Comment by BadBadJellyBean 2 days ago
I think especially since the UI overhaul in Blender 2.8 the project has been on a steep upwards trajectory. The software was always amazing, especially since it was free and open source, but the new UI and all subsequent improvements really put Blender on the map as a serious tool and not just an alternative for when you don't have money for the big players.
> but are awful at UX
This is such a weird trope.
For those of us who've used microsoft teams, jira, servicenow, salesforce, or basically any insanely popular (in the commercial if not upvote sense) products, it's unclear what is being compared to with these tired claims.
"Bad" comes in many shapes and sizes. Specifically, "technically competent person implementing a thing designed by a technically incompetent person" is remarkably different from "technically incompetent person implementing a thing designed by a technically competent person".
The way this plays out in practice is that those products you listed can hire actual UX designers, but many product decisions are made by people focusing on business concerns rather than product concerns, so you have competent people implementing designs by incompetent people.
Inversely, because open source software is usually built by people trying to scratch their own itches, they those people actually understand what the product should be, but, because they're usually software engineers instead of UX designers, they're typically incompetent at UX design. So you have incompetent people (devs with their UX design hat on) implementing designs by competent people (those same devs, with their "scratch my own itch" product owner hat on)
> This is such a weird trope.
No, it isn't. Lots of non-trivial OSS desktop applications are clearly made by people with no interest in aligning with expected desktop GUI behavior. From Gimp with dozens of windows to LibreOffice which is slow and has bad font rendering. And those are the 'poster apps' for FOSS desktops, lots of apps are worse.
Gimp's single window mode was made the default years ago now, so that's not a great example anymore - there's scientific software that uses that paradigm that might work better, but most of that isn't OSS. Also, Libreoffice being slow and having bad font rendering seems pretty inline with Word nowadays...
These are all products the ux direction of which is likely influenced more by corporate power dynamics. Sure, uxers are involved, the real power they have is a different question.
Everyone’s got their preferences, quality of ux is by definition subjective. That is what makes these discussions hard. Naming any examples will always have ”nah i don’t like that product” as counterpoint.
An equally weird trope us UX practitioners dumbing down UIs. It simply depends on who we are designing for.
As soon as developers actively hang out with real users in real life and genuinely observe them without intervening, i’m all for oss projects without uxers.
Disclaimer: did my master’s thesis on OSS UX.
Shameless plug: User Experience Design in Open Source: Inviting the Users
https://savolai.net/ux/user-experience-design-in-open-source...
Product & framework thinkers: Case studies.
https://savolai.net/ux/product-and-framework-thinkers-when-d...
I’m definitely going to read those, but even without doing so “inviting the users” as a concept carries a lot of potential. We were tasked to rewrite a very old windows app for backend grocery store sales in a web/Laravel/Vue application, and product spent _months_ if not longer sitting with sales reps, watching them use the old tool, and asking them what they would want to see - how does it work? Can it be more efficient? What do you dread most when using this?
The end result was a real pleasure both to write and to use.
>As soon as developers actively hang out with real users in real life and genuinely observe them without intervening, i’m all for oss projects without uxers.
Game dev here. Play tests are excruciatingly painful. Spend some time showing off a game and you can see why so much ux these days are "boring" and samey. Deviating off the beaten road takes so much extra polish compared to seeing how competition controls work and copying that.
Microsoft Teams was bad, so they rebuilt it and somehow made it worse. Then they decided to do the same with other apps, like Notepad. I switched to Ubuntu on my computer this week. Linux administration is not something I want to spend time on, but LLMs are able to help me debug why my password manager can't talk to my browser and write shell scripts to fix it... I'm able to focus on work and be done with the Microslop.
> microsoft teams, jira, servicenow, salesforce
Nobody wants to use those products either; they just exist because their default at a certain scale, or they're effectively free because they're included in your existing MS license.
For GIMP the comparison would be either Adobe stuff or what used to be Affinity products. Libreoffice is now competing maybe with MS Word but probably more often Google Docs or Markdown editors.
