Ability to freehand select (lasso) vector objects.

[Guest]

Lately I've found myself wanting to select vector objects using a freehand (lasso) approach rather than the standard rectangle method. I've been using Vectoraster from LostMinds to create vector halftone 'sheets' that I can then bring into Affinity Designer to use within my compositions, often masked within an existing shape. This works really well, but it often means I have LOTS of unnecessary objects within my Affinity Designer file (many of them are often masked) which can lead to performance issues. I've taken to removing much of the unused area of these halftone 'sheets' to improve performance, but it's a tedious process. The only way to select them is with a rectangular selection, but often the shapes are curved and and be quite intricate which means having to go in and carefully remove only a few at a time in order not to accidentally remove ones that are actually visible within the composition.

Essentially, I'd like a way to select vector objects in a freehand manner this is similar to the way we currently are able to freehand select (lasso) a non-rectangular area of pixels.

[fde101]

While I agree that would be a good feature to have, something else to consider would be a command to delete all objects which are children of a selected layer and which are invisible (and of no effect) due to being outside of the displayed area of the parent or due to an applied mask.

Maybe have a "minimum distance" parameter to the command to keep things within a certain distance of the current boundaries ("it would be visible if the object were expanded by 15% so keep it" or something like that) to give a margin for future modification to the boundaries of the shape being cropped?

[Guest]

Having it be automated would be absolutely lovely, but I'm willing to do a little manual work with slightly better tools if it meant that development resource could be better applied towards more pressing bugs and feature requests. The idea of automated 2D 'culling' of hidden/masked objects it really interesting, and it's got me thinking of what other ideas could be pilfered from 3D tools and re-imagined in a 2D context. Cavalry has already begun to do exactly this for motion design, taking ideas such as parametric design, replicators, distributors, data, and behaviours commonly found in 3D animation tools and applying them to 2D motion design. LostMinds Patternodes also borrows a visual node-based programming metaphor used in many modern game engines, Apple's Quartz Composer, etc to enable designers to generate (and animate) vector patterns based on similar (parametric, replicators, data flow, etc) concepts. While you can still use tools such as DrawBot to accomplish similar goals using code, but the ability to experiment and discover interesting (and unexpected) ideas is often much greater in visual tools—and the learning curve tends to be much lower, which also tends to bring in a larger audience.

Anyway, I've managed to completely go off on (another) tangent. Suffice to say, automation of such things would be wonderful, but there's lots of other ideas, bugs, and  feature requests that probably deserve more attention. I really just wanted to highlight another pain-point in my current workflow with Affinity Designer in the hopes that a solution might present itself in a future update.

[Guest]

I just realized Glyphs app just added this as a feature to their latest 3.0 release: