Ovum forecast that 8 out of 10 new phones shipping ‘today’ are Java-enabled – this estimate has been around since 2005 and may apply today according to Sun’s website. Informa’s 50% forecast comes from their MAPOS report I believe. The ratio of CE device to mobile device models comes from discussions with Adobe – but I recall seeing this echoed somewhere publically too (probably in terms of Flash-certified device models). But the extrapolation of device models ratio to a device volumes ratio is a *very* crude one.

BTW, read your post and enjoyed it – good work pulling together the stats.

Andress

Do you have links for some of the stats you give in the penetration section? In particular the Ovum / Informa %ages of devices with Java (I’m inclined to think its much nearer 80% than 50%, given that we’ve had over 2bn Java devices shipped already – double Adobe’s target for 2010) and the ratio of Flash handsets to consumer electronics? I was struggling to track some of these figures down for a post I did a while back, http://blog.masabi.com/2007/07/mobile-software-primer.html

It’s a nightmare to get any kind of comparable penetration stats for platforms, it looks like I actually overestimated the Flash numbers if that 2:1 ratio is correct (they only list one consumer electronics device on the Adobe site using Flash Lite so I assumed all devices were handsets).

Interesting post though!
Tom

>1. Lack of MVC model in Flash: very good point. I would
> say that I would also expect Flash 3.x to include screen-
>based primitives (as opposed to vector-based primitives)
> for more easily creating applications in the view of the
> MVC.
In fact it is more than adding screen based primitives : it is really pushing a state controller, a notion of clear data binding to uncorrelate what is displayed from its storage/representation. And yes it is something where Flex is good at, flash alone with its frame based structure won’t be able to scale up: doing many screens in flash to do a whole UI will always be long and difficult (as opposed to Flex/TAT/DigitalAirways/OpenLazslo/Silverlight/JavaFX/…and DHTML )

>2. Interesting note about middleware services. Anything
> specific you have in mind ?
In fact the achille heel of those “interpreted solutions” (flash+actionscript/svg+Javascript/java) is in their access to native capabilities of the platform: how to map the C/C++ or even Java world in th eaction script world (or javascript) to expose relevant middleware to the applications: typically an application shouldn’t be coded only in flash but some part may have to be C/C++ to do the tricky stuffs 99% of the usefull apps or services are in fact requiring. The achille heel of java are JSRs, for flash it is their standard APIs that will always restrict you too much (not sure about flash lite 2, but addressbook with specificities, sim application toolkits, middle telephony and many others may miss).

>3. System wide view. Couldn’t Adobe/OEMs sandbox the
> system-wide view, i.e. restrict the access to environment
> variables ?
Env variable? really not enough what do you do if you want to change th ecall state machine or how the sound is shared between two sound clients (ex: call is in the native world, mp3 player in the OEM sandbox: how to resolve the conflicts?…and more than that comply with the conflict handling specification, different for each phones?)

>4. Bluestreak et al: I would say these companies are doing
>business on the back of strong applications (e.g. ESG),
>but not on the strength of their platform..

Hum, Bluestreak is for example strongly pushing its mobile APIs, but I agree for now they focus highly on very vertical services…

good discussion

I have to find the time to make a recap of the UI technologies and companies around there.

Thomas

Responding to your points:
1. Lack of MVC model in Flash: very good point. I would say that I would also expect Flash 3.x to include screen-based primitives (as opposed to vector-based primitives) for more easily creating applications in the view of the MVC.
2. Interesting note about middleware services. Anything specific you have in mind ?
3. System wide view. Couldn’t Adobe/OEMs sandbox the system-wide view, i.e. restrict the access to environment variables ?
4. Bluestreak et al: I would say these companies are doing business on the back of strong applications (e.g. ESG), but not on the strength of their platform..

Andreas

as you mentionned flash may be use for two purposes:
+as a ui technology like in the prada phone but in that case you need a loy of c code to do the real work
+as an application framework…but really we are far far from it.
basically flash is lacking:
-some high level app dev ui paradigms (strong mvc) that flex is just beginning to give on pc
-some high level middleware services that air (new name for appolo) will bring on pc…and will have to be strongly extended for mobile (say hello to jsr )
-and a big one:no system wide view…and due to this last one be sure that the big 5 oem won’t completely standardize on flash; because to have a whole system view you have to master your execution environment and adobe won’t allow that i think.

anyway for now flash is a nice ui techno and yes adobe will try definitively to push it as a full dev framework but it is not the case today nor tomorrow and it is why companies like bluestreak are doing business today: because the adobe real strengths are their tools and dev community, not their client techo nor their knowled of mobility.

as a remark : i’ve read that flash lite 3 will use the actimagine technology….

Good question. Macromedia, prior to its December 2005 acquisition by Adobe was a $500 million-a-year company, i.e. about one fourth of Adobe’s size. Flash Lite brings in less than 2% of Adobe’s revenues, so it’s a small revenue source for both Adobe’s and Macromedia’s size. So if Adobe is not making much money from Flash Lite, why is it not making it free to OEMs ?

Firstly, Flash Lite royalties are doubling year-on-year, whereas Adobe revenue is increasing at a rate of 15% annually (see here for Flash Lite facts and figures ). Assuming this growth is unchanged, Flash Lite revenues might represent as much as 10% of Adobe revenues in 3 years. This is respectable revenue source that Adobe might not want to let go of.

Secondly, if Flash Lite was free (as is the Flash runtime for the desktop), Adobe would not be able to charge for Apollo mobile; I doubt the incremental features brought by Apollo mobile would be valuable to OEMs. So there would be a limited potential for revenue through Apollo mobile.

Let’s see what Adobe gains from giving away Flash Lite for free (note that this is different from open sourcing Apollo, which I do not think Adobe will do, as it will break the consistency – or relative consistency – of implementation across Flash Lite installations that Adobe is so concerned about).

Firstly, I can’t see why OEMs would not integrate Flash Lite if it’s free, especially now that the industry is beginning to see FL as a de facto standard.

Secondly, OEMs would get ‘free’ developer support via Adobe (ok, they have to pay for it or outsource it to Adobe), but it’s essentially one less product planning factor to worry about. This is particularly important for LG and Samsung who have very little developer support, if any.

Perhaps OEMs who have a longer-term software strategy (primarily Nokia and Motorola, and secondly SEMC) would resist the total penetration of Flash Lite and continue a dual-supplier strategy between Java and Flash Lite, but even then Flash Lite would grow to compete with Java in terms of penetration (50/50).

Conclusively, I see many reasons for Adobe to eventually give Flash Lite away for free, by reducing the royalties over time, so as to maintain the revenue stream as long as possible. The challenge for Adobe would then be to incentivise OEMs to buy Adobe tools for application development and content authoring, to make up for the lost revenue stream.

- Andreas

Andreas at Vision Mobile has an interesting article… it mentions FLASH LITE as a ghost platform. ‚ÄúAnd while Flash is a tactical solution, it is establishing itself as strategic platform within a large and diverse range of handset models. This is what I would call a ghost platform. It reminds me of

Interesting assessment on the future of FlashLite

Nokia must U-turn on its Symbian strategy > Rethinking application environments > The retail environment as a point of service discovery > Read my mind: adding recommendation to the mobile search mix >Flash-Inside (R)> Bye Bye Browser > Customised design manufacturers are here > Get real: Focus on Voice not Data > The Brand Vector > Behind the scenes of the SIM industry >