And plus, it wouldn't have to be changed on resolution, and it could also have several presets pre-rendered, like .5, 1, 2, etc. That would take up what, ten megs?
And plus, it wouldn't have to be changed on resolution, and it could also have several presets pre-rendered, like .5, 1, 2, etc. That would take up what, ten megs?
You still need multiple images though for different sizes of the same interface element.
At the moment, images stored on disk at multiple resolutions and the OS picks the closest one and matches up with the display resolution and desired size. For example, in Finder, pick a window with Preview PDF icons in it then select view options. Watch the icon as you scale it past 32 pixels closely and you'll see the image and text change completely. That's because the 128px icon scaled down to toolbar sized 32px becomes illegible so they switch in a different version with larger text on it at 32px.
The problem with using vectors, apart from the GPU speed, is that an icon reduced to 16x16 quartz pixels, no matter the actual display resolution may be too small to recognise so icon designers produce icons at different sizes emphasising the important elements in an icon at each size. If you just relied on computers, the vector based image will gain and lose elements depending on the scaling or anti-alias itself into a fuzzy mush. Try opening a drawing in Illustrator and scaling it up and down and watching lines appear and disappear.
That's why one vector icon won't do. Humans can tell if an icon looks right much more easily than computers so you may as well let a designer do it for you and store them on disk already. It saves the computer having to do it later.
Apple have been gradually pushing up the sizes of the stored bitmaps ready for high resolution displays. Bearing in mind the above, I think it's a better option than vectors.
I don't think Apple would build its own global mapping service. Why do all of that when the work is done already. They would likely use Google Earth. The same way they use the Google tool bar in Safari, PDF in Preview, and AIM in iChat.
I don't see the need for it to be integrated into the OS. It only needs to be an application that can integrate among other OS applications with a better interface than Google Earth.
What if the UI *was* vector based, but each time you set the resolution it then pre-renders the UI elements as appropriate for the resolution.
Sort of on-the-fly caching so that you have the speed with pre-rendered bitmaps, but then you have the flexibility of having the vector elements as "templates" -- saves having to have loads of different bitmap resolution elements.
Use the vector "templates" and then pre-render to bitmaps as required during a resolution change... just an idea
A bit silly idea though, cause one would have to render tons of icons and stuff and then write it to disk.
That's actually a very decent idea. I'd seriously look into applying for a patent for that. Sounds like something you could potentially get, plus it seems like the only way to fake res independence in the short term...
Comments
Originally posted by Placebo
And plus, it wouldn't have to be changed on resolution, and it could also have several presets pre-rendered, like .5, 1, 2, etc. That would take up what, ten megs?
You still need multiple images though for different sizes of the same interface element.
At the moment, images stored on disk at multiple resolutions and the OS picks the closest one and matches up with the display resolution and desired size. For example, in Finder, pick a window with Preview PDF icons in it then select view options. Watch the icon as you scale it past 32 pixels closely and you'll see the image and text change completely. That's because the 128px icon scaled down to toolbar sized 32px becomes illegible so they switch in a different version with larger text on it at 32px.
The problem with using vectors, apart from the GPU speed, is that an icon reduced to 16x16 quartz pixels, no matter the actual display resolution may be too small to recognise so icon designers produce icons at different sizes emphasising the important elements in an icon at each size. If you just relied on computers, the vector based image will gain and lose elements depending on the scaling or anti-alias itself into a fuzzy mush. Try opening a drawing in Illustrator and scaling it up and down and watching lines appear and disappear.
That's why one vector icon won't do. Humans can tell if an icon looks right much more easily than computers so you may as well let a designer do it for you and store them on disk already. It saves the computer having to do it later.
Apple have been gradually pushing up the sizes of the stored bitmaps ready for high resolution displays. Bearing in mind the above, I think it's a better option than vectors.
I don't see the need for it to be integrated into the OS. It only needs to be an application that can integrate among other OS applications with a better interface than Google Earth.
Originally posted by sunilraman
What if the UI *was* vector based, but each time you set the resolution it then pre-renders the UI elements as appropriate for the resolution.
Sort of on-the-fly caching so that you have the speed with pre-rendered bitmaps, but then you have the flexibility of having the vector elements as "templates" -- saves having to have loads of different bitmap resolution elements.
Use the vector "templates" and then pre-render to bitmaps as required during a resolution change... just an idea
A bit silly idea though, cause one would have to render tons of icons and stuff and then write it to disk.
That's actually a very decent idea. I'd seriously look into applying for a patent for that. Sounds like something you could potentially get, plus it seems like the only way to fake res independence in the short term...