Old blender used to have a very technical UI; a cacophony of dropdowns and small text that functioned but was quite overwhelming. Meanwhile things like SketchUp became popular because they weren't as powerful necessarily, but were very welcoming, and that's hard to do with a complex offering.
I think you misread and assumed this was a comparison to something else. It’s not.
Actually, I like Microsoft Teams.
I know this is controversial but I prefer teams to zoom and slack.
Usually being compared with their non-OOS alternatives, not random enterprise org software.
Teams are decent, wdym?
Inb4: I've used ventrilo,team speak, mumble, discord, Skype.
It looks like you only use a tiny fraction of Teams' functionality. I agree, there's little to complain about when using it for IM/voice/video calls. When you start using it for other things, especially the enterprise features, it is bad. It is a resource hog, handles navigation poorly, has poor default settings, finding installed apps can be tough, etc.
> handles navigation poorly
My current pet peeve: I’m often going back to the previous week on Monday to fill out my time sheet. So, I open the chat for a meeting last week to see how long it took, fill it out, and hit the calendar icon in teams and I’m back on the current week. It’s a painful UX flow that I’ve now built in to my brain, so help me god if they fix it.
Note that teams does include a “back” button, and also note that it doesn’t give a flip about state - it knows you were just at the calendar but doesn’t care where, so you’re back on the current week
Lots of that is momentum and politicking. Or the result of decades of concerted effort to associate your product with it's niche, from education to industry, like Adobe
Those products likely have UI / UX people behind how they look, feel and behave. ;) Except maybe Jira, Jiras always been the Excel of ticketing.
>it's unclear what is being compared to with these tired claims.
Relatively good UX. Because Microsoft, Salesforce, etc. Have full time teams of designers in tow. For historical reasons it's harder to get said designers to work on FLOSS.
Getting good UX requires professional designers, extensive human testing, and knowledge of human psychology—things historically in short supply among the OSS geek set. In the 1980s Apple ran a human factors lab that spent thousands of hours determining which interface features were the easiest to use and most efficient for many common computing tasks. This is why classic Mac OS is still the gold standard for UX. Even Mac OS X started making compromises to accommodate techie trends, rather than keeping the focus on the average user.
Because much proprietary software has garbo UX, that doesn't make the OSS UX situation not garbo.
I think the blender secret sauce is their artistic projects.
Put a bunch of artists in the same room as the developers and have them produce a work.
It ferments this amazing combination of aggressive QA testing(the artists) and top tier technical support(the developers) while focusing on real problems(the work) that really brings out the best possible product. The GIMP project would probably be better off if they invested in a couple of rounds of this.
I think blender always had an amazing, ahead of it's time interface. It did lean overly hard on knowing the hot keys, probably a product of it being an in house tool but the opensource versions have worn a lot of those rough edges off(menus to provide clues) while keeping that same super smooth workflow core.
>It ferments this amazing combination of aggressive QA testing(the artists) and top tier technical support(the developers) while focusing on real problems(the work) that really brings out the best possible product.
Devs stil run the show, and it can be a huge effort to convince them to change course on something. People were complaining about the mouse controls for years and it took until 2.8 to finally convince the org otherwise to adapt a more typical workflow.
It's a rough balance because the other extreme is artists making unrealistic demands based on how the system is architected (or worse, the deceptively simole: https://xkcd.com/1425/). So I don't really have a solution.
Part of what makes this so much of an issue is that in FOSS projects, the things that get worked on tend to either be low-hanging fruit and/or a personal peeve of one of the engineers. Everything else is at high risk of falling through the cracks and being ignored or forgotten.
It’s kind of the open source counterpart of how in proprietary software, some types of bugs tend to get perpetually kicked down the road to make room for development of features that are perceived to be of higher likelihood of increasing revenue.
