Years ago, there was all this geek drool behind this newfangled (well, kinda) UI concept called “Resolution Independence”. In oversimplified terms, it’s the notion that the size of visual elements, such as text and graphics displayed on computer screens, could be freely increased or decreased without any detriment to clarity (i.e. no fuzziness or blockiness). For more info: http://en.wikipedia.org/wiki/Resolution_independence. This topic comes up every once in a while between clients and developers.
“RI” was supposed to be the next step in graphics that would make UI much more accessible for the visually impaired, would take maximum advantage of higher resolution displays and it would also find all of your missing socks. So… what happened?
Both Mac OS X and Vista have supported it since a few versions ago and have seen a couple apps and demos here and there. There are nuances of it in the latest versions of Flash (things like 9-slice scaling). No real adoption though, at least beyond the OS level and some of their bundled applications. There is also a smattering of apps out there that have made the leap. As a whole though, at least what I’ve seen, it’s still a little quirky and quite inconsistent. All of the pieces are there, just slow adoption.
Here’s some contemplations and realities:
1. Production (or reproduction) of tens or hundreds of icons, UI slices and backgrounds at 2 to 6 times the original size is time-consuming and challenging work with little payback in many cases. The ability to up-sample a UI for legibility or a tad more “slickness” under special conditions? Maybe not enough of a payoff for the flood of additional hours, not to mention the size bloat and possible performance issues.
2. There really hasn’t been a huge leap in display resolution, with the exception of perhaps some laptop displays and mobile devices. Mainstream displays (at least for desktop) need to get over the stank of ~72-100 ppi and get way into the 200+ ppi zone (approaching low-end print res). Imagine cramming all of the pixels of a 30″+ display into a 22″. Goodbye to antialiasing-tweaking issues, sub-pixel mash (hello “snap to pixels”, hehe) and simply beautiful, print-like imagery and text on the screen. I know this is wishful thinking and that shift doesn’t appear to be on the horizon soon. Way too expensive unless there’s some sort of breakthrough. Overall, maybe its a chicken/egg thing? Kidding.
Q: Why don’t designers just develop all UI elements and iconography as vector art? That way it’s infinitely scalable and fluid.
A: Try drawing modern UI elements, such as photorealistic textures, believable lighting, smooth gradients and shadow details, with a drawing toolkit that features points, lines, beziers and fills. Not practical unless you are one of the 10 people on the planet who can do stuff like this with vectors.
Sure, SGI Irix had scalable vector icons, and they looked like clip art for a reason. Want to see a single modern icon constructed with vectors? It’s about 3 megs and just about takes down your machine to render it to screen.
Try opening it up in an image viewer and zoom in and out while panning around — not very fluid. This is another reason why ginormous bitmaps are created instead of the pure vector route for now.
To be continued… (standards, pipelines and RI for the web)
The main issue is fonts. The user should be able to set a global minimum font size for every character that ever appears on the screen, whether on the desktop or in applications. Perfect rendering of icons and logos is not important. Being able to read text is what matters. Font artists create a range of sizes for each font they design (in bitmap). OS and app writers just have to use this work. I can set minimum font size in my browser, so why can’t I do it everywhere else?
I like to sit back from the screen. Others have declining eyesight. It is too bad some lobby (for the visually impaired) didn’t make a big fuss about this years ago. If they had, this problem would have been fixed.
T - 8/13/09 2:38 PM
I completely agree with your global minimum font size comment. Type clarity is much more important than rendering of logos and icons, if you were to place all of your UI elements across a spectrum of importance. Text is the very basic essence of UI communication and user training. But I have to disagree with your comment that “Perfect rendering of icons and logos is not important.”
What I was trying to convey is that the outlining framework of the UI, which can include icons, logos, symbols, buttons, and perhaps varying panels and textures across the interface (glass/metal/plastic/tints/etc.) also factor into the overall user experience.
If your application’s icons, logos, textures, and other graphical UI building blocks fall apart as the text remains clear and true, you quickly cheapen the experience of your application. It’s acceptable these days for web UI – but for a desktop app? We should expect higher standards.
Lets say in a hypothetical UI, there could be a predominant use of icons, symbols, and other visual cues presented to the user to indicate functions, with little textual reinforcement (maybe in order to simplify visual flow). If the user is presented with a mixture of sharp textural cues against blurry graphical cues, how can the user not place emphasis on one or the other? Its all about visual continuity and we’re just settling for the shortcomings of hardware and software at the moment. (I kinda touch on that in the other part of my not yet posted rant
)
Your other comment: “Font artists create a range of sizes for each font they design (in bitmap).” Could you expand on that, I’m not sure what you mean.
craig_nanaumi - 8/13/09 10:56 PM