Flash Wars Continue

In the wake of newly-discovered malware vulnerabilities in the hoary old codebase of Adobe's Flash player, Adobe offers a Flash update aimed to plug the most recently-discovered leaks. These bugs are part of why Apple declined to allow Flash plugins on iOS devices: they make insecure everything they touch. The tag Flashwar chronicles the long road of Adobe's unsuccessful effort to prove Apple made a mistake in failing to include Flash in iOS devices.

Unfortunately, I still run into sites that want to deliver video using Flash rather than H.264. Smarten up, folks. We need this junk off our desktops and laptops, too.

'Chrome' Exploit Windows-Only? Another Flash bug?

The headlining Pwn2Own crack suggests Chrome was cracked within a few hours of the contest's open. But was Chrome cracked at all? Just as purported Safari cracks historically included what amounted to cracks of Adobe Flash, it seems this new crack is a Windows 7 exploit, functional on 32-bit and 64-bit Windows 7 with Flash installed, but not otherwise accessible.

The real culprit may be the incomplete sandboxing of Flash under current versions of Chrome.

Adobe Flash Takes Another Hit

Microsoft, Google, and Netflix have joined forces to request a standard for encrypted streaming of HTML5 video. The proposal isn't itself a standard yet, but it moves users away from the need for proprietary plug-ins for managing keys to ensure access to copyrighted content is limited by rights owners without a need for nonstandard software such as Microsoft Silverlight (used by Netflix) or Adobe Flash (used by Google's movie service).

Opening the field to any compliant developer would eliminate the lock-in of proprietary format vendors and, hopefully, increase the options for consumers to select player software based on factors like usability and performance.

Apple, which is not a co-author of the proposal, was an early mover in rejecting proprietary plug-ins on mobile devices due to security and performance issues. Adobe's effort to maintain its market for plug-ins on the desktop (after apparently losing the war on mobile devices) is likely doomed by a widely-supported standard for providing DRM-governed video.

End Of The Mobile War Over Flash Player

The long-running account of the jabs and counter-punches exchanged between Apple and Adobe over Flash Player and its suitability for mobile devices seems to have wrapped up at last. ZD Net first reported that Adobe's Flash player will not be developed for new mobile devices, mobile browsers, or mobile OSes or even their future OS versions.

All Things D notes that Adobe recently announced a 750-FTE job cut, which may be related to its lack of stomach for a platform war that, as chronicled right here (from its inception or shortly afterward through a number of pieces of evidence suggesting the ultimate outcome), Adobe was plainly losing.

Will anybody miss mobile Flash?

Will anybody see a reason to maintain desktop Flash? If it eats cycles (i.e., electricity), weakens security, is unnecessary (or mobile devices wouldn't work), and messes up accessibility on users' diverse platforms, what will the basis be for deploying content that requires Flash players on any platform?

Adobe to deliver cross-platform Flash!

After fighting with Apple over who needed whom, Adobe – after a long and drawn out battle – has created a drag-and-drop converter for developers to turn their useless-on-a-standards-based-device Flash content into something that will run anywhere HTML5 is spoken.

Th irony, of course, is that Adobe has claimed for so long that Flash is "cross-platform development software" and necessary to "browse the whole web" and that these facts were the reason everyone should just shell out for Adobe tools and deploy to Flash. Well, well. How the world turns.

The upside for Adobe is that if its tool makes high-quality HTML5, it will have a much bigger audience for its wares than were willing to sacrifice standards. The upside for everybody else will be the avoidance of Adobe's atrocious reliability and various performance and security problems and overall quality associated with Adobe's Flash player plug-in.

(Although I include it above, I repeat the link to Daring Fireball's discussion of Adobe, Apple, and Flash because it's such an outstanding summary for those of you who may not have the time to read more than one link.)

Flash: Bad For Your Battery?

Apple has removed Flash from the default configuration of the new MacBook Air, and no wonder: adding Flash reduces batter life by two hours if you have the wrong web page open – apparently, even if the offending page is out of sight in the background.

There have been enough posts here about the Apple/Adobe catfight over Flash and its relevance to the world that I wonder if Flash needs its own tag ....

