Native MacOSX Support!!
"Inkscape is now a first-rate native macOS application, and no longer requires XQuartz to operate. The minimum required operating system version is OS X El Capitan 10.11.
It has a standard Mac-style menu bar (rather than a menu bar within the window). Keyboard shortcuts now use the command (⌘) key rather than the control key. Retina display screen resolution is now supported. The build is now cleanly 64-bit, a prerequisite for macOS Catalina 10.15 and beyond. It comes bundled with Python 3 to power Inkscape extensions. "
Oh I'm so glad to see progress is being made in this area. Inkscape was just unusable for me on macOS. Now it's a bit more bearable. Keep up the good work!
First grats to the devs and hurray for high quality open source.
My impression though for OS X, that with Affinity Designer a lot of need for a cheaper Illustrator has vanished (if you need it to be free/open source/on principle then Inkscape is great)
Inkscape is a lot more versatile than Affinity Designer. So it all depends what you need to do. But also how you like to or are used to work.
Inkscape is more of a CorelDRAW than an Illustrator clone. People who come from CorelDRAW, like me, usually disrelish the way curves are edited in Adobe products.
As much as I like Affinity products (I own both Photo & Publisher), I prefer Inkscape any day over Designer (or Illustrator) for this reason.
The other reason I simply couldn't use Designer at all until approx. six months ago was that is lacked snapping of curve handles to grid intersections. It took Affinity years to implement this essential feature for people who do typography or logotypes[1].
As others mentioned: the Carbon native Inkscape (beta and now release) are even slower than the XQuartz version. Which was already slow, compared to running Inkscape on Linux on the same MBP.
For this reason I now use Vectornator[2] (free) as much as I can. For the missing bits I go back and forth with Inkscape.
Not ideal but it's only recent that the macOS version of Inkscape has seen more love from the developers. I hope that soon performance will improve and I can do everything in one app.
A tangent, but Vectornator is completely mystifying to me. Not the application itself, which is a very accessible vector illustration tool, but the company. It’s a Berlin-based startup with $5 million in venture funding, and they don’t have any kind of externally-visible business model for their free, no-IAP apps for MacOS and iOS.
(Further muddying the waters is a recent update to the iOS app that adds achievement badges. I don’t know what kind of product roadmap for a creative tool ends up prioritizing that, but that’s where they are.)
Yeah, they're showing 31 people and a bit over 5MM EUR as their funding. But their job ads are ~60 EUR/year, so that's what, maybe 1-2 years of funding? Not exactly long-term as they're claiming.
There's something else going on. Perhaps training an AI?
As a researcher who composes a figure or two every now and then, I find Inkscape to be far more intuitive than Affinity Designer. Affinity makes simple things hard. Creating arrows was not a feature until recently, resizing multiple selections would mess text objects, and changing the size of the canvas requires copying all content to clipboard and creating a second, separate canvas. Not to mention that exports, which should be simple, are also hard to understand for beginners / non-designers, and the gazillion of menus makes it impossible to ever find what you need.
After many years of waiting, I'm happy to see Inkscape jumping to MacOS too. Kudos to all devs and contributors that made this possible.
As a researcher who draw figures quite often... I use PowerPoint. That sounds like a joke, but the ratio features/learning curse is perfect. On the contrary I was never able to draw anything with Inkscape, and I give it a try every few years or so.
> Not to mention that exports, which should be simple, are also hard to understand for beginners / non-designers
That’s strange, I’m not a designer and I found it very easy to understand, and very powerful too — that is unless you don’t know there’s a separate Export Persona. Guess a tiny amount of prior experience with Sketch helped a bit in my case.
Unfortunately, this native version is quite laggy compared to the previous XQuartz releases. I wonder how it compares with a Linux version in a VM; my previous experience suggests it may work better.
Hopefully removing a whole extra layer will make it easier to profile and improve, both for Mac tools and for Mac developers. One of my main complaints about Firefox is that everything is so completely custom (even the parts that look 'native') that all my years of Mac development knowledge are useless.
I see that, for better or worse, Inkscape is still fully GTK+ even on macOS. It uses the slightly improved native features of that toolkit, but all the controls still look like they're ripped from an X11 app. The text labels don't even use Mac naming conventions.
I wonder if GTK+ is architected in a way that's going to be easy to optimize on the Mac graphics architecture.
Yeah same issue here, much worse performance than the XQuartz releases. Dragging objects, especially arrows, is extremely laggy and awful. I've moved to Affinity Designer for now, which is better in almost every way except that unfortunately it is not easy at all to add LaTeX equations. Also the UI looks somehow feels less native macOS-like than the XQuartz versions, they're still not even using NSOpenPanel which is a huge pet peeve for me
> Yes. This is related to GTK and is holding the macOS release back while Win/Linux graduate from beta. If you want to connect with the developers they can be found at: https://chat.inkscape.org/channel/team
So if you want to help resolve it, perhaps check there :)
The developers for the Mac build are aware of this issue and trying to determine how we can resolve it. As far as I understand it may be an upstream GTK problem, but if you'd like to know more (or even better, you have the skills to help) you can chat with them at: https://chat.inkscape.org/channel/team_devel_mac
This surprises me. I would imagine that most programmers would want to work on a unix-variant instead of Windows. But, that is probably just my bias talking.
Both MacOS and Windows are relatively annoying platform for which the "official" programming tools are not free. Additionally, there is a network effect: GTK applications, specifically, remain on Linux because GTK on Windows doesn't work too well.
The Windows SDK is free to download. Unless you meant free, like in free software, in which case, yes, there may be an ideological barrier to entry regarding Windows development.
