I enjoy the flexibility that Inkscape provides to create brushes of any shape. However, for more complex, tablet-based graphics, Inkscape is cumbersome and unwieldy.
Currently, there are multiple steps to select a brush and it is easy to make mistakes (copying the wrong shape etc.).
A brush panel, in my opinion, would be an attractive feature.
Adam and I were chatting about this the other day, and there are some easy wins as well as some hard ones.
1st: A blob brush. Easy to understand, good for sketching and filling in stuff. We already have the eraser tool, a blob brush would simply be the inverse.
2. Something I hope to address is the notification region in the status bar. Currently it's inflexible and messy. It's just tools and dialogs throwing messages into a function with no context.
Contrast that to something like Blender:
Graphite (clearly inspired by Blender):
Both are not the best examples (jargon + unclear iconography at places), but are more visual than what Inkscape provides.
Another related thing is that for most operations currently we put information, warnings and errors in the same notification region with no difference whatsoever. We might want a proper desktop-like notification system defining the severity of a message, with which the notification system shows it as a statusbar message, on-canvas warning or blocking popup.
There are quite a few. Probably will dig up some similar topics from the past from the archived forum as well.
I think the current tools and features are like a swiss army knife that is going way out of hand with all the new feaures without addressing the issues related to the very base of the concept.
There is no target audience, but broadly everyone.
Similarly, the know-how to use the program is backwards.
You want to draw the Swedish Flag? Great, adjust the document layout, learn how inkscape treats the coordinate system, learn about the fabric of the svg document, how to export at a given resolution. Resolution? Learn about the difference between pixel and vector art and possibly how that affects printing. What are colour profiles. Oh you won't print? What format to use hosting your image online. Embedding svg-s? Here is a html quick course... and so on.
Features that I'd appreciate off the top of my head.
A reliable engine processing the tasks.
If I type in 480 px, let's make it 480 px and don't bother me with 480.0000000000367 in the codes. Every single time.
Make the scale range larger. Svg may be hard-wired to 65536 user units (?) but why does that have to affect my work, and if it must, how does it?
Maybe with a larger range of available values, the zoom range and navigation could be opened up for bigger scales.
Looking at you affinity designer...
Another svg biggie, handling colours. Ditch the good ol' linear interpolation and improve the colours -and filters- bit depth.
And if at that, change to a reliable renderer that doesn't want to melt the cpu when looking at something complex.
While also using available resources, like multi-core processing maybe?
I agree. I think the brush should be a tool in itself.
Do you think the "Blob" function would be one of the brushes in the brushes panel?
When using a graphics tablet, it would be preferable to select a brush or brush mode with a single click, as copying an object would, additionally, require the keyboard.
I think the current tools and features are like a swiss army knife that is going way out of hand with all the new feaures without addressing the issues related to the very base of the concept.
Yes, you are spot on. Many tools, but abstract from goals and process.
I think it would be favorable if you could switch between gui layouts and customize the tools more
Challenging to achieve, I expect, but ideally, yes. If I could, I would strip away many of the tools and menu items to streamline my version. Perhaps, a config file could hold my preferences and this could be applied to the newest version when I upgrade?
Low hanging fruit: units of meters and feet & inches for those using Inkscape for ground-plans. Decimals in feet are a no go, same with four digit inches.
UI enhancement:
Inkscape does so many great things, but users are often in the dark. Optional illustrated tooltips with links to online documentation would greatly expand user awareness of Inkscape's usefulness.
Example from Autodesk Fusion:
Longstanding issue: printing reliably from Inkscape. (I know there are upstream issues. Wishful thinking, I suppose.)
@COD It is possible... and an interesting idea. I have surveyed users before (via Inkscape's social media channels). One issue with these surveys may be selection bias. The few who respond to surveys may not behave like the majority.
It may also be more valuable to survey those who have stopped using Inkscape and those who never adopted it (to avoid survivorship bias), but this would be a challenging feat.
I agree that viewing the results by theme or user objective/occupation would provide useful insight.
@pacer It depends on what timeframe we're looking to implement the feature.
I would choose long-term. In my opinion, the project would be better off with a long-term perspective generally. If the devs know what is to come, they can combine efforts, avoid duplication/overlap and keep Inkscape light and nimble. Short-term goals could lead to increased workload (refactoring/rebuilding) and poorer long-term outcomes.
Of course, I would still enjoy a blob brush if that is what is implemented.
How would that work with columns and other features like tables? Would the text be contained within a single object (draggable)? I assume the text would re-flow if you resized the frame?
Low hanging fruit: units of meters and feet & inches for those using Inkscape for ground-plans. Decimals in feet are a no go, same with four digit inches.
That sounds practical. I can think of more uses for these units.
Inkscape does so many great things, but users are often in the dark. Optional illustrated tooltips with links to online documentation would greatly expand user awareness of Inkscape's usefulness.
Yes. This sounds useful.
Do you think, with the additional images, there would be much additional lag in terms of loading/running Inkscape?
Do you think, with the additional images, there would be much additional lag in terms of loading/running Inkscape?
Hard for me to tell, I'm used to waiting for Fusion to load, but it's a pretty big app. Gimp has an extensive wait on the splash compared to Inkscape. A distraction during loading like tip of the day (which could be the content of a random tooltip) could help with loading.
Many of these could be opt-in at will... hover could have small image (or icon like the current Path Effects tooltips) with a small "more" button for a larger image, more verbose tooltip, and that larger tooltip could have a link to the manual or release notes for that specific tool or operation.
@pacer I haven't thinked thorough enough to have an idea how those features could be implemented; I'll update when there's a mockup. The mockup I posted there is a no-go, and we'll need knobs and sliders like in the new object properties dialog, and perhaps be contextually aware (only showing the text tree in a collapsible section or perhaps delegated to the Layers dialog). But it'll be clearer with a mockup.
Another thing that seems simple but is apparantly not: a pixellated preview mode.
UI would be simple (in the display mode popup) as attached.
Also I'll add a "antialiasing" toggle to the popup, beside color management and grayscale toggles.
Then Inkscape would become a godsend for pixel art creators, as vector objects are a lot easier to manage than real (oil-painting-esque) pixel art creation.
Another thing that seems simple but is apparantly not: a pixellated preview mode. UI would be simple (in the display mode popup) as attached.