Still, it's clear why Apple would want people to have to opt in to Flash instead of getting it without informed consent.

Jobs Right on Flash Quality After All?

The "it plays Flash!" feature of non-Apple smartphones may not be all it's cracked up to be. When users find that Flash loads, they may find the experience to be as technology and telecom reporter Ryan Lawler describes it for for NewTeeVee.com.

And how was that experience?

"Shockingly bad."

The site links a video of a 1GHz Nexus One smartphone on a 25Mbps broadband connection, which should be ideal for showcasing just how wrong Steve Jobs was about the potential of Flash on low-power mobile devices. Instead, the Nexus One video showcases just how accurate his assessment of Adobe's product continues to be: the UI seems to support mis-clicks and links to the wrong content; the waits are significant even when trying to download content on a fast broadband connection; unexpected errors occur and advise trying to experience the pain later; and when you finally find a site that does not result in an error (in his case, a show episode on Fox), there is not even a small segment of smooth video.

Kevin Teufel's assessment:
"That's seconds per frame, not frame per second. . . . I don't know about you, but I could not watch a full episode like that."

The little flag that appears atop your "video" is not encouraging: this video not optimized for mobile. Well, wasn't that what Adobe promised? Provide content everywhere and experience the "full web"? The whole time the video demo is underway, it's crystal clear that the experience one gets isn't the "whole web".

Seeking content from MetaCafé yielded a Hulu video, but Hulu is blocked on Flash mobile devices. MetaCafé, which was more reliable than other sites in yielding working videos, also had a video from an action movie, and the subjective video experience was so bad (the combination of skipped frames and other factors) that the things that were supposed to look quick looked spliced-in, and things that were supposed to be slo-mo looked choppy.

I don't claim 3G video is great on iPhones, but the video above demos performance on a WiFi network attached to a 25Mbps broadband connection. It's not 3G, it's the best case for the Flash/Nexus One combo.

Bleh.

HP To Face Lenovo's ARM/Flash Problems?

If the "Hurricane" tablet HP is rumored to launch this summer is really an ARM/WebOS number, how would it deliver Flash support in the face of Lenovo's ARM/Flash issues?

Hypothesis: HP will declare Flash unfit for non-Intel machines and use Apple as an illustration that non-shipment of ARM should be accepted by the market. Else, HP has Adobe working on an ARM version that is ostensibly going to be ready by HP's ship date, and is being paid to promote Flash as "the real web". It's not likely HP really believes Flash is a competitive advantage over the long run – even current Flash supporters think Adobe's Flash is not the future of the web, but is a next-18-months phenomenon.

Might HP think Flash is an advantage over the next few quarters? Since content sources haven't prepared HP-tablet-specific apps for delivering streaming content yet, HP would have to rely on HTML5 migration if it didn't deliver an acceptably-working Flash plug-in.

CBS Joins Flash Snub

Unlike ABC, which uses a custom app to provide its web site's video content to Apple's iPad users, CBS will provide all its content by delivering an HTML5 web site by next season– apparently cutting Flash out of its future video tool chain.

Web standards über alles?

Maybe Adobe's plan to offer the best HTML5 tools is the best defense of its content developer tools after all.

Adobe Flash Delays ARM Notebooks

Adobe's delayed release schedule for its Flash plug-in targeting Lenovo's new ARM-based mini-notebook is causing delays in Lenovo's release of ARM-powered notebook hardware, according to ARM's marketing VP Ian Drew.

Apple's decision not to allow Adobe to sit between users and content on Apple's ultraportable platforms seems quite rational. Imagine Apple waiting on Adobe to launch a phone based on a custom Apple chip for which Adobe doesn't prioritize optimization. Hmm. I'll just stick to imagining the phone with the Apple chip.

To think about Apple products suffering from third parties not bothering to dedicate optimization resources is s0 1996.

