A new and promising spin to vector drawing: Vector Networks

[Matthias]

Check this out:

https://medium.com/figma-design/introducing-vector-networks-3b877d2b864f#.v3izmmi5c

[Figmatt]

I also came here to post that this should be added to Designer!

 

Amazing innovation.

 

But maybe the creators have patented it?

[ALx]

Exactly! A-to-B paths have always been a strange limitation to me, especially when drawing maps where you want to join multiple paths to a single point. Interestingly, AI's Mesh Tool is something of an in between of the two methods.

 

Matt, I hope they won't patent it. (And can they? Isn't it essentially a 2D representation of multiple lines connecting to a node in all 3D software?) Open source is the future!  :)

[JackofDiamonds]

Exporting something like this would probably result in semi-predictable hacking and slashing. It's a lovely idea, but I'd make it as a separate tool rather than replacing the pen; call it the Uber-Pen or something, with triangular nodes for clarity.

[cleay]

I second this request. This idea seems far more intuitive (even after 20+yrs of the old method!) and makes the traditional pen tool behaviour seem a bit archaic...

[Peregrin]

I agree.  This would make lining up vector lines so much easier.  In the meantime, if the application doesn't have it already, an option to snap to a vector line or to another node would be most helpful.

[BatteriesInc]

Another vote here, but I suspect there will be patenting/licensing issues.  However, there's nothing wrong with striking up a discussion and see where it leads - I get the impression that the company is more web focused so they may be interested in either negotiating a feature price or becoming AD's first plugin (as one possible approach to isolate licensing costs).

[Peregrin]

I know that Campaign Cartographer 3, although its AutoCAD interface is clunky and its export capabilities laughable, has a function where you can snap a node to a curve.  it gave me the idea that even if vector networks end up patented, AD might be able to get around this by simply adding a feature that locks two snapped nodes together so that—for all intents and purposes—they function as one node unless they are unlocked.

[ALx]

I honestly don't think this can be patented: https://youtu.be/ejIIvrQk2tw  B)

[organicspot]

I’ll cast my vote for this, too. Current pen tools do feel like they’re from 1987.

 

Is (something like) this on the roadmap?

[Figmatt]

I honestly don't think this can be patented: https://youtu.be/ejIIvrQk2tw  B)

ACE!

[00Ghz]

Devs? Any response?

[Peregrin]

I honestly don't think this can be patented:

Well, that's a relief.  If this were added to Affinity, I'm sure it would be even more useful than the obvious point of more efficient vectors; the idea just struck me that if Designer someday added functionality for 3D meshes, 2D vector networks might allow the app to offer seamless transition between 2D and 3D graphics.

[Ben]

You'd be surprised how much stuff gets patented (especially by the big nasties).  Things that you thought were either obvious, common sense, or had already existed in some form for years.

I remember Atari had patents on all sorts of things. It made developing games a total minefield.

[Jilly]

It would be good to at least know the opinion of the developers on this.

 

Is it something they think is worthwhile and would like to pursue? Is it in the works? Do they have better ideas? Do they think this is worse than what we currently have? And if so, why?

[John Rostron]

I'm only a sporadic user of AD, but maps are my main use. I would certainly support this development.

 

John

[JET_Affinity]

The drawing behaviors are not as newfangled as the site claims:

 

A vector network improves on the path model by allowing lines and curves between any two points instead of requiring that they all join up to form a single chain.

 

Flash's original drawing tools does this (multiple segments joined at a single node), as did its progenitor (SmartSketch) upon which it was based.

 

Figma’s bend tool...lets you drag the curve around directly.

 

FreeHand did this since way back, and didn't require switching to a different tool to do so. You just press a momentary modifier key to bend a path. It's also the behavior of Xara Designer Pro's default Bezier tool. It's useful, but is not as efficient as the ubiquitous click and drag interface for path-intensive drawing.

 

 

 

> fills in Figma start off by automatically filling all enclosed space...If you later want to add holes, just punch the holes out directly without needing to worry about curve orientation:

 

Flash's original native drawing tools behaved similarly. Illustrator's so-called "Live Paint" emulates it.

 

...no vector editor actually shows you the direction your curve is going in.

 

Not true. Either Canvas or Draw or Technical Designer displays direction arrows at the closing node. (Not at my personal machine right now, so can't verify which one.)

 

I'm not dismissing Figma. Just pointing out that marketing claims often (sometimes innocently) make claims to breaking new ground when they really don't.

 

JET

[Jilly]

I don’t think anyone on this thread actually cares if Figma was first or is unique. In fact, it has been shown Sketchup already does this.

What we all seem to agree on is that their method does indeed make more sense than the current pen tool and we’d like to see it in Affinity Designer. By stating so many other tools have some of these options and Affinity Designer doesn’t, it makes it look even worse!

[JET_Affinity]

By stating so many other tools have some of these options and Affinity Designer doesn’t, it makes it look even worse!

 

No it doesn't. What would make things "worse" would be to expect Serif to rip out the underlying path-drawing interface to replace it with a fundamentally different one. That would constitute starting all over. Drawing features don't exist in a vacuum. In an elegant interface, they are cleanly integrated so that the interface is conceptually consistent. The path-drawing interface is foundational to the rest of the interface.