In theory, FOSS projects have more agency to correct this class of problem than their proprietary analogues do because they’re not subject to the same economic pressures. This however requires leadership with a strong vision for the project and soft skills to unify and motivate contributors to work on not-so-sexy bits, and this type of individual is rare in that space.
> But FLOSS software is mainly made by developers. Who like writing new flashy features, but are awful at UX, and making sure all the small kinks are worked out.
That is what product managers are for; someone to lead the product's direction, ensure quality control, and to instill taste. That requires being able to say when a feature is poorly implemented or outright bad and unnecessary -- it's not always just kinks. The problem is that this collides with the collaborative ethos of open source software. But when it's not done it's the users who suffer.
In foss the selector is the merging maintainer. Which is more a gardener than an architect. You need somebody with a vision for a park.
I think it's an issue of "what matters".
FLOSS software is often made people who are interested in the thing being done. The UI to do it is something that can be fixed "later". But later is always later. There's always another feature to implement before you can sit down and really fix that UI.
And then by the time they do get around to fixing the UI it seems the codebase is horribly bloated and littered with tech debt and now updating the UI would basically require a whole application rewrite. Which I have seen happen and work, but I also swear I've seen where teams spread themselves thin trying to make an updated UI version concurrently with their main branch only for the updated UI version to fall so far behind on features (or get worked on so rarely) that they abandoned it to fix it later...
The problem is knowing what to fund. It's easier if the users would pay. Which is doable for commercial use.
To paraphrase a quote from long ago:
"Public funding doesn't get you great coders, it gets you coders who are great at filling out government forms."
Getting paid to deliver a software product that someone wants advances humanity. Getting paid to make your own personal project provides jobs for politician's cousins.
Yes, because it isn't. What version they pump out has nothing to do with viability. It still lacks much of the feature set and functionally of Photoshop, enough so that it's just not a valid comparison, let alone a valid alternative. Like, I've seen people comparing Krita as the better alternative to GIMP, which isn't even designed to fill the same need. But the fact that those comparisons have even occurred should tell you enough about the state of GIMP.
It might sound weird, but I think the key factor is the rise of Youtube.
There is unbelievable amount of Blender content on Youtube. Like, probably more than all the other DCCs (Maya, 3DsMax, Houdini, Modo, etc...) combined[0]. It's beyond the DCC for hobbyists. I've seen people who think it's the only DCC. A few years ago, I met an 2D artist who started integrating 3D workflow and he genuinely didn't know the existence of Maya.
[0] I have no data to back this up. It's just my guess.
The people that use Maya have been using Maya since the days of it being the only thing available. When you have students as young as junior high school getting active in computers and graphics, they have no money. Using myself as an example, I was a frequent user of newsgroups like a.b.m.a. to find the software to learn how to use. Now that I'm a "professional" by earning money with software I "borrowed" while learning, I now pay for all of it.
Now that tools like Blender, Resolve, etc are all available for FREE, it's a no brainer why the younger folks entering into the scene are using them. Hook them while they are young, they'll use it for life. On top of that you can add any converts, once you have a features worthy, as everyone likes free. With places like Reddit and YouTube, you can even forgo support and crowdsource it.
For the same reason Maya & friends offer heavily discounted (or even free) educational licenses: people tend to stick with what they initially learned.
That approach works great when someone's first experience is in a traditional education system, but these days any interested kid will start to explore the options way before that - and all those self-taught hobbyist Youtube teachers haven't been offered free licenses to make content either!
So now you've got a decent pool of enthusiastic kids flowing into education with pre-existing Blender knowledge. And Blender is good enough for educational purposes, so as long as it doesn't significantly hamper the students post-education it is very attractive for educators to adopt as well - why bother with all the hassle of getting educational licenses when you can just download it for free?
The second Blender started to get industry adoption it was basically over for Maya. They could've saved it by turning it into a freemium product which hobbyists can download and install as easy as Blender, but it's probably too late for that now.