UPDATE: Opera's Philip Grønvold has weighed in on the HTML v. Flash debate, saying that "Flash as a video container makes very little sense for CPU, WiFi battery usage et cetera – you can cook an egg on [devices] once you start running Flash on them and there's a reason for that." Opera will continue to support Flash, "But at Opera we say that the future of the web is open web standards and Flash is not an open web standards technology." So it's good news for Lenovo: in a couple of years, Lenovo will be able to sell whatever ARM-based or other alternative hardware it wants, without worrying whether Adobe sees fit to optimize Flash adequately for users. In the meantime, though, it's nice to be Apple ....

Scribd to go HTML5, Abandon Flash

Adobe's Flash has taken another beating at the hands of standards: online document sharing site Scribd is going HTML5. Scribd's co-founder and Chief Technological Officer could not make the company's position more clear that the switch is based on technical and not merely political reasoning: "We are scrapping three years of Flash development and betting the company on HTML5 because we believe HTML5 is a dramatically better reading experience than Flash. Now any document can become a web page." (For clarity in parsing those links, the Jared Friedman quote is from Schonfeld's article at TechCrunch.)

Scribd joins Google (YouTube HTML5 beta here), Apple, Microsoft (which moved from merely sending video to iPhones in H.264 to declaring "The future of the web is HTML5"), Virgin America, and a variety of other content providers in leveraging new standards to provide a full-featured alternative to historically buggy and insecure proprietary plug-ins like Flash.

Apple is still winning the Flash chicken.

Flash In The Pan

The escalating Apple vs Adobe exchange on Flash (vs migration to open standards like HTML5) and the adequacy of open-standards alternatives has another data point:

[I]n the past four quarters, the H.264 format went from 31 percent of all videos to 66 percent, and is now the largest format by far. Meanwhile, Flash [encodings] represent only 26 percent of all videos. That is down from a combined [Flash encoding] total of 69 percent four quarters ago. So the native Flash codecs and H.264 have completely flipped in terms of market share
Erick Schonfield of TechCrunch, quoted by Philip Elmer-DeWitt in "Apple vs. Adobe: Is Flash Dying?"

Given what Apple has demonstrated is possible using open standards and modern browsers, the need for proprietary plug-ins to deliver content or interfaces to customers or other users seems solidly on the decline.

Apple Spells Out Its Position on Flash

In the tradition of his prior letter on DRM (before it died), Steve Jobs posted an open letter on Flash.

The upshot? The analysis of Apple's view here (and here, here, and here) has been spot-on: Flash is a buggy kludge designed to force people to use Adobe products and tends chiefly to decrease performance while increasing bugs. Flash has no place on Apple's mobile OS. Apple wants people to use Apple's APIs to leverage Apple's technological improvements, and Adobe's track record of exposing Apple's technology to developers through intermediary platforms isn't stellar. Adobe took ten years to move to Cocoa in its own products, and has a long record of churning out buggy, low-performance Mac software. Apple has no interest in forcing users to suffer through third-party software that is at least that bad.

Atop all this, rewriting apps to function with no mouseover won't happen, and this absence will prevent Flash apps from working as expected. Broken, unworking applications or parts of web pages are just not part of the environment Apple wants its users to experience on its platforms. This is completly justifiable. Adobe's argument that its proprietary plug-ins should be executed on Apple platforms in the name of "openness" are absurd, and should be loudly derided.

Apple is preventing users from suffering bugs, plain and simple. Apple could act differently, but its decision is entirely understandable. Adobe is justifiably disappointed (not just because it won't reach the market on Apple's mobile platforms, but because the shortcomings and lack of necessity of its Flash products have been made clearer to the world as a result of Adobe's conflict over accessing the Apple mobile platforms). At the end of the day, though, Apple has made its position clear and we'll see how users and developers react.

UPDATE: Adobe's already reacted, in the form of an interview given by its CEO Shantanu Narayen.

In the interview, Narayen claims Adobe is persecuted by Apple for promoting open content. This would, of course, be more plausible if anything about Adobe's Flash platform could be construed as "open" in any way. It's exactly as closed and proprietary as Win32 or any other proprietary API, and utterly unlike the HTML5, Javascript, CSS, H.264, and other standards Apple suggests developers consider as Flash alternatives.