To be honest, its much easier to port Linux software to MacOS than to Windows. Why? Because MacOS & Linux are both Unix-like OS.
And not only GTK+ apps easy to port on MacOS, but X11-based Linux apps also easy to port to MacOS. For example, take a look on AzPainter painting app.[0,1]
Native OS support and interface is important to me, and the reason I didn't jump on to Inkscape earlier.
It's also the reason I no longer use or support Audacity. I don't expect every open source project to be up to date with the latest whiz-bang OS release. But Audacity has had ten years to put together a 64-bit version, and haven't been able to pull it off. No Catalina support makes it a non-starter for me.
That’s a common theme with OSS applications, unfortunately. It’s rare to get UX engineers to help, and even more rare for anyone to lead “product direction” from a UX perspective.
So a lot of them make use of rudimentary UI toolkits, with admittedly sub-par interfaces that stink of design-by-committee, since the focus is more on the “substance” of the application than the presentation. But with a lot of these tools, design tools in particular, the presentation is such an integral part of the experience, it’s unfortunate to see it be the part that gets the least care or attention.
Not to say that this shortcoming is universal; blender version 2.50 made huge improvements to UX, and Ubuntu showed the importance of good UI in open source. But Inkscape, GIMP, openlibre and others are still notably behind the curve.
You don't need a rare "UX engineer". The issues here are much more fundamental.
On a Core i7 with 16 GB of RAM, dragging one object on an otherwise empty canvas gives me about 1 fps. You don't need any special "product direction" to know that needs fixing.
Half of the issues would be fixed by not using GTK+. Any GTK+ app on a Mac is going to feel like "you got a Linux app to run on a Mac", not "a Mac app". Maybe someday that will no longer be true, but it's been true for the past 20 years. Inkscape doesn't even look like an old Mac app -- these UI toolkits are moving in different directions.
I know that's not going to change, which is why I'm not optimistic Inkscape is ever going to be good on the Mac. My only hope is that someone will write a Mac app which uses the Inkscape engine, like Camino or IINA did in their respective categories. But good and inexpensive alternatives exist, so there's not a lot of incentive.
I'm really struggling with the multiple comments saying that inkscape has buggy UI or bad UX.
I used inkscape extensively about 5-6 years ago and had a really good experience. A couple of months ago I used the up to date version for some ad hoc work (designing a logo) to the same effect.
I think this is a great milestone for a great app. Thumbs up to the people making this possible!
Despite being amazing applications, Gimp and Inkscape receive these complaints a lot - and so did Blender until recently.
I think one category of complaints leveled against the UI/UX is these projects is that they are "fine". They are good enough to get real work done. And if you're fine with fine, then the objections seem needlessly nit-picky... and they are!
But these apps are for creative projects - you open them up to do creative work. It's certainly possible to do creative work in a "fine" environment, but it's a lot more enjoyable in an environment that sparks joy (tm).
The Blender transformation has been incredible: it went from "fine" to "sparks joy" despite the fact it's a massively complex behemoth. Sometimes I open Blender just because it makes me happy to see the UI. Open it up, extrude and scale a cube a bit, and close it. Ahh.
For me Aesprite (pixel editor), Ableton Live (music workstation), and Pico-8 (game maker) have the same feeling. It's just fun to work with the tool for anything... and double-y fun when you're working on something fun: it compounds the enjoyment.
When I first used Sketch when I was on Mac I had that feeling too. Going back to Inkscape after that was "fine". I really like Inkscape. It's my go-to vector editor, does everything I need, and I'm very grateful it exists. But I'd never open it up just to feel happy. I'd love to see it magically do a Blender.
>The Blender transformation has been incredible: it went from "fine" to "sparks joy" despite the fact it's a massively complex behemoth. Sometimes I open Blender just because it makes me happy to see the UI! Open it up, extrude and scale a cube a bit and close it. Ahh.
I usually reference Blender circa early 2000s as one of the worst UIs not intentionally designed to be bad I've encountered. I'd say it had steps more like "gah", "bleh", "usable", "ok", "fine." I haven't used it in awhile but I imagine its light-years ahead of what I recall.
GIMP as long as I can remember has been "ok." It lacked streamlined features available in Photoshop and most complaints for the UI were basically centered around "why isn't this Photoshop."
Unless there are significant workload improvements, no one likes to learn a new UI when ultimately they just want to get their current work done. Software is a tool for most people and UI designers should always keep that in mind before making drastic changes/"improvements."
The first shot compares it to 2.79, which is descended from the 2.50 redesign in 2009.
I just fired it up for the first time in a while (got a 3D printer!) and have had to relearn a bit, but I'm a fan overall.
Not shown in those screenshots, it now defaults to left-click selection! Significant muscle memory to retrain for that, but it's nice to have it match every other piece of software I've ever used.
Blenders ui is amazing. It lays out complex UIs and controls beautifully. I'd say I loved blenders interface even more polished than that of photoshop.
There's lots of interest in community asking if the in-house UI system they've created can be separated from the blender codebase so it could be its own GUI framework but IIRC it is not possible.
I'm on the opposite of the spectrum with Ableton -- I rather hate the UI and wish they would not reinvent the UI wheel so much. But then again I come from an FLStudio background and that programs has many of the same complaints leveled against it.
I'm shocked at how much negativity there is here for Inkscape - it's long been a go-to example for how great open source software can be, in my mind.
These days, I often use the (non-free, $26) Vexlio [1] software since it works better for my use-case (diagrams for use in Latex documents). But Inkscape is clearly the more fully-featured product for most users.