> The second Blender started to get industry adoption it was basically over for Maya. They could've saved it by turning it into a freemium product which hobbyists can download and install as easy as Blender, but it's probably too late for that now.
I was with you up until this part. I was a DVD programmer for years on the de facto standard software which was very expensive and cumbersome for some people to use. Then Apple released DVD Studio Pro which did a lot of abstraction layer simplification for people, but also meant that something could not be done in that software that could be on the big boy package. I know it's an apple/orange situation, but there are times where there's a reason for the cheaper/free price tag. I haven't paid attention to the capabilities of Maya/Blender for nearly a decade, but I'm guessing there are still things that can be done in Maya that cannot be done in Blender.
>I think especially since the UI overhaul in Blender 2.8 the project has been on a steep upwards trajectory.
100% agreed. I know a lot of people don't like that but sometimes I feel that FOSS projects are intentionally sabotaging themselves by ignoring industry standard options/conventions and instead they are following open source ideas just to be different. UI/UX is the main symptom of that. Blender could move forward and wish others could too.
Krita is another example of a good project
CAD is the next frontier where we need a "Blender moment"
We have to keep in mind though that many open source projects started as something that someone wanted and then made. It probably worked just like that person wanted and then it grew. Maybe it is because they weren't too versed in UI/UX design.
Another thing is that many classic open source projects don't have a "I want to grow my user base" mindset. Why would they. It's not like they get paid.
Big overhauls also always have the risk of alienating current users. I learned Blender on the pre 2.8 UI and because I use it rarely I still sometimes struggle with the new shortcuts.
Blender clearly benefited from the change but the real spirit of open source is: you don't like it then help fix it.
There always seems to be an incompatibility between the people who made it, the people who use it, and the people who want to contribute. The latter two often try, but the former isn't interested in the help or has a very specific vision for the project and doesn't allow any input that isn't in line with that even if it's not in conflict.
It's hard to fault anyone in that triad 100%. Open source has a way of becoming infrastructure. People come to depend on tools made by people without the resources, interest, or personality to run an infrastructure project, or who won't budge on their vision to allow contributions outside of it that might help get the project to a point where it can attract enough vision-aligned contributors.
Forking potentially shifts the problem to a new triad, so it's not an obvious solution in all cases.
> the former isn't interested in the help or has a very specific vision for the project and doesn't allow any input that isn't in line with that
I've come to call this "fenceware": technically open source due to its licensing, but community-wise it is as if the developers just throw a ball of code over the fence every few months. Sure, they let you play with it for a bit, but it is not yours to co-own.
The problem with (3D) CAD I've heard is that the Open CASCADE CAD kernel is a huge mess. So as much as they update and fix FreeCAD (and they've made a lot of good progress, but it's still very rough around the edges) they're always going to be hampered by that. And making a new CAD kernel is a massive undertaking.
While this is definitely true, for a long time FreeCAD hasn't exactly made it a high priority to properly work around that.
For example, the Topological Naming Problem (as I understand it) is made quite bad due to OpenCASCADE design - but as we've seen with 0.19 and later it is possible in a lot of cases to work around that. But that's a lot of really hard work with relatively little reward, so for years it languished on the backlog, and users had to deal with even trivial designs randomly blowing up in their face for no clear reason.
The result is a CAD program filled with footguns. Nobody wants to address structural issues, so you just pretend they don't exist, hide a half-baked tutorial on a Wiki on how to work around the worst of them, and blame the user for holding it wrong.
Commercial applications can solve this by shoveling copious amounts of cash at any skilled developer who is able to make any real-world improvement - even when it's not a perfect solution yet. FLOSS applications have to wait for a developer to come around who is masochistic enough to tackle it for free.
Question for someone who is very far away from this kind of development - why does CAD software need a kernel that’s wholly separate from the UI? Why aren’t they the same thing? I just don’t understand the abstraction that necessitates writing the software this way.