Narayen denied that its Flash plug-ins are the main cause of application crashes on Macs, but supplied no data for the claim. Narayen blamed Apple for the crashes. Sigh.

Narayen contradicted Jobs' statements about Flash's impact on battery life, but – alas – offered no evidence. One solid piece of evidence might be a Flash application running on an iPhone using an Adobe cross-platform toolkit, showing exactly how Flash impacts (or doesn't impact) battery life. The power to put on the demo is entirely within Adobe's power, because the iPhoneOS developer tools enable developers to load apps onto iPhones using synch cables. Adobe could load a browser with and without Flash plug-ins, and show a bake-off with time-lapse cameras, showing the battery death. Narayen could prove Apple wrong in a heartbeat. But ... no. A cynical person might be tempted to doubt his statements about Flash performance.

Narayen's assetion that users should decide which apps should be allowed to run on the iPhone is the closest thing he comes to articulating a useful philosophical perspective. Narayen's position is not dissimilar to lots of similar articulations one can find on the blogosphere, and no more fleshed-out or well-explained. Of course, the fact is that the power to decide what Apple will and won't carry in its store is Apple's. Since the qualitative claims about his company's products seem so full of fluff, one has a hard time working up sympathy for the view Narayen's company has been wronged.

Let's face it: Narayen's claim against Jobs' position is a shoe cut to fit Narayen himself: "It doesn't benefit Apple, and that's why you see this reaction." Ha. Apple's position doesn't benefit Adobe, which is why we see Narayen's reaction. As for Apple's position, it makes sense that having browser users – and iPhone users are browser users – hit site after site that requires mouseover to work the interface ... well, with a UI that has no pointer to effect a mouseover effect, it's clear Apple can't make it work for users. Developers would have to make iPhone-specific Flash content. Adobe's vaunted cross-platform benefit would evaporate as developers were forced to make target-specific versions of Flash apps. Apple's point becomes clear: so long as they're rewriting anyway, why not rewrite to a standard that anyone can implement, instead of to some proprietary API that might or might not ever expose the advantageous features of the platform to developers for the enjoyment of users?

Narayen's statement that the future lies with multi-platform content is likely correct. The problem with Narayen's position is that multi-platform content will likely be delivered through standards – programming interfaces capable of being implemented independently by different vendors competing to deliver a superior experience, better performance, and improved security – and this future is unlikely to require the services of proprietary API vendors like Adobe to allow commonly-demanded functionality like exposing video or accepting clicks or the command to scroll. Narayen's "multi-platform" prognistication may be where he wins the battle, but it's surely where he loses the war.

Maybe Adobe will come back with a better articulation of why Flash doesn't suck on the iPhone, but this wasn't it.

WebKit To Improve Security, Performance

The next-gen WebKit will be WebKit2. Like Apache and OpenSSH before it, WebKit will take on separation of user-facing processes and back-end logic to enhance the project with the delivery of privilege separation, sandboxing, multiple CPU saturation, and other performance and security goodness. Unlike Google's Chrome, which uses WebKit within per-page processes separated to protect the application from a single crashing window and prevent cross-page exploits but depends on Chrome for management of all the processes needed to provide the benefits of separating the application into processes, WebKit2 will offer process management for rendering so that any application wanting to leverage it can do so without unnecessary internal complexity or boilerplate insertion of code for process management.

The specific benefits of WebKit2 will be seen in time, but from the WebKit wiki page announcing WebKit2, one can access beta-version build instructions for testing purposes.

One benefit of WebKit2 will be the isolation of misbehaving instances of Adobe Flash from web pages on which no such misbehavior is manifest, so that losing a page to a Flash issue doesn't cause users to lose the state of their thirty other open windows with half-filled forms and half-read source articles. The plug-in will take down only the process to which it is plugged. Another benefit? When one closes a window and kills the associated processes, the killed processes' resources are immediately available to the system, and don't contribute to the bloated resource consumption of long-running browser applications. This will continue to be a major benefit so long as proprietary plug-ins remain a major feature of the Web.