Nice idea. In addition, would point out the no filter mode better be next to the normal mode and thatwith a key shortcut you could toggle them between.
Filter editing needs a safety shutdown button, and that would do the job.
Speaking of rendering modes, an oversampled/supersampled one could match illustrator's.
That would hide the anti-aliasing gap issue.
Another rendering mode I'd like to see is where color gradients are substituted with discrete steps instead of linear interpolation.
It'd help with adjusting handles so when you want to end up with a gradient fill you can ensure there are no flat areas -gradient start-end handles are outisde to a tangent projection of the convex hull-.
Speaking of gradients.
Non-linear interpolation, possibly increased bit depth.
Other editing options where you can change the influence of a step -as in illustrator.
Ok, those may be out of the svg specs.
Then an idea to select multiple steps and edit them simultaneously in the defs.
One use case would be adding an offset value. And make that offset value animatable.
Another one would be to rearrange steps based on % values.
Like, make selected steps even distant from eachother or distribute them exponentially.
Or rather make those two into constraints like an lpe. So that you can still change the offset value and keep the exponential distribution after.
Ok, that may sound a bit complicated, let me give an example.
This logarithmic spiral was created by using linear gradients with every step's position constructed individually.
Now, "rotating" the gradient may seem a visually trivial request but no way to do it automatically.
The mockup I posted there is a no-go, and we'll need knobs and sliders like in the new object properties dialog, and perhaps be contextually aware (only showing the text tree in a collapsible section or perhaps delegated to the Layers dialog).
Maybe it would be helpful to take some ideas from desktop publishing software?
This maybe a foolish thought, but would markdown be an option? MD is widely used, simple to learn and can implement tables, hyperlinks etc. Perhaps it could be extended for multi-column layouts?
One feature that I have been wondering if it will be a good idea to add is Adobe Illustrator's 3D materials tool.
Having the ability to easily turn any 2D flat object into 3D and mess around with the lighting and texture would be an amazing game changer for Inkscape.
One feature that I have been wondering if it will be a good idea to add is Adobe Illustrator's 3D materials tool.
While I expect 3D tools would be tricky to implement, I agree that they would add a new dimension to Inkscape. In the meantime, there are some 3D work-arounds that are perhaps easier than you might expect.
So i'm not that knowledgable about inkscape all things considered, but I do a lot of stuff with stuff like Fusion 360 and there's a couple of things i'd love to see that might work really well for all sorts of purposes in a 2D environment instead of being 3D;
1. Parametric dimensions The ability to make shapes / paths have dimensions / properties driven by variables that can be updated later
2. Constraints / "Managed geometry" The ability to constrain the positions of paths, shapes and handles to be at specific positions relative to other previously constrained geometry. This does not necessarily need to be as rigorous as what is in fusion, it could instead take the form of a list of transformations / generative steps that happen one after the other. e.g create a new node, move this node to be at the same position as this other existing node, then extend a line from it at this angle until you hit this line. When something is updated, any "virtual" geometry is cleared and the process starts again. This would mean inkscape itself wouldn't need to work out if the constraints "collide" like fusion does, as it's just following a set of instructions with some basic error checks, e.g if you tell it to extend from a point in a direction to a line and that line would never collide, that would cause an error.
Some examples of constraints / transformations could include:
Node to Node position Node to point on curve / line (percentage so this covers line midpoints) Distance / angle between nodes Extend line to line / point
This falls quite far outside of the normal destructive workflow inkscape is built on, so I'm not exactly making bets that this will become a thing, but since you're asking i'll give a couple of reasons why I think these could be a really neat addition;
The ability to construct complex artistic geometry that can be altered just by changing the parameters The ability to easily change these complex geometries later and have parts update relative to your changes to keep the pattern correct The ability to make templates with variable size and export many SVG files with varying parameters for example from a CSV or XLS, useful for designing papercraft boxes / envelopes / other things that need dimensional accuracy and could come in many sizes / shapes (Configurations) I can see the possibility of something like being able to tie it in with a grid layout for example cutting stickers, you could put in a parameter for size, constrain a dynamic offset to be some configurable distance away from the edge of the design, then exporting them in different sizes automatically tiling the right amount of times to fill a given size of paper. You'd end up with a bunch of SVGs all with the correct size / number of stickers on the page for that size. It could even go as far as being able to nest / instance these templates / configurations, for example constraining a shape, then using its parameters to then place more of it in another larger pattern, being able to adjust the parameters of each instance and constrain them further.
I think there's a lot of room for a hybrid between more serious CAD software and traditional artistic vector editing. It all being in 2D makes it a lot less of a task than all the magic Fusion can do, although I am sure it would be a huge challenge nonetheless.
I also realise that since I am not an expert a lot of these things may already have processes that people use so please let me know if anything i've said might already exist in some form.
Hopefully this all made sense. Thanks for making an awesome bit of software :)
@Lazur exactly! Also it would be great to divert Inkscape from the SVG spec and implement GEGL filters (GIMP) and GLSL shaders (OpenGL). The GEGL filters could be rendered as bitmaps and GLSL shaders can leverage WebGL for browser compatibility.
@pacer heh, Markdown will be really interesting, but probably as a separate object than regular text (and would have 2 levels of object-to-path, Markdown text > Text > Path). Also this would not have on-canvas editability, I don't see how it would work.
I think after a bit of thought I can reword and reformat my idea into "Virtual Geometry"
Fusion is designed to be fully parametric, everything you place has predefined widths, heights, links to other geometry, etc. You're not editing a flat document, any time you change anything you're editing the history of the whole project.
Redesigning inkscape to work this way would not be productive I think, so maybe the easiest way to format this is as something similar to a path effect just for ease of explanation.
So, normal geometry works like inkscape normally does. You draw lines, destructive editing, etc.
But what if there was a special object you can add that instead of being directly editable, would generate geometry based on existing information / instructions?
For a simple example of one use-case, take this shape, i'm trying to design tabs for popup card making.
(Imagine more buttons at the top or bottom of the scrolling list for adding, removing, rearranging these operations)
So at a glance this just looks like defining a bunch of boxes, but it gives me something to start with in describing the more complicated stuff.