I think it's a love it or hate it thing. People seem to either point at it as an example of how good open source software can be, or how bad open source software can be. Unfortunately I'm in the latter camp. A huge amount of this could be the difference between using it on mac and using it on linux though
The history of FreeHand is actually quite interesting [1]. Adobe acquired it in 1994 when they bought Aldus. The FTC then ordered the product to be returned to Altsys which was subsequently acquired by Macromedia in 1995. Macromedia developed it for a number of years before being acquired by Adobe in 2005. Finally, Adobe had successfully gained control of FreeHand whereby they immediately killed it.
One wonders why the FTC did not block Adobe’s acquisition of Macromedia or at least require FreeHand to be spun off into a separate company. The market has really suffered since Adobe became dominant.
omg I'm still missing freehand.. :/
Anyway inkscape it's a nice tool. (I had used freehand, illustrator, flash.. ) and inkscape it's fine (When I need to do some vector work I know it's there) but I mostly work on Ubuntu.
I love Inkscape, but I hate the UI. What I wouldn't give to have blender-esque ui in Inkscape. That would give illustrator and Corel draw a serious run for their money.
I love Inkscape, but it definitely has some problems. I hope the developers see this and understand how grateful I am that they created a tool that I use everyday and that despite the bugs I recommend Inkscape to everyone. Awesome work by the Inkscape team!
The bugs I encounter most frequently are:
1. When I create an inkscape svgs on windows and then open it on linux the distance two between lines of text changes and I have to reformat all the text. If I go in the reverse direction I have to add additional space because lines of text overlap. It is not a font issue.
2. With large SVG files containing many elements, copy pasting a single simple element can take between 30 seconds to 2 minutes. The app just freezes. Restarting the program does not fix this. My best guess is that it is using an inefficent data structure and adding a new element to this data structure scales with the size of the number of existing elements. A similar thing happens when rescaling the width of a line.
3. Sometimes parts of an element will disappear. This happens when undoing and redoing a change or zooming in or out many times quickly. If I zoom out, wait and zoom back in the element will reappear.
4. When creating a line which ends in an arrow, the arrow symbol I used will be duplicated in the arrow selector. Some drawings I make have a few hundred arrows, meaning that if I want to change the type of arrow I'm using I need to scroll through a few hundred clones of the same arrow type.
5. Occasionally when saving it will just crash and corrupt my svg. For this reason I change the filename every few saves since this bug cost me two days of work once. The last time I encountered a corrupted was years agos, but once bitten twice shy.
6. When copy pasting an element that is a member of a group the new element will, at a point in the future, decide it is a member of the group it was a copy of. Ungrouping fixes this.
I use inkscape on Linux (Ubuntu), Windows (7/10) and OSX. It is the most stable on Linux and I rarely see issues on Linux. Despite all these problems is it my favorite drawing tool and I use it in nearly every project I have.
The conventional way to save files is to write a new file and only as the last step do a rename. This solves the crash corruption problem. Does Inkscape not do this?
No idea. One theory is that it could be the case that at some point memory corruption occurred inside the Inkscape. The save succeeded but the data that written was corrupt. Once the file was saved, the write buffer was cleared and this caused the crash.
>The conventional way to save files is to write a new file and only as the last step do a rename. This solves the crash corruption problem. Does Inkscape not do this?
I did a quick read through the Inkscape source code and it looks like Inkscape writes straight to the file but I could be wrong.
My experience with Illustrator was that it made no sense, so I went and read the first 50 pages of the (excellent) manual, and then I understood everything. It taught basic principles of the application philosophy which explained everything I'd been missing so far, and which would lead me to understand everything else I'd do in the future. I never had to touch the manual again.
My experience with Inkscape is that it made no sense (either on its own, or compared to Illustrator), and the manual wasn't very good, so I googled until I found an answer. Then something else didn't make sense, so I googled until I found an answer. There was never any point at which it got simpler. Learning one thing didn't help me learn anything else. There just wasn't enough coherence of design.
For example, I drew a round-rectangle, and then I needed to resize it. (Unlike Illustrator, there doesn't seem to be any way to set the size beforehand.) Inkscape resizes the corner radius, too, which is something I've never once wanted, so that's a strange default. In Illustrator, this is controlled by a checkbox labeled "Scale Corners". In Inkscape, it's a cryptic icon on a tiny button (manual: "a group of four toggle buttons, the second from the left"). Then, regardless of the setting of this toggle, as you resize the object, it still scales the corners during the drag -- that is, the live-resize shows what you're not going to get! The first 5 times I tried it, I immediately hit Undo because I was sure it was broken, or I missed something. I've never seen any other application do this. You have to trust that you got the settings right, and imagine what it's going to look like when you release the mouse button.
Inkscape is full of little frustrations like this. I love the concept of a free vector graphics editor, but I'll always use anything other than Inkscape, given the choice.
I'm certainly surprised to hear that. I only do occasional vector graphics work, but for me Inkscape is in the too-rare category of "open-source apps with solid UIs". Admittedly, the domain is tricky, but as somebody who used Illustrator, I got that part out of the way long ago. Gimp, on the other hand, has always felt clunky to me, despite a similar amount of experience.
I thought the UI isn't that bad until I tried Gravit Designer. I don't know if it's comparable to Inkscape in terms of features, but to me as a more occasional user it's a vast difference in usability(even considering the compulsory online account with Gravit). Maybe more proficient users would feel different but in Inkscape there are so many buttons and switches visible at all times that I constantly have to look for the right one.
But again, that's just me as someone using it every couple of months; in general I always manage to get the stuff done in Inkscape and it seems pretty solid behind the UI.