 

For just one example: If the interface led the user to believe that more than two segments are joined at a single node, how would the program handle Blends? Path blends work best when the key paths have the same number of nodes. So what will be the algorithm for cleanly blending between two complicated so-called "vector networks" with different numbers of nodes? How does the blend (or the user) determine the order of those nodes?

 

A seemingly "new" approach to an interface element may look like a no-brainer "great idea" when viewed standalone all by itself. "Gee! What a breakthrough! Every program ought to have that!" It's quite another thing to try to make it mesh within the context of all the other features which make up a full-featured drawing program.

 

Again, Flash is a case-in-point. It's new-wrinkle treatment of vector drawing was clean and refreshing for those who approached it with an open mind. Like Figma, its interface treated Bezier curves as if any number of segments with coincident ends were a "single path." Unlike Figma, it used quadratic Beziers (like TrueType fonts) instead of cubic Beziers.

 

But it soon became evident that if Flash was not only going to be used standalone, but in concert with other programs, it was going to have to have serious problems interacting with the more defacto-standards. An early example was the fact that FreeHand had to provide an awkward "Flash preview" so the user could see the results he would get when the artwork was imported to Flash. Later either Macromedia or Adobe (don't recall which it was in the timeline) had to add an entire "overlay" of more common cubic Bezier path drawing tools to Flash, largely wrecking the initial elegance of the program and confusing newcomers to no end.

 

My favorite example of a poorly-made decision at the base level of the interface is the now ubiquitous insistence upon two separate primary selection tools (driven by emulating AI). That basic, foundational aspect of the interface cascades inefficiency and tedium upward throughout the rest of the program's interface structure. Illustrator's interface is chock full of such ill-conceived foundations, and it's why newcomers to the program universally find it so bewildering. That was the first thing I checked for when I first heard of Affinity. Unfortunately, that concession to Adobe-esque standard-fare was already made, and I sigh about it. But I certainly wouldn't anticipate Serif reversing that at this point in the game, because I recognize its "foundational" position in the rest of the interface.

 

So now you'll come back and say "I'm not asking Serif to replace the existing path behaviors, but merely add them!"

 

That's easier said than done. For example, most who have maintained a working familiarity with Inkscape will remember when so-called Spiro-curves were added, and the several versions of very awkward interface for it. Using it, one seemed to be drawing a path shaped one way, while the result was something entirely different. It's only just now (current version .92) been improved to the point of a clean interface in the context of the otherwise conventional path-drawing interface. And it's a wonderful and very useful feature. It's one of those things one could argue should be added to every drawing program. But it's more of a path-drawing tool behavior, the results of which work like normally-drawn paths within the context of the rest of the features. Spiro paths are not treated as entirely different (and clashing) object types.

 

The above-stated two selection tools notwithstanding, Serif is on a path toward a long-needed elegant serious-business vector drawing program. My vote is for Serif to jealously guard that elegance and determinedly avoid the "me, too" grab-bag feature glut which now permeates most of the mainstream drawing programs.

 

Affinity still has a long way to go to become "full featured." The "proof of the pudding" will be how well Serif has maintained Affinity's elegance once it is fleshed out with still missing necessary features. If it succumbs  to feature glut, we all loose.

 

JET

[Jilly]

What would make things "worse" would be to expect Serif to rip out the underlying path-drawing interface to replace it with a fundamentally different one.

 

Which wasn’t my point, and you’re exaggerating. My comment was on your statement, not Affinity Designer’s capabilities. Your statement makes it worse because it makes it seem alternative methods already existed and were known, yet weren’t considered from the start.

 

But lets assume they were considered, and just scraped. That still does not answer why! All I asked for was insight from the developers. I want their opinions on it, not your conjectures. The former will actually affect Affinity Designer’s capabilities, while the latter will not.

[JET_Affinity]

I want their opinions on it, not your conjectures.

 

 

Well, vitor, you'll just have to get over it. You clearly alluded to my post immediately preceding yours. This is an open discussion forum. Everyone does not have to agree. That's what discussion of the merits of a feature request is all about.

 

I'll post as I please. Just because you don't appreciate my "conjectures," doesn't mean no one else does.

 

JET

[Jilly]

Well, vitor, you'll just have to get over it. You clearly alluded to my post immediately preceding yours. This is an open discussion forum. Everyone does not have to agree. That's what discussion of the merits of a feature request is all about.

 

I'll post as I please. Just because you don't appreciate my "conjectures," doesn't mean no one else does.

 

Again, that wasn’t what I said. Please stop acting like I’m attacking you at every turn. I’m not. I’m asking the developers a question, and you’re just shouting after me.

[Peregrin]

On 2016-03-03 at 3:09 AM, Ben said:

You'd be surprised how much stuff gets patented (especially by the big nasties).  Things that you thought were either obvious, common sense, or had already existed in some form for years.

I remember Atari had patents on all sorts of things. It made developing games a total minefield.

 

I think most animation softwares have this feature, so I doubt anyone's patented it (unless I understand nothing about how all this works, which is entirely possible).  There is the possibility, as has already been stated, that adding such a feature would break Affinity, so that should be considered.  However, being able to draw vector networks would be quite helpful if there were ever an Affinity Animator.