The browser competition can only help users: the better the performance, the happier they will be on more modest hardware, and the longer their current computational power will satisfy their needs. And who knows? Perhaps Apple's recruitment of Mozilla's former security chief signals impending improvements to the robustness of Apple software in the face of malicious code.

iPhone OS 4.0 and the future of iPad

iPad Sales
I was at a Best Buy today replacing a failed Bluetooth headset, and I noticed iPads on display. On display – but not for sale. After selling 300,000 iPads on launch day, and 450,000 iPads through the iPhone OS 4.0 announcement, Apple hasn't yet caught up with Best Buy's demand. It hasn't yet been a week.

In that not-quite-a-week, Apple reported selling 600,000 iBooks (what Apple is apparently calling the books sold through its e-bookstore) and transmitted 3.5 million applications. Even if those apps are free apps, Apple is happy: the iPhone OS 4.0 offers an advertising API so developers can have Apple supply what the developers want, which is ad content and payment. Apple will deliver all the ads in HTML5. Apple will take 40% of that revenue for sourcing the ads and transmitting them to the freeware users. (Presumably, paid versions will offer ad-free use; if the ad platform is a success, however, "free" apps supported by ads may make good enough revenue for developers that they thrive in the enormous and growing community of iPhone OS users.) There are currently 185,000 apps in the App Store. 3,500 are iPad-specific apps. Apple's ad platform will not take users out of the app, just expose them to the ad and allow them to interact with the ad – if need be, it seems, to play a game, learn about a close-by restaurant, or buy another app.

What prayer has a potential competitor to approach Apple's mobile platform anytime soon?

Looking at mobile browser use for an indication of users' actual demand from mobile devices do do something besides make phone calls, one finds Apple's mobile platforms accounting for 64% of the entire mobile browser use. This means that all competitors combined barely crack half Apple's share. Looking back at Apple's ad-supported app revenue, one quickly realizes that Apple sells the platform with the eyeballs. As a developer, Apple provides the platform with the user demand.

Example? Pandora adds 30,000 users a day on iPhone. Per day. One platform.

What will iPhone OS 4.0 offer?
Other than a new ad system likely to change the way developers finance "free" applications, there are several features Apple addressed in the announcement.

For developers, new APIs are offered: an acceleration API for accessing hardware-accelerated math functions (useful all the time), and a set of services applications can access while they are in the background (useful when you want output from an application in the background). This means for users that applications written for iPhone OS 4.0 will be able to offer services – like positioning data, or music (or other audio like turn-by-turn directions), or ongoing two-way voice communication, or push notifications, or a message when some task has been completed – while the application is in the background. This, for users' purpose, is multitasking.

For Apple's purpose, it's the ability to offer what people want without entire applications – including their user interfaces, for example, which cannot be accessed while in the background – hogging system resources on a constrained system with only 256MB RAM and a single general-purpose processor core. In other words, the reason Apple didn't launch iPhone 2.0 (the first iPhone OS with third-party application support) with multiple simultaneous applications (the reason being fear of lazy developers running whole apps, including their graphics, in the background) has been solved in "allowing multitasking" on the iPhone.

Applications that want to access system resources while suspended to the background can do it through the new services APIs, and applications that don't need resources while in the background won't ask for them and won't get them. The apparently new paradigm of suspending applications while in the background will enable applications to maintain their exact state between users' access, which is both a win from the standpoint that opening the Address Book will potentially give me the open contact, scrolled to the part of the note with the stuff I was reading, rather than giving me part of the contact list (as it does now). This is good. It will also make users feel that applications they access after switching apps were still running and make them feel the whole thing is running like on a multitasking desktop – because from the user's observation, there will be no difference.

Users will get progress indicators for background uploads of photos and all kinds of useful stuff Apple will support in the services API.

Apple will also support blueooth keyboards, geotagging, system-wide spell-checking, and organizing apps into folders (e.g., a games folder, a folder for apps related to photos, etc.; useful if you need more than the 180 apps you could fit in Apple's old scheme, or just want to be able to find the apps faster).