I've used Adobe Illustrator intensively for a number of years, and have attempted to use Inkscape for some similar projects, and have found it quite robust!
The biggest problem is that it is almost universal in Abode apps is space to pan and drag the canvas. Inkscape had this at one time, but then removed it. Hopefully it has come back otherwise this app will never reach the power and speed of Adobe products.
For advanced users, it's all about the keyboard shortcuts!
I took a very short look at the Mac version. They made great progress, yes, but they aren’t there yet.
The menu bar seems native now, but save dialogs aren’t, and look out of place (and, IMO, ugly).
Also, clipboard support is limited. You can’t paste copied images into TextEdit, for example (the standard image format on Mac OS is PDF, so it should put a PDF representation of the copied data on the clipboard)
Does that make it unworkable? No, but there’s room for improvement.
On a stock Fedora install of Inkscape, I can think of half a dozen particular UI actions which either lead to compounding bugs or crash the program.
I'm scared to even enter custom size attributes for objects using the X/Y text fields on the top bar because it consistently blows up the value and then brings down Inkscape.
This behavior is observed across several concurrent virtual machines with different environments.
I think the learning curve is kinda steep in the beginning. It took me years of opening Inkscape, trying to build something and closing it in frustration before I finally took the time to learn the basics properly.
I love what you can do with it, but the UX is not intuitive. After you get used to the keystrokes, everything makes sense, so it is mostly a problem for new users.
Move to Python3 and some breaking changes in the extension API along with other big changes like theming, proper high DPI support, 64 bit & native app on mac. Then the usual glut of use case feature improvements to boot.
Extension authors were asked to update their extensions to the new API, so that is one of the breaking changes.
I've updated my simple extension¹ for applying matrix transformations to create drawings in an isometric perspective accordingly. The 1.0 beta's of Inkscape gave quite adequate feedback to help transition my Python code, so it was quite straightforward to do.
Initial release was 16 years ago, long time coming.
What is Inkscape? “Inkscape is a free and open-source vector graphics editor. This software can be used to create or edit vector graphics such as illustrations, diagrams, line arts, charts, logos, icons and complex paintings. Inkscape's primary vector graphics format is Scalable Vector Graphics (SVG); however, many other formats can be imported and exported.”
Adobe is much easier to learn & use and works well with Adobe’s other 30-50 products depending on what is being done. Professionals tend to use it, so if you need to work with others, likely the best choice.
Inkscape is free, open-source, and not cloud based — and is less user friendly, has a much higher learning curve. GIMP, which is comparable to Photoshop, works okay with Inkscape, but is not comparable to the ways Adobe’s products work together.
Personally, I enjoy using Inkscape because I value the freedom it provides. Including knowing that if I create an digital asset with it and backup the build used it create it (OS, Inkscape, etc) - very likely I’ll be able to edit the file as it was when I created it; same is not likely true for Adobe products, especially over long time periods.
Highly recommend if you’re using either that you also understand the SVG file format and stick to it as much as possible.
If you have any additional more specific questions, happy to try and answer them. Love vector based graphics.
Just another perspective but I find illustrator CC to be inscrutable. Maybe older versions were more reasonable, but I definitely and wholeheartedly prefer inkscape and use it almost exclusively. The last time I tried to use illustrator it was a total mess. I could barely create a new document without a tutorial.
SVG is a specification, how it’s rendered and what’s supported vary from application to application, version to version, etc. Personally, seen some very nasty bugs related to moving from one rendering engine to another. If you’re working with professionals, I assure you few if any would want the additional over head of a teammate using another tool.
Beyond that, as I mentioned, GIMP is not comparable to Photoshop. If you’re going to be using Photoshop with vector based digital assets, not aware of any reasonable justification not to use Adobe’s vector application over Inkscape.
Beyond that, Adobe’s asset management and workflow management applications don’t compare to anything that’s available to anything that’s available within Inkscape.
If you’re aware of any counter claims, would welcome any more specific thoughts you have on topic.
I use Illustrator professionally. Some of these things may have changed since I last looked at Inkscape but I found it to be lacking in these features that are core to my workflow:
* support for CMYK work, as well as spot colors, and the whole dark art of color profiles - last I checked Inkscape was RGB-only and there are things involved in professional printing that you simply cannot do without CMYK and spot support
* global color swatches (change the swatch, everything you drew with it changes to match)
Also having 20y of Illustrator files on my hard drive that I regularly pull elements and appearance stacks out of for reuse is a compelling argument for me to stick with AI.
Inkscape was the only free, viable vector image (svg) application in Linux that I could find a few years ago. On the other hand, GIMP is the Linux go-to for raster imaging (everything else) but AFAIK GIMP only works with vector images by converting them to raster.
I find the 1.0 release candidate also completely unusable, and have filed https://gitlab.com/inkscape/inbox/issues/2307. It looks like just now it might finally be getting some attention. (I did deliberately express my opinions much more strongly than the first time.)
But until multiple critical issues are fixed, I can’t use this, and must stay on the 0.92 series.
> I think dropping all this feedback in one place is probably better for triage than me opening lots of individual bugs
You presumably know better than this. The standard, decades-old advice for submitting issues and bug reports is the opposite -- one problem per issue submission, so discussion and triaging individual issues is possible.
With most community-supported open source projects, casual drive-by braindumps are likely to be ignored. Reduce the friction to the developer as much as possible.
