OBS is an awesome app.
Its interface is a bit "awkward," but not too bad. It has many tools that are not available elsewhere.
Its interface is a bit "awkward," but not too bad. It has many tools that are not available elsewhere.
The other day he had an issue with a slight echo in the audio. He mentioned it when we were out of the house. When we got home I had the idea that maybe he'd added the Audio Input Capture twice ... but I couldn't think of why that would happen.
When i mentioned it to him, he'd already figured it out and solved it. He had added two audio inputs when he was trying to capture his headset mic and his desktop mic (with phone taped to it, playing music) ... I dunno what the point of this story is ... OBS has a weird interface, but it reminds me of the good old days, when you just made it work and the users just figured it out.
I think the polished UI/UX we have these days makes us complacent. Afraid of poking and prodding.
If your product isn't quite opinionated on the one right way to do things (and only a specific kind of thing) you can create a really shiny UX. If your product needs to be many things to many kinds of people, stuff ends up a bit messy but only to the detriment to extreme beginners.
If that's what polishing is please leave OBS all smudged and gross haha
The problem is explained in...
Design for the Novice, Configure for the Pro
https://bothsidesofthetable.com/design-for-the-novice-config...
Beyond the survey, are there usage logs we can share or limited telemetry we can opt into?
Make sure you do it differently than how Muse Group did it with Audacity (specifically the announcement and explanation of why and what parts) [0]. Which I'd argue is not particularly difficult to do.
[0]: https://arstechnica.com/gadgets/2021/07/no-open-source-audac...
Love OBS, everyone involved in it should legitimately feel proud of their work. Great software.
E: Survey done :)
By the way, as a community we should look at the changes Blender landed in 2.8, which were (I think) considered a major UI/UX redesign that increased accessibility to the broader public.
My sample size of 1 is a friend who is a professional UX designer. Once, I tried convincing them to help with some OS project, but the response was "that sounds like work, and I don't work for free".
Which has a point, I guess. Open Source contributors have normalized pouring countless hours for free into their projects (when not paid by a commercial entity). But outside of this bubble, I guess people don't see it with the same eyes. Even though in the end, it benefits everyone.
It just requires (semi?)professional knowledge to write good code, like it takes professional knowledge to write good UX, so "why would anyone do that unpaid". Maybe makes sense in the spirit of an artist who gives their art for free, but not in the mindset of a professional who charges good $$$ for an hour of work.
EDIT: Typos and unfinished phrases.
In the project that I developed, and turned over to a new team, we finally (after about ten years) got someone who is a professional UX manager/developer to step up, and he's been doing an awesome job.
In my case, my personal motivations run deep. I do a "better than pro" job on all my open source. I'm a passably decent designer and UX guy, so that helps.
But even I think the Calibre UI is pretty awful.
You have software domain knowledge and experience working with spaghetti interfaces assembled by developers rather than designed by experts. You find them more functional because it fits your existing mental model of an interface. However, these interfaces are vastly less functional to everyone else.
An interface that only facilitates beginner workflows is poorly designed. I rarely see these in the wild, but the culprits are usually non-designers pantomiming their understanding of ‘designed’.
I do, however, regularly see developers misinterpret interfaces designed to facilitate advanced functionality for non-developers. Focused views with controls and readouts obscured as appropriate, but functionality exposed through key shortcuts or other means are what most people expect, and it’s a whole lot easier for a developer to learn those usage models than for non-developer users to learn how to use a hundred buttons and readouts acting as a thin wrapper around an API.
The secret probably is having well defined roles: whoever writes the code shouldn't be involved in interface design and vice versa, unless they prove themselves good at it, as the two tasks require very different sets of skills. This is easily accomplished in commercial software writing where roles are defined and salaries paid, but in the FOSS world where everyone and their cat want to be the famous developer, the UI designer role is rarely recognized for its importance.
Calibre's UI isn't illogical or nonsensical, and it can do everything you probably want when it comes to managing an ebook library, converting between book formats, and taking files and off ebook readers. But it doesn't feel like anyone involved with the UI design ever really thought about it from the perspective of "how do I make the common tasks easy" rather than "how do I ensure all the functionality is exposed".
That is, when designing something, and especially when there's nothing to yet dictate what the form should be, let function guide you to it.
If you were to ask "how should a table lamp be designed", I would point to its function -- to give diffuse light and allow switching it on and off -- as the primary guide. Make that work well. Don't let the form get a life of its own that will distract from the function. Rather, the form should serve the function and make it beautiful.
I’ve been in media production for about a decade. You’d be surprised how many high budget projects use OBS for live streams. It really does get the job done.