taharvey

I've been trying not to judge, but a year into using Affinity, the key bindings often make no sense. I don't know where your key bindings derive from, but many common ones are opposite every program I've every used... Adobe, Every MacOS app, the old Freehand & Macromedia stuff.
Why be crazy? Why not use standard key bindings? It's a huge productivity suck.
Duplicate: Option-drag, not Cmd-Drag. (This is universal: every program I can think of on the Mac is option-drag).
The command is already is reserved territory fo, well... Commands! Because you overload it, drag duplication is crazy making. Every time I press command to execute a key binding while dragging... it copies the object!  Resize around centers: Option-drag, not Cmd-Drag. Again this almost universal. Please... do not overload the Command key bindings.
Constrain aspect ratio:  Affinity has non-locked aspect-ratio when dragging a corner... but opposite behavior with images! Consistent UX expectations is more important than which behavior you chose. Pick one way: Adobe defaults to unconstrained, with shift adds constraints. Omnigraffle, Pages/Keynote, are default constrained, with an unlock constraint button, but you could use shift in reverse.

Agreed. Please add HSB! 
 
HSL is a werido. The 'L' adjusts not one variable, but two. Where S is an expected 0-100%. The L 100% is really at 50%, adding black down to 0%, and adding white up to 100%. HSB is more sensible, 1 variable for each slider. Its better usability has made it more dominate in all the drawing and design apps. I think HSL would be dead if it weren't for javascript.
 
If you want to get real progressive, also change LAB for HSLuv/HSBuv, which provides a human perception based color picker, that humans can actually use! (I've never met anybody who thinks in LAB)
 
Finally, provide a mac color picker button in the Affinity color tab (not just menus)

Indeed. Its is more important to maintain in-system consistency, then cross-platform consistency.
I use 50 different applications on  a regular basis, its not worth anyone of those apps dancing the jig, while other are doing the waltz. Its at best annoying, but mostly unproductive.
It's not worth debating which is your favorite way. On the Mac, there is effectively only 1 way to drag copy. Just do it.

This ongoing thread cracks me up.
There isn't a real linux desktop market, and its very unlikely to evolve, for multiple well-known reasons. Even Linus has long ago dismissed that the "year of the linux desktop" will ever come. Linux dominates the embedded and headless server space, but isn't a compelling market for users outside of a few developers who don't come up for air above the terminal-editor level. The very culture of the linux core audience, keeps it from developing a canonical full-stack framework and singular OS distro, high quality UI/UX, or attract the non-programmer designer and business talent to ever make the market successful.
Anybody who is betting on a linux market to emerge for consumer software is kidding themselves. Hasn't happen in 25 years, no indicators that the market changing.

It is difficult to organize non-trival projects in AD without pages. For example, I  have a website with 40 views, each with its own layers showing the base view, notes, actions, and datamodel.
 
But in addition to these basic layers, every page/view has custom layers to manage overlays, popovers, sliding panels, menus, special action behaviors, or  in complex views multiple data-models. If I had to flatten all layers for 1 master artboard view, I'd have 100s of layers to manage, and it wouldn't be clear which artboard view they go with. In addition, managing shared layouts with shared layers, is a very helpful concept.
 
Artboards are not practical solution for non-trival design projects. Artboards are nice for managing aspect ratio variants. Both pages and artboards are used to good effect in Sketch, which with omnigraffle problems own most of the UI/UX market.
 
 

Icon design does have describable rules. Simplicity, clarity, meaning, recognizability. 
Your examples of Apple is a good example. An Apple with a bite out of it. Very clear imagery, very recognizable. Even still, the Apple logo of the 1970s was too muddy and busy. They fixed it, and simplified it, now it is very good icon design.
 
Here is a random grab of App icons from my computer. Some flat (Deltawalker, Omnigraffle), some 3D (Motion, Sketch), some simple logos (Sourcetree, Postico, WhatsApp, Zoom), some concept-identifiable illustrations (Xcode, Pages, Dashboard). Very different approaches, but all relatively nice design. 
 
Affinity stands out. Its super amateurish. Its design is very muddy, not simple and iconic – like say the Dropbox icon. Yet, all of the that complexity doesn't read as a identifable 'thing' or idea like say the Xcode icon which is clearly illustrating the idea of building and blueprints.
 

Icon design does have describable rules. Simplicity, clarity, meaning, recognizability. 
Your examples of Apple is a good example. An Apple with a bite out of it. Very clear imagery, very recognizable. Even still, the Apple logo of the 1970s was too muddy and busy. They fixed it, and simplified it, now it is very good icon design.
 
Here is a random grab of App icons from my computer. Some flat (Deltawalker, Omnigraffle), some 3D (Motion, Sketch), some simple logos (Sourcetree, Postico, WhatsApp, Zoom), some concept-identifiable illustrations (Xcode, Pages, Dashboard). Very different approaches, but all relatively nice design. 
 
Affinity stands out. Its super amateurish. Its design is very muddy, not simple and iconic – like say the Dropbox icon. Yet, all of the that complexity doesn't read as a identifable 'thing' or idea like say the Xcode icon which is clearly illustrating the idea of building and blueprints.