There comes a point when so much is broken or in need of changes, especially when most of it’s essentially in one area (in this case, the win32 GTK theme it’s using) that that traditional wisdom (which I normally follow) ceases to be true. I know that were I developing on such a thing I would genuinely prefer one big report to begin with than a large number of tiny bug reports. I might or might not then split it up, depending on how much effort things would be. In this case, I think that most of the issues could be fixed (or mitigated) by one person modifying the theme stylesheet, fairly quickly. Then there are some things that I simply don’t know whether they’re the same underlying thing or not, and splitting them out if they’re the same root cause is a nuisance when you’re a developer—you just end up needing to coalesce them again.
Considering that the small group of developers are all volunteers, I think you owe it to them to write your bug report in a way that is actually usable by them. One way would be creating one bigger issue with your suspected root cause (the win32 GTK theme) and a long, enumerated list of the specific issues, their reproduction, etc. (including screenshots). Ideally, each item on the list links to a new issue that follows the standard model for bug reports. Reacting as strongly as you did in the issue thread will definitely not get anything fixed.
Please consider the human and, here, the limited resources of the developers.
> There comes a point when so much is broken or in need of changes, especially when most of it’s essentially in one area (in this case, the win32 GTK theme it’s using) that that traditional wisdom (which I normally follow) ceases to be true.
I think it's even more true for cognitive behavioural reasons. A volunteer-driven project in particularly works best with small, easily digestible problems that can be individually doled out to (or accepted by) people with limited time. It's beneficial because it shows progress to participants and issue submitters.
I just read the second report, both the rambling source and your irredeemably rude response to the volunteer developer who chose to respond to your nonsense. You should be ashamed of yourself.
The problem with Inkscape, as well as other open source gui apps, is that there's nobody to lead a holistic UX QA process, and so they're full of enough bugs, edge cases, and general UX fail, that it's a chore to use the app, and you're constantly worried that you'll hit something big enough that you can't complete your work. I've had this happen often enough with libreoffice that I've stopped using it.
I really hate this statement. You're asserting that because the project doesn't have somebody in a dedicated role they have bugs, unspecified issues, and subjective "problems". Not only is it a logical fallacy, it asserts nebulous "faults" that can't be addressed because they're so ill defined.
None of what you're asserting is true of Inkscape. I've been using it for 10 years and while the UI/UX isn't familiar, there's nothing wrong with it. It isn't chock full of bugs or "edge case" (whatever the hell that means) and like all UI/UX it has rough edges AND that's no different from any other sufficiently complex Application.
Does it follow Adobe's UI/UX paradigm? No. Is that bad? No. If you started out using an Adobe competitor in the 90s then you might be familiar with alternatives, that's where I started and I can't stand the UI/UX in Adobe products to this day.
Now, I agree that Inkscape has a fairly solid UI as far as OS software goes.
BUT, in general, UX and interaction design simply is not a reductionist process the way implementing functional requirements is. It is an inherently holistic endeavor, an art more than a science, and that is why it helps tremendously to have a single competent person in charge who has a vision of how things should work.
I understand that to the typical programmer mindset it is frustrating when people come and complain that the UI is clunky and unintuitive without being able to express their frustration as a list of actionable "this does A but it should do B" tickets. But that's just how it works! Understanding and empathizing with the users simply requires a different skillset than programming.
I don't think anyone disagrees that the interface is less than perfect for them. And there is no one ideal user interface for everyone.
That's how software goes, it's a conglomeration of all the features for all of it's myriad of uses. What would benefit one use case might be detrimental for another, and no one UI/UX developer is going to be able to fathom all of those different use cases or conceptualize a UI that's ideally suited for them all.
What you do often get with dedicated resources is a direction drive and clarity of vision that is not in line with what everyone wants and their UI/UX ends up driving people away.
I don't use half of what Inkscape has to offer and I use Inkscape for two very distinct and very different use cases. And for both of those use cases Inkscape is far and away the best piece of software.
For web development, Inkscape is my go to SVG editor. I don't use it as a vector drawing tool at all, if the drawing and filter capabilities went away it wouldn't matter much to me. But the ability to quickly manipulate canvas sizes, add and remove objects, and tweak paths is essential.
When I use Inkscape as part of a CAD/CAM solution engraving, the use case and what UI/UX elements are important is completely different than when I'm optimizing SVG Assets for the web. Certain UI/UX elements I require as a web developer could vanish and I wouldn't care as a CNC operator.
I'm reasonably certain that much of what I value in Inkscape isn't of much use to someone who is producing vector art.
That's the nature of most mature sufficiently complex software. They don't start out as a set of user requirements cleanly laid out that are developed in a straight shot waterfall process.
Instead they grow organically over time and are a reflection not just how things are implemented but also of when they were implemented.
The difference is it with software that has product management, it can grow and shrink. Features can be dropped, deprecated, or re-worked into a more appropriate presentation. With OSS it seems like everything is additive, since you have no idea what users are using what features.
The blender 2.50 release was awful for me, since it broke so many of my workflows, but the UI absolutely needed a revamp, and I have incredible respect for the team throwing so much away to make something better.
I wish more open source software would delete more things in order to provide a better or more thought out user experience.
I’ll also note that telling users that their sub-par user experience doesn’t mean that the application has bad UX is foolish. That’s what user experience is. And it’s not a unique position; look at the comments here. The number one complaint is UX. These people ARE the users. I can’t imagine more direct feedback than that.
I’d love to see a case study of how Blender managed its massive 2.8 UX overhaul. It’s the only major open-source application I know of that’s pulled something like that off.
2.49 was the last one with the really inscrutable old UI, 2.5 was a big overhaul but still had a lot of basic functionality buried behind hotkeys or menus, and 2.8 is the relatively recent one with a lot of that basic functionality exposed.
