on with life.  The reasons I haven’t filed any bug reports is because:

• The bugs are on old code; I found a bunch of issues in Flex 1.5, but Flex 2 was in alpha… so who cares about AS2?
• The bugs in the SDK come in 2 forms; extremely minor, so you just work around them, or make tincy monkey patches.
• If they are major, you just create your own component (aka, you can’t have negative padding using a TabBar since the tabs depth is hardcoded to their order, thus you can’t put a Tab “on top” of all others).  Could I fix it?  Not in a reasonable time frame modifying such low level classes in the SDK… but why?  It’s quicker for me to create a quick custom TabBar that does the majority of what I need it to do, give more accurate time estimations on doing so, and I can ensure I’ll complete it to do exactly what I need it to do in case something else comes up.
• With the ones that DO come up now… who cares, Flex 4 is coming out and 3.2 works well enough for me.

Frankly, I think my opinion is skewed.  I’m sought out specifically for my Flash experience, and ability to create custom components that don’t fit the common mold.  So, the thought of having alignment issues in an mx:Form component is a pipe dream.  Maybe the bugs I perceive as “minor” are in fact deal killers for some?  Who knows.

Additionally, I don’t mind using forked / hacked code.  I used to only use component code that I wrote out of fear, but I’ve been finding more and more that a lot of libraries out there, once you learn how the original developer expected you to use it, and you put some unit tests on it, you’re good to go.  BulkLoader, fluint, SuperImage, and Cairngorm/PureMVC… I’d be in not-so-good shape without them!  So, if someone creates a forked / different version of the Flex SDK, I’d use it if it works as advertised.  I can run a diff to confirm if you’re full of bs or not anyway.

Bottom line, while I may not have any immediate vested interest in the Flex SDK having it’s open sourceness improved, I certainly recognize that I, my peers, and the industry can be benefited from it if the contribution level improves.  For me, jury’s still out on the marketing side.  Although I’m not qualified to judge open source, having it end, or less resources put behind it seems wrong to me.  I just don’t have a lot of experience in the open source world, so not really sure what I can contribute, but maybe you do?  Details here.