Comments

Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts

1. I think it's entirely feasible that Adobe web apps could be as powerful, and probably exponentially more powerful, than a desktop app. Think folding@home and the total compute power a relatively small number of PS3s have contributed to the cause.

   http://fah-web.stanford.edu/cgi-bin/main.py?qtype=osstats

   Enough said.

   Posted by: dave | October 18, 2007 11:39 AM





2. 
   I think it's probably feasible, but it is not necessarily desirable. In my opinion, the main benefit of web apps is that they are accessible on any machine, everywhere. A bunch of apps benefit from that - email, simple writing/editing, etc.

   However, sometimes you need the interface that a specific physical computing provides you to be productive - I for example, don't write code on my iPhone, nor would I want to use Photoshop. The benefit in that case of a web app is minimal (which is not to say the app can't connect to the internet, that's a whole different matter)

   Thinking about it, this is actually why people want native apps on the iPhone - you don't necessarily need the access-from-any-device functionality of a web-only app as you carry the device with you all the time.

   Posted by: Einar Vollset | October 18, 2007 11:42 AM





3. @Dave: I admit I don't know much about the technical aspects of cloud computing, but I think with Folding@Home you have a ton of computers contributing to the same task... that's not the same as a bunch of people simultaneously executing different complex tasks.

   Also, I think it is less about the back end power to execute a filter or render a video or rotate a vector image, etc... and more about the pipe coming into your home being able to handle it. I just don't know if the broadband infrastructure in the US will be up to handling that in 10 years -- I'm not holding my breath. (Maybe I'm being too pessimistic.)

   Posted by: Josh Catone | October 18, 2007 11:48 AM





4. While it seems like a daunting task when considering the infrastructure and technical challenges, it certainly does not seem impossible when one considers the state of web infrastructure and technology in October of '97...

   Posted by: Benjamin DiGregorio | October 18, 2007 12:32 PM





5. Having your apps "online" does not have to mean that all the processing must be done within today's browser/flash technology, or that all of the processing happens on the server. "Online" can come to mean "the application launches quickly from wherever I am in the world and the data is not tied to one computer".

   The web browser and flash are one way to get to that version of "online". Alternatively, launching a future version of Photoshop (e.g.) from my desktop or a web site (won't matter) may take these steps:

   • check if I "own" a copy
   • verify that the software is on my system, then downloading or updating any parts that aren't up-to-data
   • let me run the software with my full system processing power
   • load or cache all my data from the web cloud
   • clean-up, if necessary, when I'm done
   if all those steps happen, and usually are very very very quick, then that to me counts as an "online app". I hope this seamless form of online app is what Adobe is shooting for, because then we won't be hobbled by any browser/script/banwidth limitations.

   Posted by: Brent Noorda | October 18, 2007 12:42 PM





6. The high-level processing necessary to do a lot of the things that Adobe's popular offline apps do -- managing multiple layers, calculating the math behind vector graphics, applying complex filters -- these are all things that are complex, and I would guess are beyond the capabilities of Flex or Flash.

   The generality of Actionscript is such that all of these things can be done in Flash (with sufficient effort) right now. Being able to calculate (for example) where a bezier curve should lie given its control points is just math, and well within Flash's capabilities. It's a different trick to be able to constantly recalculate and display that path as a user drags a control point around. Flash's limits are more likely in the area of real-time feedback during user manipulation than they are the simple act of calculating the end result.

   Posted by: Jpfed | October 18, 2007 1:09 PM





7. Feasible? Yes, I think so. Desirable? Not so much. I'd much rather Adobe, having nearly gotten to the point where the core layering/filtering capabilities stretch across their suite with similar capability, finish that task and then update it, heavily. The architecture is long in the tooth, IMO, and rather than chase the trend of the moment, I'd much rather they innovate.

   And by innovate I don't mean adding another marketing bullet point, half fleshed feature.

   Posted by: Brian | October 18, 2007 1:48 PM





8. Feasable, yes. Desireable, yes.

   Feasable because, well it just is. Already, VMs such as the JVM have demonstrated that you can write code that executes at *very* close to native compiled speed. While flash's AVM2 is not quite there, it is very fast. One more iteration and I think it will be in Java league. More importantly, as you point out, Hydra gives you low level access to the graphics processor. This means that hardware acceleration will be accessible to flash apps. Once this is mainstream, game over.

   Desirable, yes, because there is no benefit to apps being tied to one specific computer if we can put it in the cloud and the performance is equivalent. Whether its gmail, or photoshop I would much prefer to be able to work on my laptop and then go to my desktop and have everything accessible. But beyond just universal accessibility, it also means upgrades are automatic, and it makes collaboration *much* easier.

   It is fascinating to read people suggesting that these kinds of apps would not benefit from cloud computing. The general lack of ability of people to imagine beyond their current circumstance never ceases to amaze me.

