Adobe To Offer Ad-Supported, Web-Based Version Of Photoshop
from the free-as-in-beer dept
There's been a lot of talk about traditional software companies shifting their model to emphasize software as a service instead of traditional installed applications. Companies like Microsoft have talked about delivering ad-supported hosted versions of their software, driven largely by the popularity of Google's services that use such a model, though it's starting to push subscription-based services for enterprises as well. Still, for all the talk, there hasn't been a lot of movement in this space when it comes to consumer applications, though enterprise vendors such as Salesforce.com have delivered their software over the web -- with quite a bit of success -- for some time. However, Adobe now says it will deliver a free, web-based version of Photoshop supported by advertising. The company says the hosted version will serve as an entry-level version of its Photoshop and Photoshop Elements software, and not have all the features of those applications -- but it won't have their prices, either. The company says it realizes that there's a large group of potential customers that aren't interested in paying to buy its software, so the ad-supported model gives it a way to earn some revenue from them, as well as a marketing tool that could convince them to purchase one of the installable versions. One would imagine there are some kinks that will have to be worked out, such as the hosted version's ability to deal with large photo files, but the business model overall seems pretty sound, and should continue to find favor with other software companies.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Reader Comments
Subscribe: RSS
View by: Time | Thread
I am Jack
[ link to this | view in chronology ]
Ablowbe
[ link to this | view in chronology ]
Re: Ablowbe
Seriously, I suspect that this would be useful for the same kind of people who would buy Elements or PaintShopPro (I own both), not professionals who would (and should) buy the full-blown package.
Although Adobe often has a dismal record with creating easy to use software, this could be a very exciting development and another long overdue nail in Microsoft's monopoly's coffin.
I got nothing against MS making software, some of their stuff is good, but they should compete fairly with other companies, and the lengths they go to in order to avoid that shows that even they realize they can't do so well competing on quality.
[ link to this | view in chronology ]
Remote Services
Local software installations do still have their advantages.
[ link to this | view in chronology ]
have they studied this at all?
There's no real issue with the GUI or the bandwidth, some nicely coded AJAX would do the trick and with fat enough pipes you can shuffle images back and forth between client and server, do zooming, cropping and basic editing.
The problem is with the serverside DSP. Image processing is very expensive in CPU cycles. Even after you have transferred a local copy to the server each transformation is intensive. If you want the fancy capabilities of Photoshop, which at heart are matrix manipulations, z-transforms, linear algebra and so on you will only get about 5-10 client connections on a 3.8GHz box. It doesn't help if you use clusters or distributed processes or any other technique to spread the load, you still have a uniformly high cycle cost and commensurate real cost per operation. You can throttle back the ops and schedule like an old school JCL, but then you get a very sluggish system which artists and experienced uses will just laugh at.
There's no way on Earth you can build any sort of economically viable model on this. It's a money pit.
Things like image processing, sound and video editing are why we have powerful clients.
Things that work well for remote processing are mostly text and numerical applications with a low processing overhead per transaction and storage/database intensive backend that adds value the user wouldn't have locally.
But I suppose these clowns just want to jump on the latest trendy bandwaggon without thinking it through. Some PHB at Adobe had a brain fart during a strokey-beard thinktank committee and decided that this was a "cool". Perhaps I can sell them my study? :) But to be honest it's common sense to anybody with half an iota of technical nouce.
[ link to this | view in chronology ]
Re: have they studied this at all?
[ link to this | view in chronology ]
Re: have they studied this at all?
There's no way on Earth you can build any sort of economically viable model on this. It's a money pit.
Funny, they said the same thing about indexing the web too. Good thing the search engine pioneers didn't bother listening.
But to be honest it's common sense to anybody with half an iota of technical nouce.
Right... You just keep on telling yourself that.
[ link to this | view in chronology ]
Re: Re: have they studied this at all?
Did "they"? Show me a link to any statement by a credible commentator who said that.
And while you're at it let's see your calculations for a decompression, anti-aliased resize, rotate, and re-encoding of a 600x800 JPEG image (the most common workflow for a typical user who edits photographs). Please take into account memory usage, DMA thoughput etc. when showing your working. Turn that into a per transaction cost analysis for your own design of a hardware architecture capable of serving 100,000 requests per day and demonstrate any revenue model that can pay for those boxes in the server farm. If you like do some experiments to back up your claims. You may use libmagick, gandalf, libdevil, mng1,libmorph libraries or use the gimp toolkit. Don't forget your Apache+squid frontend to your process management system capable of serving a peak of several hundred simultaneous requests.
The we'll see if you've got Jack, or if you just know Jack.
Hmm?
[ link to this | view in chronology ]
Re: Re: Re: have they studied this at all?
Oh, just in case you were waiting...
No, I won't be responding to your troll.
Go take into account your own limited mindset.
[ link to this | view in chronology ]
Re: Re: Re: Re: have they studied this at all?
This is Jack crying a river.
This is Jack being a dooshbag.
[ link to this | view in chronology ]
Re: Re: Re: have they studied this at all?