The start position, width, height, and any other things about a box, line, curve, etc, could be driven by 2 sources:
1. The parameters list I mentioned before, which would store values that can be referenced in a formula typed into the box 2. References to existing geometry, either virtual or "real" user placed geometry (normal paths / objects)
I can imagine an interface similar to the one Blender uses for a similar object selection purpose;
So, in this example although I didn't picture it, once you define the first box, the other boxes can reference specific nodes / relative line positions etc within the existing boxes.
So, if I wanted to change the upper tab size to, say, 30mm, I can change it in the parameters list and the shape will update.
From a process perspective, every time a change is made the virtual geometry can be calculated from the beginning, wiping any existing stuff it made and starting from scratch.
For an example of the parameters interface, I think the fusion parameters page is pretty much what I had in mind for it;
So for the inkscape example I showed before, you might have
Then those values can be typed into the virtual geometry menu as formulas, and out comes the shape with those dimensions.
For a more artistic example, take this shape
In virtual geometry this might look like;
Line 1 -> Starts at relative position (0,0) extending `height` upwards
Line 2 -> Starts at Center of line 1 minus `width / 2` on x axis, extending `width` on x axis
Join Points 1 -> (`percent` along line 1) (`percent` along line 2) (`1 - percent` along line 1) (`1 - percent` along line 2) (Close Path)
Join Points 2 -> (Start of line 1) (Start of line 2) (End of line 1) (End of line 2) (Close Path)
Changing any of the variables percent, width, height would automatically reconstruct the geometry.
This could also be linked to "real" geometry, so for example you could draw lines 1 and 2 yourself, then have the virtual geometry reference your 2 lines. Then you could drag the lines around and the virtual additions would follow your real geometry.
There's a ton of detail I could add as to how this would work in my ideal vision of it, but I'll leave it there for now. I have no idea if anything already exists that can do this, but if it doesn't exist then it would add an entirely new way of working and make it much more useful for so many projects both artistic and practical.
As a primarily CorelDraw user, I have some thoughts.
I'll start with an outline of what I hoped to accomplish in this situation.
I have a mirror at home that is damaged, I would like to tastefully cover up the damaged spots with vinyl.
This involves creating a simple design with some pictures I grabbed of pngwing, and tracing a cutline around the parts I would like to be cut off from the vinyl.
I would like to start by saying that I greatly appreciate the collective efforts of the Inkscape team and that a good number of my feedback may just be a lack of familiarity with Inkscapes workflow, but in my opinion program makes better decisions for certain UI elements that I think Inkscape would benefit to improve on.
1) Quick Page size editing, here it is in Coreldraw, when no object is selected, which is a neat feature.
2) The top toolbar isn't very context aware, for example, in Coreldraw, if I select a bitmap, I get an updated toolbar, which contains a bunch of handy features that I'd want to use such as a rotate menu with a specific angle box, mirroring shortcuts for horizontal and vertical flip, etc
Contrast this to Inkscape, which is mostly just the same
Another example is with path tools, in CorelDraw if you select two (or more) paths you will get all of your most useful path tools such as weld, trim, etc
In Inkscape, all of those features are in the path menu, or via keybinds. It would be great if Inkscape adopted a more dynamic top toolbar.
2) Universal Centre to page button, simple addition that would make my life soooo much easier, currently you have to:
- Open Align and distribute, change it to relative to page, then press e and c, by contrast on Coreldraw you can just press P, regardless of your align and distribute settings.
3) Bitmap crop feature, no comparison to be made here, this feature doesn't exist at all in Inkscape, you can jankily get around it by using clipping, I guess.
Which I would be okay with, if the clipping feature were better.
4) A feature akin to the "PowerClip" feature from CorelDraw, comparatively the Powerclip feature is much more efficient than the clipping features in Inkscape
The process in Coreldraw is as simple as right click object 1, press powerclip inside, and select object 2.
In Inkscape it involves Moving object 2 above object 1, then clipping inside, issue is, it makes object 2 invisible, meaning that if you wanted object 2 to be visible you have to make a duplicate first.
5) Better Contouring tools, Inkscapes contouring tools is a bit unintuitive to use, and without a clear way of doing just a selection path boundary, makes it rather painful compared to CorelDraws implementation which is very seamless.
In conclusion, I want to reiterate that I am extremely grateful to the Inkscape team, they have gotten the closest to an open source alternative to profession design to print software (and pretty much the only option that works on linux as CorelDraw won't function at all even via WINE) , I think with some additions and tweaks, they can really get there.
I've put together a short comparison video, that hopefully highlights my thoughts (and less of my inexperience with Inkscape 😅)
Let's keep in mind that Inkscape is a vector illustration tool (and native svg editor at that). There are hybrid programs like Corel and Affinity that have raster/bitmap contexts, but there are plenty of reasons for Inkscape developers to focus primarily on vector illustration (e.g. Adobe Illustrator and PhotoShop are not mashed together.)
It's highly unlikely that Inkscape will get true parametric CAD modelling with constraints, which requires a solver engine that itself is a massive bit of programming. Even if the solver from FreeCad or another FOSS CAD program is available for plugging in, CAD integration would be a heavy lift.
2) Universal Centre to page button, simple addition that would make my life soooo much easier,
Yes. I can see that being a useful feature. Also being able to adjust the size (both height and width) of an object to match page size would be good (helpful when creating a background layer).
Currently, for quickly changing page size, I find it easiest to create a rectangle and "Resize page to selection".
@Pacer another useful little feature that popped to mind in CorelDraw is that if you double click the rectangle tool it automatically creates a rectangle the size of the entire page
I realise that my comparison to CAD directly was probably not what I was actually thinking of, regardless hopefully I communicated the idea well enough, it would still be very functional without "constraints" as much as "commands", just a series of instructions for creating and modifying "virtual geometry", it would be up to the user to work out how to "constrain" it to be how they want, but being able to set up a series of non destructive steps on, say, a path effect would make some things much easier and not require a whole CAD solver.
I would be interested to see a list of use cases for your "virtual geometry". I think I understand the concept, but I don't know what projects would benefit. I was thinking this method could be used for creating a pack of cards (where there are many objects sharing dimensions), but I don't think it would benefit the artwork?