Since iPhone OS has over fifty thousand games, Apple has also integrated into iPhone OS 4.0 an API with which developers can plug their users into a social gaming network – to access information about what games people are playing/recommending, who has the high score, and so on. This social network aspect of the new API could become very interesting to see develop as a way to multiply the value of users by causing them to reinforce successful applications with recommendations, sales, and competition (that will cause more app use, that may drive more ad revenue for Apple and for developers ... I imagine a racing application in which you pass billboards advertising things game users might want to buy, and each time the developer gets a fraction of a cent from the brief impresion ....). Very interesting.

The synopsis of the iPhone OS 4.0 presentation: "So let's review. Multitasking, folders, enhanced mail, iBooks, enterprise, Game Kit, and iAd. And these are just seven of the 100 new user features. Now we are releasing a dev preview today."

iPhone OS will run on all the currently-shipping platforms, and on iPod Touch 2nd gen and the iPhone 3G. To get all the iPhone OS 4.0 features, one needs a phone that's a year old or younger: iPhone 3GS or iPod Touch 3rd gen. The platform launches for the iPhone this summer, and for the iPad this fall.

The iPhone OS v.4.0 will not support Flash or Java. (Apparently, even through cross-compilers, though there's some speculation the ban is based on their impact on multitasking rather than being a result of pure philosophical opposition.)

The Future ...
iPhone has gone from "not a business tool" to being deployed in 80% of the Fortune 500. Most of the professionals I know now seem to be toting iPhones. Apple's iPad has been embraced by one New York hospital reported to me, which is configuring users' Citrix apps to join its LAN remotely and enter prescriptions for patients without having to get to a fax machine. It'll be psychiatry from the couch, on an iPad, while your explicit app runs in the background. Longtime computer-industry geeks are using Apple products as their main portable computers, and Apple's platform puts it in a position to flexibly build machines out of any parts it thinks will give it the best bang for the buck, thus solidifying its competitive position by allowing it to pick the best parts for whatever application Apple chooses to target.

Apple owns the catbird seat.

Flash Chicken: iPad Update

For those of you just tuning in, Apple has been playing chicken with Adobe's Flash by expecting the world to use something else to accommodate users of Apple's mobile products. Adobe has responded to Apple's criticism that Adobe is a lazy developer that ships bug-balls that breed security problems by explaining that Steve Jobs doesn't understand Flash and that everybody really needs Flash and can't get along without it, and Adobe looks forward to Apple coming around and allowing the Flash plug-in Adobe has developed.

So far, Apple has been winning this game of chicken. Content providers and Flash competitors eager to enjoy the competitive advantage of reaching Apple's mobile customers have taken steps to allow content access without Flash. The latest of these is ABC, whose online video content can be accessed via an iPad app. The ironic twist? ABC advertises its iPad app in a Flash banner.

Presumably, Apple expects open-standards solutions like HTML5 to pose a Flash replacement, as nobody is going to deploy an Apple-only solution. Rich media tools are important to the web, but support for novel interfaces isn't trivial and can't be bolted onto platforms like Flash that assume moving pointers and other elements that don't exist on the iPad and would require users to interact with Flash content completely differently than they do with everything else on their handheld unit.

iPad: Battery Exceeds Expectations

Both the Wall Street Journal's Walt Mossberg and BoingBoing's Xeni Jardin agree that Apple's battery claims hold up to real world demands -- and then some. The BoingBoing review notes that graphics hardware acceleration available to native iPad applications provides a performance one doesn't see on the web: "This ain't Flash video over WiFi, folks. You'll feel sad going back to chokey http embeds." Mossberg noted that he used his own notebooks perhaps 20% of the normal amount of time while reviewing the iPad, mostly when he had to edit a long document or view Flash (which Apple is not going to support on its mobile platforms).

Although Mossberg credited Apple's initial battery life at 15% over Apple's claims, BoingBoing went further:

Manic, nonstop use revealed a number of things: battery life is better than I anticipated. I got a full day of constant internet-connected use (it did not leave my hands) on one charge. More than 12 hours, with heavy video and gaming, and screen cranked up to full brightness.