Its like saying Trend Micro's housecall antivirus wont work cause you have to upload the entire contents of your hard drive to the server and boy that server will only be able to handle 1 or two people at once.
Or you download a drm'd app with an ad server built in and do all that crap on the client. sheesh. Why do it the hard way? Especially when you just bought macromedia. Flash anyone?
[ link to this | view in chronology ]
Re: have they studied this at all?
I fully agree with you re the limitations of both CPU bandwidth and network bandwidth. A graphics app does take a lot of run time and transferring potentially huge images over the internet to have a bit of work done on them then returned seems non-sensical and slow
However I guess it depends on how modular you make the app, how much is done server side and how much is done client side
I've given this about 5min thought so am fully prepared (and half expecting) to be shot down in flames, but I'd make a basic container to hold the picture in and links to tools (so the basic client object would donwload quickly) then have any big tools seperate modules to download when required
So the first time you used say the 'sunburst' effect it takes a little longer but if its still in the temporary internet files next time...
Face it - there are a lot of graphics controls the average person never uses
Have the whole thing kind of neutered so it won't work at all unless the online site gets to serve adverts (like the old dongles) and Bobs your uncle
Sure there will be people who can hack out these controls but I would suggest that these are the people who will just download a CS2 torrent anyway
And I think there is a market I have [ah-hem] friends who have copies of software because they use them once in a blue moon, and don't want to shell out loads just to crop a photo occasionally (if push came to shove they'd probably just stick to MS-Paint). At least this way the average person like this who isn't going to buy your product anyway will earn you some revenue
The REALLY interesting bit for me will be if they manage to get context specific ads based on whats in the picture.... ;0)
[ link to this | view in chronology ]
Re: have they studied this at all?
Yes probably only good for basic editing at first, a few steps above Paint for people who don't often need or use Photo software or would rather use familiar Photoshop tools instead of learning some bulky junk image-editing app that came with their printer.
[ link to this | view in chronology ]
Re: have they studied this at all?
However, what they SHOULD do is to create a smaller, much simpler local version of photoshop with banner ads, and release it for free. Even make parts of it open source, and integrate it with existing plug-ins, allowing users to get an idea for what the photoshop user experience would be, and get the community involved.
This way, they can monetize some of the users that either were unable to use photoshop in the past, or illegally downloaded it (not to mention that it's never bad to have a bunch of young people/students/tomorrow's knowledge workers trained on your software.
Of course, some people might release a mod to remove the banner ads- but anyone that smart/motivated would have pirated the software anyway, so that's really not a problem. To deter this, you'd just have to patch, patch, patch.
-Max
check out my blog
[ link to this | view in chronology ]
uh
[ link to this | view in chronology ]
re: uh
[ link to this | view in chronology ]
The Cutdown CS2 is already here....
Paint.NET is free, has layers, can open PSDs with a free contributed plugin, looks pretty...
What more could people want if they aren't in a job that requires the full CS2 - which means they should be able to buy it...
[ link to this | view in chronology ]
CS not on my shopping list
then... until CS. Then I parted ways with Adobe.
I don't like the software activation scheme and
web apps are even less appealing. They'll get no
more money from me.
[ link to this | view in chronology ]
CS not on my shopping list
then... until CS. Then I parted ways with Adobe.
I don't like the software activation scheme and
web apps are even less appealing. They'll get no
more money from me.
[ link to this | view in chronology ]
Two Words
[ link to this | view in chronology ]
Where?
I see that this is a web offering, but maybe adobe's just going to bundle up a client that runs all transforms in in your browser? Server might just store "saved" photos for location independence and serve up applets and ads.
Wouldn't that be the smart way to handle this? This isn't a move to replace photoshop with a web-app, it's a move to keep potenital new users from going elsewhere by providing (and controlling) access to an entry level package.
[ link to this | view in chronology ]
Re: Where?
But, let's entertain your idea for a moment and see what advantages it might offer to all parties. A uniform and updateable interface is one possibility, but if it has to send modular code blocks to the client when a function is updated you have the whole problem of platform compatability (even with Java) and security that comes with the traditional code distro issue.
The system I was asked to evauate was serverside, so that's where my assumption lies, maybe Adobe have a different take on it, but I don't really see how it offers any advantage to cripple a clientside app by making it require internet connectivity, That's the sort of twisted thinking you would expect from clowns like Microsoft.
[ link to this | view in chronology ]
Re: Re: Where?
if you're talking over the heads of most people on techdirt, chances are good that you're pretentious.
[ link to this | view in chronology ]
ignorance does not behoove you well
The proposal has a technical flaw. How am I to communicate that without referrence to the technical details and anecdoatal eveidence of similar systems I have had experience with?
Simply making an unqualified comment that "its a dumb idea" wouldn't be very helpful or rigorous would it?
Perhaps you might be encouraged to explore some of the terminology I use, do a little research and enlighten yourself .