Here's a more complex example. When I went to design this in inkscape I realised that I didn't know exactly how big, how many, or how wide each shape would be. I also wanted to vary the width and height of each one as it curves around, and doing all that manually and consistently with only a mirror path effect / rotate copies just wasn't possible, so I ended up writing code to generate an SVG file in the browser locally, and tweaked it until it displayed correctly on the preview. Notice how the shapes at the bottom are thinner, their scale is not uniform which I don't think can be done with just rotate copies
The basic element of this can be described in virtual geometry similarly to the one above, constructing lines then connecting at different points along them
These could then be instanced with a command similar to rotate copies, but within the virtual geometry effect instead. Then the parameters for each instance can be varied to produce an effect similar to the one above.
Another thing I find myself doing a lot is needing to realign geometry when multiple paths are necessary, take this example of a petal with internal connecting lines. Moving the inner or outer part of the petal leaves the inner lines without a connection.
In reality on this design the internal paths are supposed to connect at a given percentage along the inner and outer curves. If these internal joining paths were virtual (Or indeed, the whole thing), they could be generated to join at those percentages, with a configurable amount of bending, and moving either of the petals would leave the internal part still attached.
This is another one where I would have loved to have these tools. It's pretty simple, but if I could generate it "virtually", it would mean I could extend the pattern out as many layers as I like, tweak the dimensions and see what the result is.
Here's another example of what i've been doing with popup cards
Prototyping these takes a lot of fiddling, iteration, changing sizes of different elements. The ability to construct these shapes from parameters would make a HUGE difference.
Right now, if a tab is too short or long, I have to go through a few steps, including drawing alignment lines to snap to and making sure the fold lines are correct, since I need to make the fold lines separate so I can use my cutter's embossing feature to make perfect crease lines.
Sometimes tabs need to be the same size as some other tab, parameters would make this change automatic as the sizes of different sections can be driven by the same parameter rather than having to double check and mentally reconstruct it to work out if I also need to change this other tab's size too
If it was virtual, I could change some parameters and come out with shapes that are perfectly sized relative to each other.
I kind of see this a bit like a simplified, strictly linear, more directly controlled version of Blender's geometry nodes. You add a new piece of virtual geometry (Adding a path effect to something) and it can reference parts of your existing design to add live updating geometry, or just generate them from scratch using parameters.
Thinking about it, a path effect would definitely be the best place for it to go (if it *was* going to be a thing), as path effects already have the non destructive behaviour of regenerating their output geometry when things get updated.
I can probably come up with an endless list of possible examples of things I would do with this, just the ability to define lines and boxes from parameters for the card cutting would be awesome, fusion is a somewhat useful tool for stuff like popup cards but its export options are annoyingly bad without a paid SVG plugin, and it's still not really designed for artwork in any form, or a hybrid of destructive and non destructive workflows.
Visio is dying. The software is getting more and more dated and it doesn't integrate with Office365 well. I have been thinking about what it would take for users at my company to move to Inkscape. The ability to quickly and easily add text to an object or path is a big one. In Visio you click on the shape and type, double click to edit. The process in Inkscape to flow text into a path is considerably more complex. Adding an easy way for less skilled users to add text to a shape would be huge.
@yellek agreed, having more moodboard/flowchart-like functionality would be nice. Do you have any mockups in mind or how we can design this in Inkscape? My current thought is that we can bolt it onto the Connectors tool, which is already sort-of used to create flowchart-like elements.
Also just to level everyone's expectations: this is a place to dream big, and is no way related to what would actually be implemented (except if you can write the feature/fix or pay/ask someone to do so). But idea pools are always good.
My use cases probably don't go to flow charts - If I need to do that I'll use a proper architecture tool like Archi. Just being able to easily add text to a shape is the biggest one for my colleagues.
Community Café The Future of Inkscape
Inkscape.org
A space for users to discuss thoughts on the future of Inkscape.
Attach mock-ups of new features/tools or UX improvements, for discussion.
(These will be placed in the showcase below).
I enjoy the flexibility that Inkscape provides to create brushes of any shape. However, for more complex, tablet-based graphics, Inkscape is cumbersome and unwieldy.
Currently, there are multiple steps to select a brush and it is easy to make mistakes (copying the wrong shape etc.).
A brush panel, in my opinion, would be an attractive feature.
Adam and I were chatting about this the other day, and there are some easy wins as well as some hard ones.
1st: A blob brush. Easy to understand, good for sketching and filling in stuff. We already have the eraser tool, a blob brush would simply be the inverse.
A rudimentary example is Boxy SVG's blob brush. It has:
To improve on it is simple:
2. Something I hope to address is the notification region in the status bar. Currently it's inflexible and messy. It's just tools and dialogs throwing messages into a function with no context.
Contrast that to something like Blender:
Graphite (clearly inspired by Blender):
Both are not the best examples (jargon + unclear iconography at places), but are more visual than what Inkscape provides.
Another related thing is that for most operations currently we put information, warnings and errors in the same notification region with no difference whatsoever. We might want a proper desktop-like notification system defining the severity of a message, with which the notification system shows it as a statusbar message, on-canvas warning or blocking popup.
There are quite a few. Probably will dig up some similar topics from the past from the archived forum as well.
I think the current tools and features are like a swiss army knife that is going way out of hand with all the new feaures without addressing the issues related to the very base of the concept.
There is no target audience, but broadly everyone.
Yes, you can even create tutorials with a comprehensive attitude, like https://www.inkscapefriends.com/inkscape-for-artists-step-by-step/
but with all seriousness, would you want to learn about tiled clones before you want to hit the canvas inking manga sketches?
Or learning how to use the text tool before creating fashion drawings -if that is your focus.
And other big ones, even if seem reasonable.
Learn how to use gradients and draw drop shadows BEFORE delving into graphic design and understand you better avoid those.
For that particular reason I think it would be favorable if you could switch between gui layouts and customize the tools more, like in blender.
(Object mode, edit mode -> sketching mode, drawing mode, lettersetting mode, layouting mode, pixelart asset exporting mode etc.)
Also to have a variety of targeted manuals.
Similarly, the know-how to use the program is backwards.
You want to draw the Swedish Flag? Great, adjust the document layout, learn how inkscape treats the coordinate system, learn about the fabric of the svg document, how to export at a given resolution. Resolution? Learn about the difference between pixel and vector art and possibly how that affects printing. What are colour profiles. Oh you won't print? What format to use hosting your image online. Embedding svg-s? Here is a html quick course... and so on.