Jardin extols iPad content by likening it to a Harry Potter book -- a responsive source with moving pictures, magical, allowing a content experience involving personal tactile control, not seen elsewhere. If Mossberg says the device answers 80% of his portable computing needs, this thing may be a plausible notebook replacement for some users, and if not a notebook killer, maybe close enough for Apple's purposes.

Time's writer Fry concludes his article with what is likely the most heartfelt endorsement:

It is possible that the public will not fall on the iPad, as I did, like lions on an antelope. Perhaps they will find the apps and the iBooks too expensive. Maybe they will wait for more fully featured later models. But for me, my iPad is like a gun lobbyist's rifle: the only way you will take it from me is to prise it from my cold, dead hands.
The iPad Launch, by Stephen Fry

Flash Chicken: Apple Winning

Remember the iPhone/iPad-related squabble between Apple and Adobe on support for Flash?

(Jaded Consumer offers some opinion here and here on Apple's nonsupport of Flash on new platforms.)

The next chapter, after Microsoft's decision to support iPhone by sending it H.264 instead of expecting iPhones to speak Microsoft-ese, is here. Virgin America has dropped Adobe's Flash to court iPhone users. Virgin's experience was interesting: its site used Flash for things like transitions between advertisements, which could be adequately handled without the heavy load of Flash. Logically, Virgin dumped Flash because it didn't need the headache and it didn't want its customers to have to endure the wait, especially on thin-pipe clients like roaming mobile devices. Apple isn't moving to support proprietary plug-ins, content providers are moving to drop the crazy expectation that client platforms will install proprietary plug-ins to support the content they want users to get hassle-free. By standing with standards, Apple seems to be winning the Flash chicken game by waiting for the world to take advantage of improving standards.

Given that most folks' use of Flash seems to be to show video, HTML5 could lead lots of sites down Virgin's path toward avoiding the headaches caused by Flash.

RIP Flash?

Apple's pogrom against CPU-hogging proprietary plug-ins, announced with the iPhone's nonsupport of Adobe's Flash, continues with the iPad and its "screaming fast" A4 processor.

Perhaps Apple's plan to push developers toward standards, and away from competing content-creation suites, calls for mobile devices to drag developers into a standards-based future. And drag them, it might. Microsoft changed its back-end tools to send MPEG-4 rather than proprietary content to iPhones, for example.

On the other hand, many sites are simply horribly broken on the iPhone: they sniff that you use a mobile platform, redirect or rewrite all URL requests to a URL that users aren't trying to reach, like the URL of the domain's mobile-version home page (not a mobile-version of the linked page, a mobile version of the home page of an entire news site like the Washington Post), and thereby break every single link made to any page in the entire domain. The fact these dufus site developers can't bother to give an already-HTML site to a mobile device suggests that revising a site to do it properly is simply beyond their pathetic powers.

So, what's the real reason to halt flash?

Easy: Flash sucks. Flash is the number one cause of crashes in browsers on MacOS X. (Heck, it crashes browsers on Ubuntu Linux and on MS-Windows, too.) And not only browsers: according to Apple's Bertrand Serlet, "plug ins" (read: Flash) constitute the number one cause of all application crashes on MacOS X. Flash is also implicated in security problems on every platform for which Flash is released (and spawned third-party Flash security tools). And the number one use of Flash – delivery of video – doesn't even require Flash; HTML 5 supports plugin-free delivery of much of the content, like streaming video, that plug-ins like Flash were designed to support. Use of Flash to create full-featured applications is certainly possible, but it also sucks: it doesn't behave like any users expect, and for example won't support cut-and-paste with any other applications and will be totally unaware of the spellchecker that all your MacOS X applications inherit from upstream Cocoa objects. Worse, unless Flash sites are carefully designed, making useful links to content deep within Flash sites can be impossible for regular users. And we've seen how much effort most site designers seem to be willing to spend making a site usable by a customer's browser. Got accessbility aids? Unless your Flash site designer knows to sniff for them, Flash will simply defeat all the accessibility aids that your computer offers all your non-Flash applications.

Don't expect Apple to support Flash on mobile devices Apple holds out as examples of innovative and high-quality user interface. Flash won't be designed for multitouch or any other technology Apple may choose to offer.