I don't follow the blender development process too closely, but I think Blender Guru's videos kicked off a lot of the 2.8 design work: https://www.youtube.com/watch?v=UWacQrEcMHk
From my extremely limited involvement in the community it seems like it was started by one artist publishing a very dreamy vision of "this is what the future of Blender could look like" to his followers. It seems like people liked it enough that a lot of it was eventually incorporated. It might have also started a wave of enthusiasm for more people to become involved.
"I've had this happen often enough with libreoffice that I've stopped using it."
But did you had it with inkscape?
Also, what was the issue with libreoffice that blocked you from completing your work?
Because, I can imagine many issues with libreoffice, but mainly with microsoft office documents, importing them etc. but experienced no issues with the actual program - except it is horrible from a designers point of view, but I also despise microsoft office. So if I had to designe a page, I usually used - Inkscape. Which I like a lot more than Illustrator.
- Style settings that would permanently break when changed (changing the style back to what it was before would not fix it).
- TOC links that stopped working
- TOC items that disappeared
- Inconsistent alignment (where the same style would be off by a few pixels on different pages)
- Inconsistent page headers (where some elements would not show, depending on how many pages you have)
- Scripting that yields different visual results each time you run it (this is where I gave up)
- Disappearing cells
- Mysterious blank pages that you can't delete without breaking everything else using the same page style
- Inconsistent results when printing (what you print is not what the document shows)
This is using purely native document formats, not cross-format. Just before giving up, I downloaded the source code to see what was breaking (in the scripting). And then I understood.
I can't remember what issues I had with Inkscape since it was awhile ago. I only remember being aggravated enough to invest time in learning a replacement tool and redoing all my work.
Gimp has tons of UI bugs as well (rectangle select gets stuck in subtract mode, for example), but so far nothing quite bad enough that it can't be fixed by saving, restarting, reloading.
"Scripting that yields different visual results each time you run it (this is where I gave up)"
Ah yeah, I remember, this is why I stopped digging more into it, after weird starting results.
And I remembered my main issue with inkscape: text editing.
I would actually consider it broken beyond repair at the current state. Which is sad, as otherwise I am really happy with inkscape.
But I also very clearly remember the statement from the main dev from inkscape, when people ask him, if the next version of inkscape could have feature xyz:
he then just replies with a one liner, his bank account number.
Please provide a link to the message in question. It sounds serious.
The main problem with your story is that there is no "main dev from inkscape". There's a rag tag group of fairly junior developers and amazing support volunteers.
But, I could see how the story could get twisted over time. I'm pretty sure I've never shared my bank account details and I'm the worst bastard at Inkscape for asking users to contribute, all users, in any way they can. This is because contributions are lifeblood, not users. Adobe is great because 90% of their users are contributors where as 0.01% of Inkscape users ever contribute.
So consider contributing your passion about Inkscape at chat.inkscape.org there's a ux team and everything.
"Please provide a link to the message in question. It sounds serious."
I did not follow development of inkcape and the message I mean, was from around 10+ years ago .. in an interview with whom it sounded like the main dev at the time(don't remember the name). And the statement was a bit hyperbole as well, to get across the point, "you want a lot of things, so send a lot of money, so I can hire lots of people to make them happen, otherwise be patient"
So no worries, no one today is (afaik) abusing inkscape ..
My point isn't that these bugs exist, or even the maintainers' willingness, time, or ability to fix them.
My point is that people don't understand the necessity for a "vision guy" to direct UX in any program of appreciable size. And it's not like these kinds of people don't exist, or that they'd be unwilling to help. People just don't seem to even WANT them taking on these tasks on their projects.
And that's a real shame, because many of these projects are technical wonders that are borderline unusable for want of a loving touch.
I get your point, but don't you think it is somewhat complicated?
When you are the main dev in such a project, would you be willing to give this huge responsibility to some guy who might loose interest after a few weeks? What then?
Also, I have seen people doing graphical sketches of possible UI's and people think nice - but it is a whole different thing to actually implement this (especially with so little money involved), when the to do list involves 10 million other things.
Also, usually the problem is the designers are not coders and quite often the code base is not designed in a way, where you can just swap out the UI for a new one.
But yes, I'd love to see more collaboration of designers with a vision and coders.
What was your point, with "he then just replies with a one liner, his bank account number"?
That seems a reasonable way to support additional features in an open source project, especially if the volunteer developers have no personal interest in the feature.
A one-line comment would be a bit brief. "This feature isn't planned, but one of the core developers would be willing to take on the work for payment. If that's interesting to you, we will estimate the cost."
"What was your point, with "he then just replies with a one liner, his bank account number"?"
That I am not criticizing the inkscape devs, for having implemented such a bad text support. They had other priorities and people like me, who would have liked it, did not pay enough money, to have it implemented. Simple as that. So the other comment was not neccesary.
I've used both LibreOffice and Inkscape enough to know that Inkscape is very capable of doing complex and detailed work. If you have a design process then you can make Inkscape work into that fairly easily. I know UI designers that are not able to take advantage of Inkscape's freedom however that speaks more to their innate curiosity and creativity than Inkscape's capabilities.
Of course, as soon as you go outside that, you run into the fragmentation prevalent in the Linux desktop. For light computer users or people who can get by without a bunch of specialized graphical apps, it works well and is very intuitive.
Are you claiming this is true about Inkscape? I agree with you on Openoffice/Libreoffice (although we disagree about the cause). But Inkscape has always struck me an exception.
If you are saying it has a bad UI, could you say more about what you tried to use it for, and what other vector-graphics program you've spent a lot of time in?