It is much like a game might use a physics engine, or a new language might use the LLVM backend. To overly simplify, a CAD kernel will keep a list of operations (make a cube of this size here, drill a hole of this depth here, round these edges but not those). And combine that into a final volume. These responsibilities only get more and more complex as a part gets more complex - so using a pre-built engine allows CAD software to focus on tools and workflows to translate human instructions into lower-layer kernel geometry: the UI/UX. It also crosses into compatibility, if you use the same Kernel as another CAD it is much simpler to export/import from them. Otherwise you would have to reimplement their kernel (or enough of it), or be stuck exporting triangulated versions of the final volume - sort of like converting an image from vector to raster.
Same reason a browser uses a separate library for image decoding, or font rendering. A CAD kernel is a very complicated piece of heavily specialized math. The UI itself is there to let the user construct the input data for the CAD kernel and to display the resulting output. Doing that translation in a user-friendly way is already hard enough without having the kernel smeared out all over the rest of the application.
The Blender project is the model I hope FreeCAD can eventually follow. Like digital animation, the 3D digital design field has a pretty rough selection of tools and the UI on all of them leaves a lot to be desired. FreeCAD has been on an upward trajectory in the past couple years as more people lean into the project out of frustration over increasingly hostile pricing from the commercial solutions. KiCAD has seen incredible advances since CERN started pouring resources into it, I'm sure Netflix money is going to help Blender. Now to get some large engineering shop to consider FreeCAD as their exit path to Siemens/et al...
Unfortunately it’ll be a lot harder for CAD because of all of the other lock in like PLM/ERP integration. A good PLM is half the product. I know a good amount of companies that do not use solidworks because their PLM is absolutely crappy (but I haven’t been a MechE for a couple years now so things could have changed)
To be fair animated 3D modeling is a complex task so the UI can only get so simple. Even commercial tools require training and have challenging interfaces.
Another example is Gimp. People like to bag on it for having a terrible interface, but when they say Photoshop is so much better I have to wonder what magical version they are using. For me the differences between the two are marginal, but that may be because I learned how to use Gimp first and have to hunt around Photoshop's interface more.
> To be fair animated 3D modeling is a complex task so the UI can only get so simple
The interface doesn't have to be simple. What it should be is conforming to established UI patterns and conventions. Blender used to be incredibly unintuitive even to people who had never used any other 3D modeler before.
Some complexity is of course an inherent part of the domain, and you can never truly get around it. At best you can stick to sensible defaults for the basic user and initially hide the most complicated parts until they are needed.
On the other hand: there is no reason whatsoever something as trivial as drawing a circle has to be as complicated as launching a rocket in GIMP. It certainly doesn't lack the technological scaffolding, and UX-wise people would already be ecstatic if they just cloned how 30-year-old version of Microsoft Paint did it. So why doesn't it have a Circle Tool yet, despite the massive amount of requests for it? The only possible explanation left seems to be an active disdain towards basic non-technical people: the UI is hard to use because they want it to be hard to use.
This is my perspective as well. I've been a big FOSS junkie and, in ~2015 or so, Blender had a repute similar to GIMP. (A free, worse version of proprietary tools).
By the time I picked Blender up in 2016 (before 2.8!) it felt pretty mature, but I used it (still) because it was the one that was free and which worked on Linux.
The time and energy I put into learning Blender feels like an investment that has paid off amazing dividends.
(I'd also picked up Godot at the same time, with much the same story of elation on its adoption rate).
It's the same with KiCad!
Version 5 was kinda-sorta usable - but buggy and painful. In practice everyone would tell you to just download the nightly build of version 6 instead, as the UX improvements were massive. It became a genuine joy to use, and with the death of EAGLE the no-brainer choice for every hobbyist.
Since then development has raced ahead, with regularly scheduled released chock-full of both small quality-of-life improvements and new features focused on professional use. It's still a tier behind the likes of Altium, but these days KiCad is a very solid choice for everything but the most high-end PCBs.