Features that I'd appreciate off the top of my head.
A reliable engine processing the tasks.
If I type in 480 px, let's make it 480 px and don't bother me with 480.0000000000367 in the codes. Every single time.
Make the scale range larger. Svg may be hard-wired to 65536 user units (?) but why does that have to affect my work, and if it must, how does it?
Maybe with a larger range of available values, the zoom range and navigation could be opened up for bigger scales.
Looking at you affinity designer...
Another svg biggie, handling colours. Ditch the good ol' linear interpolation and improve the colours -and filters- bit depth.
And if at that, change to a reliable renderer that doesn't want to melt the cpu when looking at something complex.
While also using available resources, like multi-core processing maybe?
For example look at that:
https://inkscape.org/forums/wip/frosty-mountains/#c51603
Why is it unreachable with filtering although the tools suppose to allow everything?
It's all a tease.
Those are the features that can hardly be captured with an informative mockup.
I assume you are likely looking at more presentable tools and features wishlist.
Will post some of those later, just got to get these out of the way.
Same with tools I don't use or avoid using.
I think that brush panel could really make a change for illustrators using a tablet.
Having said that I do know inkscape would get alot of hate for it if raster brushes weren't implemented at the same time.
@ltlnx Wow! Great idea. I think a blob brush would be SO useful.
Agreed. Sketching is the thing which Inkscape is missing.
With traditional art, sketching is the first step. It is fundamental to planning.
It fits with our mental process... We have a rough idea about what we want to achieve and this concept is refined over time.
I agree. I think the brush should be a tool in itself.
Do you think the "Blob" function would be one of the brushes in the brushes panel?
When using a graphics tablet, it would be preferable to select a brush or brush mode with a single click, as copying an object would, additionally, require the keyboard.
Also, this will be a welcome improvement.
Perhaps, if the notifications are terse, a ? icon within the status bar could trigger a panel with a more verbose explanation?
Yes, you are spot on. Many tools, but abstract from goals and process.
True. It would help if documentation/tutorials were outcome-focused rather than tool/feature-focused.
Challenging to achieve, I expect, but ideally, yes.
If I could, I would strip away many of the tools and menu items to streamline my version.
Perhaps, a config file could hold my preferences and this could be applied to the newest version when I upgrade?
Low hanging fruit: units of meters and feet & inches for those using Inkscape for ground-plans. Decimals in feet are a no go, same with four digit inches.
UI enhancement:
Inkscape does so many great things, but users are often in the dark. Optional illustrated tooltips with links to online documentation would greatly expand user awareness of Inkscape's usefulness.
Example from Autodesk Fusion:
Longstanding issue: printing reliably from Inkscape. (I know there are upstream issues. Wishful thinking, I suppose.)
Is it possible to set up a user survey at the Inkscape website when someone downloads Inkscape?
Be interesting to see how users are actually using Inkscape.
Graphic Design/Illustration?
Web Design?
Maker/Crafter (technically what I call CAD/CAM users)?
Half dozen more cats?
New video: https://youtu.be/1_b2wYFXHG0?si=F-yyrIiJVbG9oLuN
@pacer It depends on what timeframe we're looking to implement the feature.
Short term: Blob brush tool. Spec as described there.
Middle term: Blob brush tool + brush shapes from clipboard. Similar to Calligraphy tool.
Long term: A mode in the vector brush tool (instead of stacking up, make them the same path).
Keeping the discussion going: I'd like a paragraph tool.
Complete layout control, winding and flowing paragraphs, margins on textboxes, object wraparound as an attribute.
This was my initial mockup for the dialog, Moini hates it for being overly technical but it could be improved and refined:
@COD It is possible... and an interesting idea. I have surveyed users before (via Inkscape's social media channels). One issue with these surveys may be selection bias. The few who respond to surveys may not behave like the majority.
It may also be more valuable to survey those who have stopped using Inkscape and those who never adopted it (to avoid survivorship bias), but this would be a challenging feat.
I agree that viewing the results by theme or user objective/occupation would provide useful insight.
@ltlnx
I would choose long-term. In my opinion, the project would be better off with a long-term perspective generally. If the devs know what is to come, they can combine efforts, avoid duplication/overlap and keep Inkscape light and nimble. Short-term goals could lead to increased workload (refactoring/rebuilding) and poorer long-term outcomes.
Of course, I would still enjoy a blob brush if that is what is implemented.
@ltlnx I like your text layout idea.
How would that work with columns and other features like tables?
Would the text be contained within a single object (draggable)?
I assume the text would re-flow if you resized the frame?
@TylerDurden
That sounds practical. I can think of more uses for these units.
Yes. This sounds useful.
Do you think, with the additional images, there would be much additional lag in terms of loading/running Inkscape?
Hard for me to tell, I'm used to waiting for Fusion to load, but it's a pretty big app. Gimp has an extensive wait on the splash compared to Inkscape. A distraction during loading like tip of the day (which could be the content of a random tooltip) could help with loading.
Many of these could be opt-in at will... hover could have small image (or icon like the current Path Effects tooltips) with a small "more" button for a larger image, more verbose tooltip, and that larger tooltip could have a link to the manual or release notes for that specific tool or operation.
@pacer I haven't thinked thorough enough to have an idea how those features could be implemented; I'll update when there's a mockup. The mockup I posted there is a no-go, and we'll need knobs and sliders like in the new object properties dialog, and perhaps be contextually aware (only showing the text tree in a collapsible section or perhaps delegated to the Layers dialog). But it'll be clearer with a mockup.
Here was the list of topics until 2019.
https://alpha.inkscape.org/vectors/www.inkscapeforum.com/viewforumce40.html?f=28
Multipage support is implemented since and cmyk support is on the way.
Besides there are alot on that wishlist which still make sense.
Another thing that seems simple but is apparantly not: a pixellated preview mode.
UI would be simple (in the display mode popup) as attached.
Also I'll add a "antialiasing" toggle to the popup, beside color management and grayscale toggles.
Then Inkscape would become a godsend for pixel art creators, as vector objects are a lot easier to manage than real (oil-painting-esque) pixel art creation.
