incubator-flex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Daniel Reicher <danreic...@gmail.com>
Subject [OT] What are we doing here?
Date Mon, 12 Mar 2012 21:42:41 GMT
Apologies in advance if this ruffles some feathers...

As a Flex developer and, of late, professional Adobe apologist the past few
months have been challenging. More recently, it seems that frustration has
reached the point of rolling boil.

Let me first state unequivocally that I still believe Flex is the best RIA
solution available today for a wide swath of applications and the current
ecosystem will live on for many years to come in some capacity. What
remains to be seen is what that capacity is and if the ecosystem grows,
shrinks or stays relatively the same. To be honest, given the current state
of affairs, I don't feel very optimistic about the trajectory of things.

I don't think Flex is going to go away any time soon, but I (personally and
selfishly) don't want to be the guy locked in some broom closet maintaining
some second-class citizen, line of business Flex application(s) because
noone understands them enough to rewrite them in HTML/JS/CSS and noone has
the appetite for greenlighting any new projects based on Flex. If I wanted
that, I'd learn Fortran.

Here is what I "know" today. First, the people making technology decisions
(the suits) have read just enough to "know" that Flex is Flash and Flash is
a bloated and insecure technology, doesn't work on their iPad or iPhone and
HTML5 is the way forward. Adobe has "abandoned" it and has explicitly
stated that HTML5 is the best option for RIAs in the future. So, unless
you're doing video or gaming - nothing to see here, move along.

They "know" all of this for some very basic reasons. (1) There is a very
large ecosystem around JS/HTML5 and a lot of buzz. (2) That ecosystem is
just as interested in money as anyone else and are selectively pulling
information to suit their agendas. (3) Apple doesn't support Flash! Adobe
abandons Flash mobile! These are the sensational headlines of the stories
that are in the rags. (4) Adobe fumbled some messaging along the way. (5)
Most importantly, nodbody is changing the narrative.

So every day I come to the mailing list for a ray of hope. Looking for some
glimpse in to how we as a community are going to respond. How we as a
community are going to change the narrative. How, with still (arguably) the
best RIA technology available today we're going to start acting like it as
a collective. We discuss cross-compilation to JS, compilers, build
environments, component proposals, coding standards, infrastructure (jira,
test frameworks, etc.). And, frankly, it makes me want to scream.

Decisions are being made TODAY on what technology to use for upcoming
projects. Decisions are being made TODAY regarding whether to continue to
grow Flex projects or cut losses and begin migrating them to HTML5. We have
a stable, extensible and proven solution for providing cross-platform
(including mobile - even Apple!), rich, immersive and visually appealing
software TODAY. We have professional tools TODAY. We have a solid set of
components out of the box TODAY. Why does it feel like we're playing catch
up or having to sing for our supper?

The decisions being made today are being made by people who don't care one
bit what build technology the Flex SDK uses. They don't care if the SDK
compiler is efficient and quick. They don't care how its tested or even if
its tested - they care that it works. They want to know that given a set of
problems it offers a set of solutions and building blocks. They want a rich
component set and something that works on the desktop, tablet and phone
without having to build 4 or more different applications. They want to see
that the framework is improving, growing and is supported by someone or
something.

Considering all of that, the only thing I don't see happening is the
growing/improving and I'm beginning to get a little nervous (with a side of
impatient). I don't need anything from the framework, but I do need Flex to
be viable in the marketplace and that, at this stage, requires a growing
community. It requires changing hearts and minds by making it easier to
choose Flex for organizations and as much as it pains me to say, if Adobe
has now become a bottleneck to that happening, we need to move forward
without them and integrate any additional work they make available when it
is available. If there are things we absolutely need from Adobe - lets
prioritize them and ask for them or build them ourselves. If there is more
to come, lets create a branch for Adobe that is pristine to commit to and
figure out what to do with those commits when they happen instead of
holding up community progress waiting for them.

Please understand I mean absolutely no disrespect to folks like Alex, Carol
and others who are clearly doing everything in their power to make this
transition happen quickly and I appreciate their tremendous effort and
constant communication on where things are and where they're going. I
believe them to be an invaluable resource and hope that they are made
available to the community for a long, long time.

I've seen numerous component and layout proposals get raised up only to
wither and die because nobody knows what to do with them. Where do they
belong? Even if they belong in the "main" SDK, they can't because that is
locked down for Adobe's commit. Do we need four different color pickers?
Probably not, but its better than zero. If they don't belong in the "main"
SDK then where? I don't feel like "something else" is a valid answer here.
Where should they be committed? Its been over a month since the initial
framework commit and nothing has changed. Is that a valid representation of
the discussion and effort put forth by the community?

I'm perfectly willing to admit that I may be completely out to lunch on my
reasoning or I may just be tilting at windmills here. It certainly wouldn't
be the first time I missed the big picture and will *gladly* retract
everything I've said if and when I'm proven wrong.

That said, can we please come to some actionable decisions and, you know,
act on them?

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message