I think this is going to be the Photoshop where people <i>seriously</i> start considering a web-based alternative. Not because CS6 will be poor software, but because I think the division among users who use Photoshop for photo editing/professional photography/restoration/stationary/brochure design and those who use it for designing web sites and applications is getting greater. The latter needing less of the options available in Photoshop today, let alone CS6. I say this with over 10 years of experience as a designer/developer – Photoshop needs a Basecamp simplicity-inspired derivative. I think as we move to more institutional, common & predictable UI (if we do), we'll find that we're doing very simple tasks and that those tasks can be taken from Photoshop and automated for the purpose of web/app design and execution. I know there are a few alternatives to Photoshop already, but of all the ones I've seen (yes, even Pixelmator) none of them really go for <i>just the web industry</i>. I think the market for such a tool (hopefully web-based) would be pretty great at this point – and given the audience – using state-of-the-art browser API wouldn't be much of an issue.<p>Long ago, Adobe had this product called "Photoshop Server" or something similar. I believe it was based off of Macromedia's Generator (which generated server-side Flash content). The concept was neat - place variables within a PSD, upload it and allow the server to render out based on what you wanted to show. The problem with this, of course, was that it was proprietary, heavy, expensive and overall pretty crappy. It would be really interesting to see someone create a web-based web design/development application that interacted with some OSS library that most hosting companies would allow (like gd, ning, etc), read .json and interpret data within the image.<p>Probably a bad idea, but how big would a .json-based image be compared to the average jpg? Pretty big, I'll bet. But I wonder with gzip compression how manageable could it be? Would be really cool to define information (text inside of the image, accessibility, SEO capabilities) within a .json-based image. No license required and it would be readable by any human or software.<p><pre><code> {
"dimensions": { "width": 3, "height": 3 },
"data": [
["rgba(255,255,255,1)","rgba(255,255,255,1)","rgba(255,255,255,1)"],
["rgba(255,255,255,1)","rgba(255,255,255,1)","rgba(255,255,255,1)"],
["rgba(255,255,255,1)","rgba(255,255,255,1)","rgba(255,255,255,1)"]
]
}
</code></pre>
Yep. That would be huge. I suppose that is a massive digression of the topic but this would be my ideal image format, outside of svg/canvas of course.