It is definitely good enough to build your small consumer business electronics business around, which means there are suddenly a lot of potential users willing to throw a few bucks at it to solve the remaining small papercuts and missing features they run into.
That shows the importance of listening to users. I too tried to learn Blender before the UI overhaul, but with prior 3ds max experience, Blender was infuriatingly counterintuitive; for example, it used the right mouse button instead of the left to select objects. Felt like those deliberately annoying demo pages that make you select phone numbers from drop-downs and click on moving buttons to submit forms.
The context was also weirdly random, probably with some logic for longtime Blender users but just weirdly random.
The usual context for modelling, [[[ Mode(model/uv/anim) -> Object/Mesh selection -> Face/Line/Vertex selection ]]] that is found [[[ (top-to-bottom)-(left-to-right) ]]] since Blender 2.8 and most other programs used to be placed [[[ middle of screen-top of screen-middle of screen ]]], just an insane order and that stuff was actually defended by Blender-die-hards (that probably used keybindings for these context switches anyhow).
There is still things placed "weirdly", but once we got past that it became immensly better (and not rage-quit worthy).
I remember when Blender first forked from NeoGeo's old code: it was clunky, alien and just plain weird. But even then the slashdot crowd was remarking about how snappy the UI was, once they figured out how to use it.
Other open source projects should take note. It seems like UX is a complete afterthought for most and any suggestions for QoL improvements are met with hostility by the small fervent community telling everyone to go fork themselves.
Somewhat relevant XKCD: https://xkcd.com/1172/
Where are all of the open source UI/UX peeps? Why do they not exist? Why are so many devs accepting of the open source concept and yet apparently no UI types are by comparison? The number of open source UI peeps rounds to zero.
What is it about design/artsy types that makes working on open source anathema where coders will do it just for the lulz?
It's not that UI/UX people don't want to contribute, it's that the coders have to be convinced that UI/UX matters enough to start including designers' contributions. The type of people making FOSS stuff also tend to be the people who prioritize code, make "good enough" interfaces, and see UI/UX work as fluff. This is thankfully less true today than it was in the past, but it's always been part of my experience around FOSS.
I'm definitely one of those strictly-programming people.
Sure, I can probably hack together a sorta-kinda technically-usable UI, but I know I'm awful at it. In my professional life I quite early on realized any attempt on my side is just a waste of time and effort, so these days as a mostly-backend developer I don't go beyond sticking bare unstyled HTML elements on a page to demonstrate basic functionality. I'll leave all the design stuff to the people who are actually good at it!
It's not that hard though to make a drastic improvement over that. Basic CSS isn't complicated. Sure, there's quirky things, but it's still possible. As "over the weekend" type of projects, I have taken websites/apps with UIs that I've thought were interesting and recreated them with HTML/CSS. You learn a lot very quickly, and then your UIs start to suck a lot less. It's still coding as far as writing in and IDE and testing/fixing bugs. You just get to skip the compiling part of it!
Almost no one is being paid to make desktop apps any longer. And the UI/X discipline did not make it to the web for whatever reason. The last gasps became designers and settled on more padding, rounded corners, and hidden scrollbars. Most of which are pretty but counterproductive. Dead discipline except for very large products.
The number of job reqs I've seen posted for a dedicated position in the last ten years could be counted on one hand with several fingers left over. I'm interested in the subject so pay attention.
As mentioned the responsibility tends to have been subsumed by designers, but believe few designers study HCI. Part of the reason interfaces get harder to use every year.
It's a self-reinforcing loop. Once a FLOSS tool becomes good enough, it'll start to attract professional users, who are willing to invest in it, which makes it even better. And it is quite hard for commercial players to compete with free.
But FLOSS software is mainly made by developers. Who like writing new flashy features, but are awful at UX, and making sure all the small kinks are worked out.
So most FLOSS software gets stuck in a "death by a thousand papercuts" scenario, where it has enough features to technically be usable but it is painful enough to use that no professional would ever adopt it.
Blender got out of it. I really hope more projects will follow their example.