but sure don't want my motorcycle to operate anything like a sailboat. While I appreciate the analogy I think this is fundamentally wrong and been proven by the workflow in Affinity Designer which is basically Photoshop's workflow as a vector program and it's beautiful to work with, in fact they've even made some major improvements. Everything works like you expect it to, and the keyboard shortcuts are practically identical. I've actually started using it over Adobe in many projects. They have. But that's neither the path to full-featured vector graphics functionality nor to making "everyone's life easier," just as adding a few raster-based conveniences in a vector drawing program is not the path to full-featured raster imaging. Remember the people in the early 2000's that said things like "Why would you ever want to stick a camera and an mp3 player in a phone? That wouldn't make anyone's life easier." Not saying it would be easy, but having one tool that can do both is not only possible, it's often practical as the two disciplines are very interlinked, Photoshop and Illustrator wouldn't need to be cross-compatible at all if they weren't. It's just going to take a company brave enough to try it, and I think some already are. whereas InDesign's is far better. My experience with inDesign is limited, think I made a few magazine covers in it about 5-6 years ago. but again I found it's workflow less intuitive than I would have liked. I just makes me laugh that other companies with design programs that serve as alternatives to Illustrator, InDesign etc have chosen to mimic Photoshops workflow more than Adobe has for its own programs. To me it's just obviously corporate and office politics getting in the way of making the products better. All their programs probably developed by seperate teams in a bubble and they've gotten too stuck in their ways to make fundamental changes because they don't want to rock the boat with similar set in their way customers, and consolidating the workflow across the programs would mean admitting one is better than the other and upsetting a lot of people in the process even if it makes sense in the long run. Autotracing is, in most cases an amateurish last-resort practice. There is no "conversion" of a set of pixels into a set of mathematically-defined curves. There is only, at best, a crude simulation. The autotrace program, for a very simple example, doesn't recognize that a human's pupil is a circle. Someone drawing a proper vector path does. There's pretty much mathematically-defined ways of describing everything in the known universe, pixels are hardly where the laws of physics break down and stop making sense, they are obviously mathematical themselves. So there is of course a way to do it, just because it hasn't been done yet doesn't mean it isn't possible. Thought experiment: What would you consider a "perfect" autotrace of a raster image? In terms of programmatic accuracy, wouldn't that be a perfect vector square accurately drawn around each pixel? Although entire vector-based, that would have zero technical benefit. That, in a nutshell is the gross misconception of autotracing as exists in drawing programs. It has no visual or conceptual discernment (i.e., no shape recognition). It only has an "accuracy" adjustment. Surely the answer to this is very simple, a tracer should be able to perfectly reverse a vector that has been rasterized, given it was done at a reasonable size i.e.you can't see tons of pixelation at 100-200% zoom. Think about this logically for 2 seconds, if a vector can create a perfectly identical raster the same way at any size, why is the opposite so hard to imagine being possible? Obviously it's more complicated than that or they would have done it, but if it were true that there was no possible way to mathematically define a raster into a vector accurately, the opposite wouldn't be possible either. Again not saying it's easy, but as I said just because no one's done it well yet, doesn't mean it's not possible, especially with the rise of machine learning
... View more