JD on EP

One part of the presentation suggests that "make flash open" may not be achievable... or maybe not.

It's the line that says "what open is not: No Single-Vendor "Lock-In"". (Ironically, I can't link directly to that part of the presentation -- the application state is not exposed in the URL. ;-)

At first I read that as "no single-vendor engines can ever be called 'open'", but then I wondered whether single-vendor engines would no longer be "not open" (and therefore objectionable) if there was no "lock-in", which would then invite the speaker's current definition of "lock-in".

Recap: When people say "I won't use Flash technologies because they're not open", then I'm really trying to find what they actually need, what specific changes can be made, in order to avoid the label, and avoid the objection.

Maybe it was the title that got me started off wrong: "The Open Web and Its Adversaries". This seems to set up an opposition... I'm not certain who he thinks is an adversary to something else, nor why.

Part of me wonders whether he's saying something like "There is no room on the Web for contributions by a single group which is not the universal group". Not sure.

Hey John,

The issue of single-vendor lock-in is simple: if you can't either switch vendors, or else fork a copy of an open-source Flash Player, and continue to support your presentations in the face of incompatible or just different (performance trade-offs) behavior of future Flash Players, or even just because you want to do your own custom hacks to the Player's open source codebase and make your content users run a custom Player build you provide, then it ain't "open".

I know, this sounds crazy to you. Why wouldn't everyone want a single version that Adobe controls? Why wouldn't everyone want all the docs and hand-holding marshalled in favor of the latest supported single-version? Well, you never know. But "open" isn't about what you think is best -- it's about what other users think is best, possibly wrongly, and choose to do in opposition to your central control system.

Suppose there's a better Flash/HTML integration than Apollo, and people want to use it. Say it requires forking Tamarin and some openswf.org or gnash workalike. Say it becomes quite popular on cutting edge sites. Would Adobe's customers be in hot water legally for using such a thing? Would Adobe do something other than send in the lawyers?

A less extreme case: various hackers try Apollo for a while, along with Ajax and WPF/E. They are not ready to "lock in" to any one system, but the budget for hedging by trying all three is running out. At the end of the day, these hackers have to make a choice. Do they favor something controlled by one vendor, which may or may not meet their needs? Or do they try something controlled by a set of extensions on top of evolving, user-innovation-directed browser functionality?

It's not obvious that most content hackers are better off picking Adobe or Microsoft in this scenario. There are subtle short- and long-term tradeoffs.

But it's clear what "Open" means: no single-vendor command and control -- no "lock-in" to a single-latest-version runtime. More IE compatibility shimming; more Firefox and Safari first, then IE porting, development. Or perhaps using a code generator such as GWT or Laszlo, which handls the cross-browser portability hassles but involves less lock-in than the for-$$$ tooled approaches from Adobe and Microsoft.

This is a cost, but it's not obviously so high that buying into the single-version utopia from Adobe or Microsoft wins.

/be

[jd sez: So, does that boil down to "if adobe doesnt publish sourcecode to player for others to rewrite and rebrand as they wish then it is "not open" and is the adversary?" If not, then what's the summary instead?]

Something like that, yeah.

Sun finally is open-sourcing Java. Years late and megabucks short; it may still help.

Flash is nearly ubiquitous, and pretty popular; yet it ain't all that when it comes to text, a11y, focus handling in browsers, and 3D. Sure, Adobe can improve those things in version N+1 or N+2. So can browsers improve their vector graphics and video. In the latter case, at least some browsers use open source, and the standards for graphics and video are open -- public, not encumbered by patents, subject to revision by a community process of some kind.

Don't get me wrong: Adobe can be responsive to its community, and command loyalty to its products. But open is a well-defined state whereby you can substitute or fork. That's not the case with Flash. Do you see my point?

/be

[jd sez: Is anything which does not follow your "open" development model the adversary, or is it just something different, to be taken on its own merits?]

It's not the "development model" that's "open", it is the set of content standards. Safari is not all open source, but you can substitute Safari for Firefox to browse the web and run "rich Ajax web apps"[*] if something bugs you about Firefox. IE7 and Opera are closed source, but if you want to substitute either for Firefox, you can (modulo add-ons of course). Substitutability is independent of open vs. closed source in general.

My use of "adversary" is meant to identify single-vendor lock-in. In the context of Microsoft as well as Adobe fielding proprietary runtimes against the browser, two adversaries competing against each other could accelerate Closed Web evolution -- or one could drive the other toward opening standards and possibly even open source.

But open source is a development model, and I was writing about content formats, production and consumption models, and distributed innovation. Open source fosters user innovation networks, as Eric von Hippel's research shows, but it is not required.

/be

[*] Safari, or rather Webkit, still is not supported at first launch by a surprising number of "Ajax" apps, possibly because of web compatibility bugs -- but it seems to be getting more support over time.