Nice idea. In addition, would point out the no filter mode better be next to the normal mode and thatwith a key shortcut you could toggle them between.
Filter editing needs a safety shutdown button, and that would do the job.
Speaking of rendering modes, an oversampled/supersampled one could match illustrator's.
That would hide the anti-aliasing gap issue.
Another rendering mode I'd like to see is where color gradients are substituted with discrete steps instead of linear interpolation.
It'd help with adjusting handles so when you want to end up with a gradient fill you can ensure there are no flat areas -gradient start-end handles are outisde to a tangent projection of the convex hull-.
Speaking of gradients.
Non-linear interpolation, possibly increased bit depth.
Other editing options where you can change the influence of a step -as in illustrator.
Ok, those may be out of the svg specs.
Then an idea to select multiple steps and edit them simultaneously in the defs.
One use case would be adding an offset value. And make that offset value animatable.
Another one would be to rearrange steps based on % values.
Like, make selected steps even distant from eachother or distribute them exponentially.
Or rather make those two into constraints like an lpe. So that you can still change the offset value and keep the exponential distribution after.
Ok, that may sound a bit complicated, let me give an example.
This logarithmic spiral was created by using linear gradients with every step's position constructed individually.
Now, "rotating" the gradient may seem a visually trivial request but no way to do it automatically.
Here is the svg source for reference:
https://openclipart.org/download/228444/lspektr5.svg
Maybe it would be helpful to take some ideas from desktop publishing software?
This maybe a foolish thought, but would markdown be an option? MD is widely used, simple to learn and can implement tables, hyperlinks etc. Perhaps it could be extended for multi-column layouts?
One feature that I have been wondering if it will be a good idea to add is Adobe Illustrator's 3D materials tool.
Having the ability to easily turn any 2D flat object into 3D and mess around with the lighting and texture would be an amazing game changer for Inkscape.
https://www.youtube.com/shorts/hj_G1ewHouU
Welcome to the Inkscape forum @SpeeQz1
While I expect 3D tools would be tricky to implement, I agree that they would add a new dimension to Inkscape.
In the meantime, there are some 3D work-arounds that are perhaps easier than you might expect.
Here is a video I created showing how to emulate a pattern on a 3D sphere:
https://office.inkscape.org/nextcloud/index.php/s/eskqNCH9FBZkNSi
UPDATE:
@SpeeQz1 just to illustrate that these things can be done without 3D tools, I created the following graphic using Inkscape:
So i'm not that knowledgable about inkscape all things considered, but I do a lot of stuff with stuff like Fusion 360 and there's a couple of things i'd love to see that might work really well for all sorts of purposes in a 2D environment instead of being 3D;
1. Parametric dimensions
The ability to make shapes / paths have dimensions / properties driven by variables that can be updated later
2. Constraints / "Managed geometry"
The ability to constrain the positions of paths, shapes and handles to be at specific positions relative to other previously constrained geometry. This does not necessarily need to be as rigorous as what is in fusion, it could instead take the form of a list of transformations / generative steps that happen one after the other. e.g create a new node, move this node to be at the same position as this other existing node, then extend a line from it at this angle until you hit this line. When something is updated, any "virtual" geometry is cleared and the process starts again. This would mean inkscape itself wouldn't need to work out if the constraints "collide" like fusion does, as it's just following a set of instructions with some basic error checks, e.g if you tell it to extend from a point in a direction to a line and that line would never collide, that would cause an error.
Some examples of constraints / transformations could include:
Node to Node position
Node to point on curve / line (percentage so this covers line midpoints)
Distance / angle between nodes
Extend line to line / point
This falls quite far outside of the normal destructive workflow inkscape is built on, so I'm not exactly making bets that this will become a thing, but since you're asking i'll give a couple of reasons why I think these could be a really neat addition;
The ability to construct complex artistic geometry that can be altered just by changing the parameters
The ability to easily change these complex geometries later and have parts update relative to your changes to keep the pattern correct
The ability to make templates with variable size and export many SVG files with varying parameters for example from a CSV or XLS, useful for designing papercraft boxes / envelopes / other things that need dimensional accuracy and could come in many sizes / shapes (Configurations)
I can see the possibility of something like being able to tie it in with a grid layout for example cutting stickers, you could put in a parameter for size, constrain a dynamic offset to be some configurable distance away from the edge of the design, then exporting them in different sizes automatically tiling the right amount of times to fill a given size of paper. You'd end up with a bunch of SVGs all with the correct size / number of stickers on the page for that size.
It could even go as far as being able to nest / instance these templates / configurations, for example constraining a shape, then using its parameters to then place more of it in another larger pattern, being able to adjust the parameters of each instance and constrain them further.
I think there's a lot of room for a hybrid between more serious CAD software and traditional artistic vector editing. It all being in 2D makes it a lot less of a task than all the magic Fusion can do, although I am sure it would be a huge challenge nonetheless.
I also realise that since I am not an expert a lot of these things may already have processes that people use so please let me know if anything i've said might already exist in some form.
Hopefully this all made sense. Thanks for making an awesome bit of software :)
Svg comes with diffuse filtering. However the low bit depth makes it practically unusable.
Manual suggests to blur the result and live with the sloppy quality.
http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Filters-Lighting.html
But if we had a proper way of 3D shading based on a bump map, the magic would start if we had the ability to use (generated) normal maps.
and as a side effect, apply images as environment maps.
https://www.youtube.com/watch?v=xLPRHNiXE6w
There could be a dropdown library showing a collection of similar spheres, that could be stretched to any shape.
And it doesn't even need a real 3D geometry, just a normal map which can be used as a source for proper displacement.
Similar in nature to using the bend lpe, but with more complex shape possibilities.
(For example on using the path effect https://openclipart.org/detail/212914/spam-iron-man; Far from Sorayama, just to illustrate the idea.)
@Lazur exactly! Also it would be great to divert Inkscape from the SVG spec and implement GEGL filters (GIMP) and GLSL shaders (OpenGL). The GEGL filters could be rendered as bitmaps and GLSL shaders can leverage WebGL for browser compatibility.
@pacer heh, Markdown will be really interesting, but probably as a separate object than regular text (and would have 2 levels of object-to-path, Markdown text > Text > Path). Also this would not have on-canvas editability, I don't see how it would work.