This site is called "Techdirt". It is a forum for exploring technology. I'm sorry that technology is complicated and involves a lot of jargon. I also find jargon which is outside my domain very confusing and occasionaly rant at some of the legal and financial topics discussed here. But at least I listen, learn, ask questions and do research to fill in the gaps in my knowledge. I find it challenging and educational to explore the ideas and jargon of other people, it's worth making the effort.
It's a bit like us being on a forum for brain surgeons and you saying, "All this talk about about the hippocampus and neurons has me lost,"
And @ Jack, you made the troll with your asinine and glib insults, I've merely given you the oppotunity and challenge to put your money where you mouth is. I'll take your lack of substantial response for what it is.
I'm well aware of my very limited mindset. I consider that a virtue. Unlike you I'm well educated enough to be under no illusion I know everything. The more you know the more you realise how little you know, whareas you've only shown me hubris and arrogance so far.
Guys, if you want to engage in a conversation about technology at least come to the discussion well informed or with the humility to listen and learn.
[ link to this | view in chronology ]
Re: ignorance does not behoove you well
You made many assumptions that since you failed miserably to come up with a working model in the past, that it is obvious that noone else could possibly come up with a working model.
THAT is ignorance.
[ link to this | view in chronology ]
This is Jack shutting up and looking for a different thread to pretend to matter in.
[ link to this | view in chronology ]
jack needs a bean
[ link to this | view in chronology ]
what more can I say?
Admittedly:
i) this is only ONE persons analysis.
ii) its a "back on an envelope" sketch based on a few limited heuristics
iii) I've never actually *built* such a system (but many of comparable complexity)
So. Jack, I've not "failed" to make such a system work. Moreover, you have not succeeded, not even in understanding the issues involved.
@kneeL , last year I worked on an OpenMosix cluster with 64 nodes, so if you think I'm "afraid of using processing power" you need to revise that. Old and bitter yes, I'll grant you that :)
Perhaps Adobe can make this work. Let's see in 12 months and you can taunt me with "told you so".
Perhaps, as is common, the author has partly misunderstood what the proposal is. I'm working to the assumption that it's a serverside image processing system for mass use. As Enrico and Neal have both pointed out maybe that assumption is completely in error and we are all misunderstanding what the software really will attempt to do.
And I hope Adobe can make it work since it would be a useful tool, perhaps even one that I will be using at some point, it it works.
As it stands the proposal sucks because it will always be more economical to use local processing power and most people have more local cycles at their disposal than the server can offer.
Meanwhile my beef is really that I've put forward a thoughtful and reasoned analysis and got very little in reponse but mocking. It would have been nice if we could have stayed on topic and come up with a few insightful counterarguments. I would have liked that because I could have called up the developer I spoke to in December and said "Hey, have you ever considered... maybe we could re-evaluate this".
I think the distinction between "cautious and sceptical realism" and "mocking cynicism" is what is now on trial here. And Jack, Andy and kneeL have basically proved that the lowest common denominator of thought, the shouting empty taunts of the mob, count as "argument" in this forum.
[ link to this | view in chronology ]
Re: what more can I say?
[ link to this | view in chronology ]
Nice speculation.
You have no idea how Adobe is going to accomplish this. They have some of the best minds in the world on staff. Since they haven't posted any technical facts about the system, everything posted here is pure conjecture.
I'm no programmer, but even I can think of solutions here.
How about using the clients pc to process info? We use a java app like this that works on every machine we've tested.
How about working on a proxy image? All the editing takes place on a very low resolution copy of your file until you finalize at which point the large image either takes some time or is emailed to you when it's ready.
Speculate more noobs.
[ link to this | view in chronology ]
I am Jack
I'll just let your owns words do the talking...
But I suppose these clowns just want to jump on the latest trendy bandwaggon without thinking it through. Some PHB at Adobe had a brain fart during a strokey-beard thinktank committee and decided that this was a "cool".
Right, really thougtful and reasoned there.
And Jack, Andy and kneeL have basically proved that the lowest common denominator of thought, the shouting empty taunts of the mob, count as "argument" in this forum.
You managed that all by yourself.
But to be honest it's common sense to anybody with half an iota of technical nouce.
Since you seem to be insulting my depth of technical knowledge, while at the same time begging me to solve the problem you failed miserably to solve, here's a little clue on a good place to start.
If the limiting factor is as you declared it, is CPU power, then where can we get loads of cpu power from? I would give big blue a call and ask for pricing and scalability options on their utility-computing services. CPU power is not a limiting factor... Your imagination is. Hence my original statement that you simply have blinders on.
[ link to this | view in chronology ]
Probably still won't work on a Mac
[ link to this | view in chronology ]
jack still doesnt know jack... but he likes nuttyb
@nuttybar... processing a lower resolution proxy image is not a viable solution to processing power issue. why you ask? because when the user decides to zoom into a part of the image a new zoom croped proxy image will have to be uploaded and reprocessed by the server to match whatever changes were made to the original. your ideas dont hold water in real life situations.
[ link to this | view in chronology ]
Compatibility of AMD with Photoshop Elements
[ link to this | view in chronology ]