I always found Inkscape next to unusable compared with Corel Draw, Illustrator and Affinity Designer. Tried it several times, always gave up. In v0.92 (I know, not latest version) if I moved a point near the end of a path, points near the other end of the path suddenly started moving. Totally weird. I do not know how people can do serious work with this.
Off topic for this conversation, but I agree with you on LibreOffice. The only reason I use it is because I want to use an open data format that’s more likely to exist for a longer time. I hate its ’90s style UI, the UX, the fonts, layout, menus, keyboard shortcuts that are very different from the more popular proprietary alternatives, etc. The worst of this is that it doesn’t have a working copy paste I can rely on. It seems to have its own clipboard that dislikes the OS clipboard. All these experiences make it impossible to sell it to someone who doesn’t care much about proprietary vs. open.
P.S.: I donate money periodically to The Document Foundation, the organization behind LibreOffice.
- Touch pad and touch screen pinch to zoom (YES!!!)
- The canvas can be rotated and flipped (probably great for hand drawing, would be way more useful if possible by touch)
- GTK 3
- Pressure sensitive pencil
- Interactivity for UI mockups
- Autosave is on by default
- Plotter extension automatically converts things to paths
- Unified raster tracing
- Better Mac OS integration
While the UI has apparently been polished a whole bunch, I think most of my other comment moaning about it still applies - and that's really sad, especially when you want to market this as a mile stone release everyone should try. You'll still end up turning off a lot of people, I'm afraid.
I discovered this one by accident in the 1.0 beta. I still don't know which shortcut did that, but I definitely like having that option. Great when you are making woodworking plans and want to work on a section of the drawing where some parts are not aligned horizontally or vertically.
> - Touch pad and touch screen pinch to zoom (YES!!!)
It seems like there is still some work to do on this. Trying to use the pinch gesture crashes Inkscape 1.0rc1 (09960d6, 2020-04-09) instantly and reliably on Mojave.
Pressure sensitivity was only working on the caligraphy tool, not on the pencil! This is an huge improvement - The rendering of the pencil tool is way faster for me and i suppose pencil lines are plain paths with an strokeWith, so the resulting svg would have way less complextion on the path data, i guess. Not checked yet. :- )
Inkscape helped me bring several projects to life, thank you to all devs. The last boardgame I released was completely designed with the previous version of Inkscape
Thanks Microsoft...
At least you can override this.
Would be helpful tho if you actually "smartly" "screened" stuff instead of resorting to the "this file is new/unknown and thus potentially dangerous" metric.
I've used inkscape regularly for large print scientific posters for many years. It definitely used to struggle in that regard.
It's gotten significantly better about handling high resolution figures and using many layers in recent years. At least on Linux, it's quite capable of handling complex files, though deeply nested layers are fundamentally a bit clunky in the UI. (Inkscape is more focused on using groups the way layers are sometimes used in Illustrator. Layers are really just a named group under the hood, if that makes sense. At any rate, it's often best to rethink your workflows a bit if you're used to having illustator files with several hundred nested layers. The UI in Inkscape semi-deliberately discourages that workflow, while Illustrator deliberately encourages it.)
Printing calibrated colors isn't it's strongest suit -- I'd recommending exporting and using something else for the actual final print. However, that may have changed too. I haven't tried recently -- I've mostly taken the "export to PDF and tweak the profile when printing the PDF" approach.
I recently had a major struggle with it's inability to export to CMYK. I eventually found a way to get my graphics into CMYK by exporting to PDF plus an assortment or ad-hoc scripts for conversion. That seems like a real deal-breaker for most professional print applications. I still love Inkscape though.
Inability to export to CMYK seems like a huge drawback for the printing industry. RGB colours cannot be represented correctly in print. Being able to work in and export to CMYK is pretty much necessary for any kind of professional printing work.
When necessary, my trick has been to export inkscape graphics to Scribus when print colour support is needed. Scribus has extremely good spot/CMYK colour support, but a rather weird UI (I'm told it's modelled on QuarkXPress).
I was reflecting on the Open Source UI debate recently because I was helping a family member do some document layout stuff in Scribus. Certainly these tools can have a daunting UI but the existence of these free alternatives (Blender, GIMP, Scribus, Inkscape, etc) is so important, especially as the existing tools in these ecosystems move towards an even more unaffordable subscription pricing model.
The ability for people without access to either pirated copies or student subscriptions to start playing around with, for example, Blender as a way to get started in 3D modelling, is a huge benefit to the world.
And it was ultimately easier to do what we needed in Scribus than in Word so I think there's an extent to which we have Stockholm syndrome for existing UI.
This is quite exciting! I have started using Inkscape to draw diagrams for use with LaTeX a few years ago and I like it a lot. (There is an option to export PDF+LaTeX with allows you inline code.) For more precise stuff which requires programmatic handling (e.g. algorithm visualisations) I can just write out an initial SVG and edit that by hand afterwards. Maybe now is a good time to look into some customisation options, it seems there are lots of them with the 1.0 update.
For discussion purposes, moving from Beta to RC doesn't make a new topic, i.e. the discussion tends to be substantially the same.
Edit: Even when a project goes from one version to another, there's rarely anything new in the thread about the new version, because the comments are nearly always about the project in general [1]. For example, in the present case, you'll find a few comments about XQuartz being removed (something which sounds like a major new development), but the vast majority are just about Inkscape in general. There's nothing wrong with that, but we need to limit the number of times it happens, or HN's front page would mostly be discussing the same projects over and over. Obviously there are exceptions in unusual cases, but when it's the same project, with the same version, going through two different statuses (like Beta to RC or actual release), you will be hard pressed to find the slightest difference in the threads.