Really wanted:
- Dimensions like length and area in the UI (like Xara Designer Pro)
- More fluid strokes based on pressure ( the current rendition is so-so. Have a look at Illustrator and Xara)
- Export to DWG and HTML
- Layered PDF export
@tiggerbiggo
I'm not familiar with Fusion 360, but this seams interesting. Could you give a visual example of how it would work?
This looks interesting. What sort of user-interface would you expect for a feature like this? It would be good to see a visual example of the UI.
@Pacer
I think after a bit of thought I can reword and reformat my idea into "Virtual Geometry"
Fusion is designed to be fully parametric, everything you place has predefined widths, heights, links to other geometry, etc. You're not editing a flat document, any time you change anything you're editing the history of the whole project.
Redesigning inkscape to work this way would not be productive I think, so maybe the easiest way to format this is as something similar to a path effect just for ease of explanation.
So, normal geometry works like inkscape normally does. You draw lines, destructive editing, etc.
But what if there was a special object you can add that instead of being directly editable, would generate geometry based on existing information / instructions?
For a simple example of one use-case, take this shape, i'm trying to design tabs for popup card making.
(Imagine more buttons at the top or bottom of the scrolling list for adding, removing, rearranging these operations)
So at a glance this just looks like defining a bunch of boxes, but it gives me something to start with in describing the more complicated stuff.
The start position, width, height, and any other things about a box, line, curve, etc, could be driven by 2 sources:
1. The parameters list I mentioned before, which would store values that can be referenced in a formula typed into the box
2. References to existing geometry, either virtual or "real" user placed geometry (normal paths / objects)
I can imagine an interface similar to the one Blender uses for a similar object selection purpose;
So, in this example although I didn't picture it, once you define the first box, the other boxes can reference specific nodes / relative line positions etc within the existing boxes.
So, if I wanted to change the upper tab size to, say, 30mm, I can change it in the parameters list and the shape will update.
From a process perspective, every time a change is made the virtual geometry can be calculated from the beginning, wiping any existing stuff it made and starting from scratch.
For an example of the parameters interface, I think the fusion parameters page is pretty much what I had in mind for it;
So for the inkscape example I showed before, you might have
tab_width = 30mm
upper_height = 30mm
lower_height = 80mm
total_height = upper_height + lower_height = 110mm
Then those values can be typed into the virtual geometry menu as formulas, and out comes the shape with those dimensions.
For a more artistic example, take this shape
In virtual geometry this might look like;
Line 1 -> Starts at relative position (0,0) extending `height` upwards
Line 2 -> Starts at Center of line 1 minus `width / 2` on x axis, extending `width` on x axis
Join Points 1 ->
(`percent` along line 1)
(`percent` along line 2)
(`1 - percent` along line 1)
(`1 - percent` along line 2)
(Close Path)
Join Points 2 ->
(Start of line 1)
(Start of line 2)
(End of line 1)
(End of line 2)
(Close Path)
Changing any of the variables percent, width, height would automatically reconstruct the geometry.
This could also be linked to "real" geometry, so for example you could draw lines 1 and 2 yourself, then have the virtual geometry reference your 2 lines. Then you could drag the lines around and the virtual additions would follow your real geometry.
There's a ton of detail I could add as to how this would work in my ideal vision of it, but I'll leave it there for now. I have no idea if anything already exists that can do this, but if it doesn't exist then it would add an entirely new way of working and make it much more useful for so many projects both artistic and practical.
Hopefully this makes sense XD
As a primarily CorelDraw user, I have some thoughts.
I'll start with an outline of what I hoped to accomplish in this situation.
I have a mirror at home that is damaged, I would like to tastefully cover up the damaged spots with vinyl.
This involves creating a simple design with some pictures I grabbed of pngwing, and tracing a cutline around the parts I would like to be cut off from the vinyl.
I would like to start by saying that I greatly appreciate the collective efforts of the Inkscape team and that a good number of my feedback may just be a lack of familiarity with Inkscapes workflow, but in my opinion program makes better decisions for certain UI elements that I think Inkscape would benefit to improve on.
1) Quick Page size editing, here it is in Coreldraw, when no object is selected, which is a neat feature.
2) The top toolbar isn't very context aware, for example, in Coreldraw, if I select a bitmap, I get an updated toolbar, which contains a bunch of handy features that I'd want to use such as a rotate menu with a specific angle box, mirroring shortcuts for horizontal and vertical flip, etc
Contrast this to Inkscape, which is mostly just the same
Another example is with path tools, in CorelDraw if you select two (or more) paths you will get all of your most useful path tools such as weld, trim, etc
In Inkscape, all of those features are in the path menu, or via keybinds. It would be great if Inkscape adopted a more dynamic top toolbar.
2) Universal Centre to page button, simple addition that would make my life soooo much easier, currently you have to:
- Open Align and distribute, change it to relative to page, then press e and c, by contrast on Coreldraw you can just press P, regardless of your align and distribute settings.
3) Bitmap crop feature, no comparison to be made here, this feature doesn't exist at all in Inkscape, you can jankily get around it by using clipping, I guess.
Which I would be okay with, if the clipping feature were better.
4) A feature akin to the "PowerClip" feature from CorelDraw, comparatively the Powerclip feature is much more efficient than the clipping features in Inkscape
The process in Coreldraw is as simple as right click object 1, press powerclip inside, and select object 2.
In Inkscape it involves Moving object 2 above object 1, then clipping inside, issue is, it makes object 2 invisible, meaning that if you wanted object 2 to be visible you have to make a duplicate first.
5) Better Contouring tools, Inkscapes contouring tools is a bit unintuitive to use, and without a clear way of doing just a selection path boundary, makes it rather painful compared to CorelDraws implementation which is very seamless.
In conclusion, I want to reiterate that I am extremely grateful to the Inkscape team, they have gotten the closest to an open source alternative to profession design to print software (and pretty much the only option that works on linux as CorelDraw won't function at all even via WINE) , I think with some additions and tweaks, they can really get there.
I've put together a short comparison video, that hopefully highlights my thoughts (and less of my inexperience with Inkscape 😅)
https://www.youtube.com/watch?v=oOwUwl9nU54&feature=youtu.be
I know I could have done some of those steps better, (I was avoiding keybinds that involved taking my hand of my mouse).