   Posted by: Hank Williams | October 18, 2007 3:20 PM





9. I was just working in an After Effects project that (with source material) was 93 Gigs. Have fun with that one on the web.

   Posted by: Austin | October 18, 2007 6:50 PM





10. Benefits to the end user -- well maybe not too many. You could build in the desirable cloud features in to the desktop application and everyone would get the same benefits.

    But how about benefiting Adobe? Well lets see.

    • Moving to something like Flex/Flash (once it becomes powerful enough), would mean saving a lot of cost to port the apps to various platforms, since the web platforms are inherently cross platform.
    • Web software is nearly impossible to pirate.
    • Patches are delivered instantly without distibution issues.
    • ActionScript doesn't have the same support as C++/Java/.net and so Adobe will have to build the libraries themselves. Their platform dev tools will get pretty good when they finally port all their apps. This means more AIR/Flex Builder sales.

    Now is this feasible? Sure! Quad processors, multiple GPUs are already here-- all you need is to build this scalable support in to Flash/AIR and everything is possible. Hell, you can do layers and effects today in Flash! Check out:

    http://www.fauxto.com

    Posted by: Gautam | October 19, 2007 1:02 AM





11. If they have a 10-year goal it's probably feasible - given that they control the Flash player, it's a case of taking it in a direction where it supports all the 'hard bits' of CS3 - lest we forget, much of non-performance intensive aspects of Photoshop are already written in LUA (a scripting language) and CS3 uses an Adobe custom framework for x-platform Windows/PC development already.

    On one level that is completely different from Actionscript / Flash / Flex, but over a 10 year period it would be possible to do - first push towards using a custom version of AIR to replace the existing framework / LUA, then tackle the remaining intensive parts - mostly A/V processing.

    You can see the early steps with Flex - i.e. the latest version caching frameworks / components and only downloading new modules.

    A final step will be supporting PhotoShop and AfterEffects plug-ins as downloads.

    I'd guess the goal on this would be moving CS toward being a subscription service, rather than the current purchase/upgrade cycle model.

    Posted by: JulesLt | October 19, 2007 3:00 AM





12. I think it is feasible and reasonable. Moreover, it might help Adobe reduce the cost for offline application product.

    I hope to meet many Adobe application on online as soon as possible.

    Posted by: my2pac | October 19, 2007 8:03 AM





13. Besides the (possible) technical issues, this would be a great step forward. Indesign in a collaborative webbased environment ... when can I sign up? This would have a huge impact on the premedia/prepress field. Instead of relying on a more or less closed business this would force these companies to be more open and collaborative. Better start re-thinking strategies and business models.

    Posted by: Peter van Teeseling | October 20, 2007 9:46 AM





14. I believe it's totally feasible. noone knows what will come in the next few years.

    have you seen last week's preview of flash 10, developers will be able to include c and c++ code in flash? they were able to run Quake 1 in flash. it doesn't take much more to make a video processing app for flash. for me it's just a matter of time, and which startup or hacker will first going to do it.

    Posted by: heri | October 22, 2007 7:33 AM





15. I am not sure if we always need to talk in terms of "online apps" and "desktop apps". Flash is a runtime, it runs on the desktop. There is no reason one can't beef it up to run any complex application. Is it just how you initiate it?

    I echo most comments on the post. It's time we take a break from such questions - "Is it feasible for Adobe to have its entire catalog of software online in 10 years?".

    They seems very short sighted.

    Posted by: Ardor | October 22, 2007 11:18 PM





16. I'm with the other commenters. Something important is missing from the post -- Web technologies are going to improve in 10 years. The brave CEO-talk may be puffery anyway, but we shouldn't ignore the possibility that it's real.

    In 10 years we may not be writing things in slow, incompatible JavaScript, thank God. Look at what Microsoft's doing with Silverlight 1.1 and the CoreCLR, and consider that Adobe's in a good place to introduce its own improvements, since so many people have Flash.

    Posted by: RF | October 29, 2007 9:43 AM





17. Adobe CEO Bruce Chizen is a buffoon.

    First his dumbass comment about Adobe users have clients with deep pockets and are willing to pay more for his software; now this.

    I can only hope 10 years from now that jagoff will be a distant memory from the halls of Adobe.

    Posted by: andrew | October 29, 2007 10:27 AM





18. I agree with Ardor, this is just marketing bullshit from a big corporation (Adobe). Lots of software companies already deliver complex software over the web (including Adobe) - its called "downloading and running it".

    The process could be improved, e.g. by having a standardised web-software sandbox (a la Java) or a trusted "installation" process, but the innovation and progress progress is going to be made by the OS developers, the browser developers and by improvements in networking infrastructure (i.e. massive bandwidth) - and not by (albeit influential) graphics software companies like Adobe.

    Posted by: sasserstyl | October 30, 2007 7:35 AM