That makes this post a duplicate by HN's rule (see [2]). I won't bury this one, but when this comes up again in a few weeks, let's not have a third thread.
Nothing against this project, of course. It's just that front page space is the scarcest resource HN has [3].
I wish you were right, because it would be a lot less work. But herd dynamics alone don't produce anything like HN's front page. HN is actively curated and always has been. The site is a set of feedback loops between the elements of its system: community, software, and moderation.
The reason why unregulated hivemind behavior composes to a dumb and unstable outcome is not that the people contributing are dumb. It's that they're only using a small part of their attention. If everyone here would deeply reflect before upvoting, flagging, posting, etc., about whether that action making the site more interesting in the long run, we'd get different results. But people don't put that level of attention into what they're doing here, nor should they—it's a lot of work, and everyone has more important things to worry about. What that means, though, is that the hivemind is mostly acting reflexively rather than reflectively (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...), which is bad for a site like this.
The solution is to have a small number of users (i.e. the mods) specialize in worrying about that, i.e. putting full attention into the site as a whole, and reflecting on it and regulating it (or helping it self-regulate) at that level. Then you get complex feedback loops between the big system (the community: large population, sporadic attention) and the little system (the moderators: tiny population, focused attention), which makes the site more unpredictable and interesting. Edit: software is also a critical part of this.
To have paid staff manually micro-regulating like this isn't what I call a solution. You're simply prodding the herd with a stick, and your message is virtually drowning amongst the other ones.
You cannot be imperative, top-down to a community and expect that to work.
"For discussion purposes, moving from Beta to RC doesn't make a new topic" <-- kinda trying to convince
"For the purposes of HN threads, there's no difference" <-- almost resorting to authority
That tone and kind of discourse is ... unfortunate.
It's similar to the Wikipedia principle, don't say something is amazing, explain WHY it's amazing. Then you will perhaps convince 10s or 100s of people that will not enforce your rules but perhaps convince other members of the same. I just hope your rule is right :)
In other words, a community cannot be sustained by being regulated by privileged people, poking them here and there. A community needs to regulate itself and their individuals need to grow themselves.
By the way, such a (humble) message would be worthy of being colored in a distinct way to give it more priority.
But given the abysmal lack of development (and also the current state of features, ew) on the HN site, I doubt that such a feature would ever emerge. In fact, the only thing that makes HN worth visiting is its users. The website itself is horrific.
I had no idea it is also actively maintained and reaching stable, I hope vector graphics make a big comeback in many more construction-art kind of ways than just mobile apps or webverts.
They finally updated the UX. Though the only thing I notice is my theme sticks better (could be a matter having gtkrc set locally)
Anything related to inkscape and gimp being powered up and having the functionality of Illustrator and Photoshop (without the bloat and background services) is very openly welcomed.
The stuff that I use something like inkspace for (Pen tool, SVGs) and Photoshop (very very basic photo editing, resizing, exporting) doesn't need a huge hulking application.
What drives the development of these apps? I'd guess Adobe making the applications heavier/bulkier over years, and pushing a cloud model where an app tries to become a bigger part of daily life than 9/10 people want it to be.
My guess is most Illustrator users like the app, but don't want a daemon running in the tray 24-7. And why invest in saving files in Adobe when there's solutions that work for all files, (and it may require a local network filesystems for sizes that big, cloud won't be practical)
I am an incredibly huge fan of Inkscape and the monumental efforts put forth by the developers of this amazing piece of software. I cannot do half of the things I've done in Inkscape that I've tried to do in Visio.
Are the tons of glitches and incoherent UI fixed? I recently tried Inkscape again (on arch, so I assume a relatively recent version) and it was still pretty bad (although better than a few years ago) :s
Snapping is a nightmare. Layers are kind of hard to work with. Elements don't keep their ratio by default. The way the canvas border works by default makes no sense. The UI is just not sensibly arranged in many places.
Since they talk about crashes I hope at least the spontantous crashes while copy/pasting are gone now. A nice bonus would be touchpad zooming (at least moving around now works smoothly).
It's still the best FOSS vector editor out there, but 1.0 is a big release. I hope it now has the polish it deserves.
Inkscape snapping had always worked well for me. I love the fine-grained control available on the snapping panel. I do find the layering / grouping controls confusing.
I am not familiar with the package of arch, but changes could have been done. When in doubt and need, try an vanilla version from source.
You can adapt the layout in the preferences.
> Since they talk about crashes I hope at least the spontantous crashes while copy/pasting are gone now.
Yes, copy-pasting had issues, but if you adapt to duplicating instead almost all problems will be gone.
> A nice bonus would be touchpad zooming (at least moving around now works smoothly).
Touchpad zooming will be in the 1.0 release, called pinch-to-zoom
When I am using an FOSS tool, i try to learn it as an tool, meaning that I want to understand the reasoning for some work flows. By learning the tool I want to show my appreciation to the devs.
On mouseover you will see that the '%' sign enables/disables snapping, which is almost always the desired behaviour.
Even with elimination of XQuartz, there is still little rationale to use Inkscape, if you are actually trying to be productive towards a result.
Just pick up Affinity Designer.. One time purchase... awesome UX, on par with Illustrator. For simpler, quicker result, Keynote has awesome diagram tools.
It has a standard Mac-style menu bar (rather than a menu bar within the window). Keyboard shortcuts now use the command (⌘) key rather than the control key. Retina display screen resolution is now supported. The build is now cleanly 64-bit, a prerequisite for macOS Catalina 10.15 and beyond. It comes bundled with Python 3 to power Inkscape extensions. "