It appears the embed didn't work, here is a direct link:
https://www.youtube.com/watch?v=oOwUwl9nU54&feature=youtu.be
Let's keep in mind that Inkscape is a vector illustration tool (and native svg editor at that). There are hybrid programs like Corel and Affinity that have raster/bitmap contexts, but there are plenty of reasons for Inkscape developers to focus primarily on vector illustration (e.g. Adobe Illustrator and PhotoShop are not mashed together.)
It's highly unlikely that Inkscape will get true parametric CAD modelling with constraints, which requires a solver engine that itself is a massive bit of programming. Even if the solver from FreeCad or another FOSS CAD program is available for plugging in, CAD integration would be a heavy lift.
WRT Corel bitmap operations (crop etc.) there are some limited extensions, but the shape-builder tool can now offer some fast cropping/clipping functionality. Inkscape will not likely have a generous selection of tools that can be found in GIMP; however, images can be linked and updated in GIMP, with the results shown in the Inkscape workspace.
@No.Chill
Yes. I can see that being a useful feature. Also being able to adjust the size (both height and width) of an object to match page size would be good (helpful when creating a background layer).
Currently, for quickly changing page size, I find it easiest to create a rectangle and "Resize page to selection".
@Pacer another useful little feature that popped to mind in CorelDraw is that if you double click the rectangle tool it automatically creates a rectangle the size of the entire page
I realise that my comparison to CAD directly was probably not what I was actually thinking of, regardless hopefully I communicated the idea well enough, it would still be very functional without "constraints" as much as "commands", just a series of instructions for creating and modifying "virtual geometry", it would be up to the user to work out how to "constrain" it to be how they want, but being able to set up a series of non destructive steps on, say, a path effect would make some things much easier and not require a whole CAD solver.
@tiggerbiggo
I would be interested to see a list of use cases for your "virtual geometry". I think I understand the concept, but I don't know what projects would benefit. I was thinking this method could be used for creating a pack of cards (where there are many objects sharing dimensions), but I don't think it would benefit the artwork?
Can't seem to post the reply I planned, it just takes me to a "new comment" page. Will try again in a bit, I copied it into a notepad.
Here's a more complex example. When I went to design this in inkscape I realised that I didn't know exactly how big, how many, or how wide each shape would be. I also wanted to vary the width and height of each one as it curves around, and doing all that manually and consistently with only a mirror path effect / rotate copies just wasn't possible, so I ended up writing code to generate an SVG file in the browser locally, and tweaked it until it displayed correctly on the preview. Notice how the shapes at the bottom are thinner, their scale is not uniform which I don't think can be done with just rotate copies
The basic element of this can be described in virtual geometry similarly to the one above, constructing lines then connecting at different points along them
These could then be instanced with a command similar to rotate copies, but within the virtual geometry effect instead. Then the parameters for each instance can be varied to produce an effect similar to the one above.
Another thing I find myself doing a lot is needing to realign geometry when multiple paths are necessary, take this example of a petal with internal connecting lines. Moving the inner or outer part of the petal leaves the inner lines without a connection.
In reality on this design the internal paths are supposed to connect at a given percentage along the inner and outer curves. If these internal joining paths were virtual (Or indeed, the whole thing), they could be generated to join at those percentages, with a configurable amount of bending, and moving either of the petals would leave the internal part still attached.
This is another one where I would have loved to have these tools. It's pretty simple, but if I could generate it "virtually", it would mean I could extend the pattern out as many layers as I like, tweak the dimensions and see what the result is.
Here's another example of what i've been doing with popup cards
Prototyping these takes a lot of fiddling, iteration, changing sizes of different elements. The ability to construct these shapes from parameters would make a HUGE difference.
Right now, if a tab is too short or long, I have to go through a few steps, including drawing alignment lines to snap to and making sure the fold lines are correct, since I need to make the fold lines separate so I can use my cutter's embossing feature to make perfect crease lines.
Sometimes tabs need to be the same size as some other tab, parameters would make this change automatic as the sizes of different sections can be driven by the same parameter rather than having to double check and mentally reconstruct it to work out if I also need to change this other tab's size too
If it was virtual, I could change some parameters and come out with shapes that are perfectly sized relative to each other.
I kind of see this a bit like a simplified, strictly linear, more directly controlled version of Blender's geometry nodes. You add a new piece of virtual geometry (Adding a path effect to something) and it can reference parts of your existing design to add live updating geometry, or just generate them from scratch using parameters.
Thinking about it, a path effect would definitely be the best place for it to go (if it *was* going to be a thing), as path effects already have the non destructive behaviour of regenerating their output geometry when things get updated.
I can probably come up with an endless list of possible examples of things I would do with this, just the ability to define lines and boxes from parameters for the card cutting would be awesome, fusion is a somewhat useful tool for stuff like popup cards but its export options are annoyingly bad without a paid SVG plugin, and it's still not really designed for artwork in any form, or a hybrid of destructive and non destructive workflows.
Many extensions have a live preview and have numeric inputs with choice of units etc. (e.g. the many laser box extensions).
WRT the petal wit connecting lines, IIRC, there is a Attach Path LPE that might be useful. https://wiki.inkscape.org/wiki/Release_notes/0.92#Attach_Path
Visio is dying. The software is getting more and more dated and it doesn't integrate with Office365 well. I have been thinking about what it would take for users at my company to move to Inkscape. The ability to quickly and easily add text to an object or path is a big one. In Visio you click on the shape and type, double click to edit. The process in Inkscape to flow text into a path is considerably more complex. Adding an easy way for less skilled users to add text to a shape would be huge.
@yellek agreed, having more moodboard/flowchart-like functionality would be nice. Do you have any mockups in mind or how we can design this in Inkscape? My current thought is that we can bolt it onto the Connectors tool, which is already sort-of used to create flowchart-like elements.
Also just to level everyone's expectations: this is a place to dream big, and is no way related to what would actually be implemented (except if you can write the feature/fix or pay/ask someone to do so). But idea pools are always good.
My use cases probably don't go to flow charts - If I need to do that I'll use a proper architecture tool like Archi. Just being able to easily add text to a shape is the biggest one for my colleagues.