Is Photoshop for Chrome OS going to kill GIMP?
There are two interesting points that seem to prey on people's minds today: is Project Photoshop Streaming for Chrome OS the first step to Photoshop on Linux? Will it, therefore, threaten GIMP in its own habitat? All good questions, let's dive into this.
Executive summary aka TL;DR
Don't panic. In fact, never panic.
Why did Adobe go for Chrome OS?
The first thing we need to understand is that being a de-facto monopolist in the graphic design market means that so far Adobe can do pretty much as they please. Here's some meaty data to confirm this. Back when we wrote on the Creative Cloud in May 2013, the latest information was that Adobe ended Q1 2013 with 479 thousand paid Creative Cloud members. And now here's an excerpt from their latest financial report:
Adobe exited Q2  with 2 million 308 thousand paid Creative Cloud subscriptions, an increase of 464 thousand when compared to the number of subscriptions as of the end of Q1 fiscal year 2014.
As you can see, Adobe increased Creative Cloud subscriptions by ca. 1.4 million users over a year despite all sorts of complaints about forcing users to give up and embrace the SaaS model.
Nothing is forever, of course, but it does look like Adobe is going to keep its market dominance for years to come.
So why Chrome OS? Basically, it's one of the ways Adobe is trying to stay relevant as a software vendor. Here's an interesting bit from NPD Group's report from July 2014:
Since the start of the year (January – May), Chromebook sales within the U.S. Commercial Channel increased 250 percent year-over-year and accounted for 35 percent of all channel notebooks sales. According to The NPD Group Distributor and Reseller Weekly Tracking Services, total notebook sales through the U.S. Commercial Channel increased 36 percent, desktop sales jumped 24 percent, and overall PC client volume rose by 1 million units so far this year. Windows notebook sales were flat and Macbook sales increased more than 20 percent.
Not too bad for a product that only started shipping in 2011. And then there's the whole topic of the Google Chrome browser that does have a major piece of the market's pie too.
Is this the beginning of Photoshop for Linux?
It seems to be one of the most common assumptions about this news so far: Chrome OS is Linux, ergo Photoshop for Chrome OS must be Photoshop for Linux. It also seems that Adobe's FAQ on Project Photoshop Streaming hasn't reached the masses yet, because it gives a pretty good idea:
Project Photoshop Streaming is identical to the Photoshop you’d install locally..., however, instead of being installed on your local machine, it is running in a virtualized environment so can be accessed from any Chrome browser or Chromebook.
In other words, Photoshop wasn't ported for the web or to Linux. It was "merely" adjusted to run in a virtualized environment and save all your project data to Google Drive. Moreover, Adobe has no immediate plans to replace desktop apps with virtualized apps:
Creative Cloud products will continue to be available as local download and install.
While this doesn't exclude the possibility of CC apps becoming true web apps in the future per se, this is simply not happening just yet. Neither is there any indication that Adobe is interested in porting Photoshop (or any other CC apps) to Linux.
Will this move threaten GIMP et al.?
Now that is a very sensible question. Here are some important considerations.
Pricing. Adobe is rolling out Project Photoshop Streaming to educational institutions first (North America only for now), with other target groups to follow. This makes a lot of sense: if you want loyal customers, start growing them early on.
Ideally, that's where free/libre applications are supposed to have an advance over proprietary ones even with academic licensing in place (Adobe provides an up to 60% discount for teachers and students on available subscription plans). But the topic of extrapolating skills acquired with GIMP to Photoshop, as well as being able to survive in a Adobe-centric (or, rather, PSD-centric) company is a huge can of nastily looking worms. Students expect to land a nice job, after all, what with student loans looming over them.
Performance and color fidelity. Since we are talking about a native app running in a virtualized environment, it looks like we are talking about VNC-like communication where you basically interact with JPEG screenshots of the actual user interface that a server keeps sending to you. That is, you get lags (digital painting, anyone?) and, potentially, JPEG compression artifacts. Providing reliable color management over network would be another tough task to accomplish. It's going to be interesting to see how Adobe will tackle this.
Vendor lock-in for file storage. Within the Streaming project, Photoshop was customized to save everything to Google Drive, which is only to be expected with Chrome OS. That's what you get for a $199 laptop that works as a thin client only.
Project data storage. Even with events like recent bump of available disk space at Dropbox up to 1TB for $10 a month, handling large bitmap project data from Photoshop is going to be an issue when it comes to cloud-based storages.
Not exactly a novel solution. In late 2013, rollApp released GIMP, Inkscape, and more free software as Chrome apps. This solution turned out to be a double-edged sword. There is no innovation there, as the company brings no added value to customers except forcing them to use cloud storage (and that's a questionable benefit). However customers appear to like what they get: the vast majority of reviews of both virtualized GIMP and Inkscape is positive despite of the aforementioned JPEG compression artifacts, interaction lags, and lack of updates (both GIMP and Inkscape provided by rollApp are badly out of date).
Summing it up
Don't panic about the threat for GIMP: there's still time, and Adobe most likely doesn't care all that much. And don't get too excited about Photoshop on Chrome OS either: there's much to be done yet.
With introduction of the Project Photoshop Streaming, Adobe is placing one vendor lock-in over another, which, admittedly, many will happily embrace, while others will stick to either Photoshop CC or earlier versions.
Should you adopt it, you are likely to deal with network latency, connectivity issues, and insufficient color fidelity as well. You might live happily with those deficiencies, you might be dragged off by GNU activists, kicking and screaming, or you might stay indifferent and keep using existing software. Either way, polishing this will take some time, and whatever free apps follow this route, they will have to deal with the exactly same issues.
It's rather obvious that web-based apps are getting smarter and more capable. In fullness of time, they might wipe out the software ecosystem as we know it. Which is why it's great to see free/libre projects such as Gravit and Metapolator emerging.
One breaking point for making this free/libre cloud-based software sustainable will be solving the data storage issue. Today, locking users to a particular cloud storage of data is practically the core of SaaS businesses, and it's not easy to devise a revenue model that would involve both free software and federated or self-hosted storages. It's entirely possible that in a future SaaS-dominated world, libre end-user software might stay marginalized.
It is, however, worth thinking about (especially if you are of radical FSF persuasion), how the entire community could grow reality check muscles and start figuring out, if there's a way to stay relevant, when SaaS is a dominant model. Because we'll live to see the day. In fact, it's about 8:30pm the day before.