The UX designers I knew had a bit of a circular problem: the poor UX put them off too much to develop that passion for FOSS that drives devs to work on it for free. They’d first have to puzzle out what things are supposed to do and how they’re supposed to go about achieving whatever they’re looking to do, basically reverse-engineering user flows that were never properly engineered.
I don’t have a solution for this and I don’t know if they’re representative, but acknowledging the importance of UX is a start for sure.
Honestly there are some things I feel Microsoft and Apple do still do better and since they are UI-first considerations, Linux will likely never keep up.
For example: Apple has app packages. Imagine a flat pack to install by just dragging into a folder. Thats it all dependencies are there in the package. When you double click, the default action is to run a predefined binary within the package. It’s basically a folder whose default action is launch a binary within rather than open and browse. If you want to uninstall, just drag to trash. The OS keeps receipts and when I sees a package is sent to the trash, it handles the cleanup for you.
Flatpaks kind of do this but not really. It’s really clever and there’s no reason Linux can’t do this other than it just doesn’t face the same design pressures Apple did when they made it. Apple was GUI first, Linux had a GUI as an afterthought.
Microsoft has the plus of turn it on and it’s already there. Anymore, that’s pretty much all it has. Linux is getting much better with easier to use graphical installers and pushing themselves into a system much faster than windows does. But if a user doesn’t understand how a computer boots, that they can tell the computer to just boot something else, it’s overwhelming to get to the Linux-is-easy part. I feel like there should be a single click installer that creates a recovery partition, tells the uEFI to boot into it, wipes the original windows partitions out and installs to the computer, then removes itself, lets Linux reboot and reclaim the entire drive with a partition resize.
I feel like there should be a single click installer that creates a recovery partition, tells the uEFI to boot into it, wipes the original windows partitions out and installs to the computer, then removes itself, lets Linux reboot and reclaim the entire drive with a partition resize
Most linux installs are easier than Windows. It’s not a Linux issue, it’s an “already installed” issue. I wouldn’t want installations to be any easier than they are because the steps you’re involved with are steps that someone needs to give input or you risk destroying everything you had. Any OS installation that isn’t OEM, linux or Windows, starts with the assumption that there might be something important that it shouldn’t touch. Linux has OEM installations that from the perspective of the user receiving the computer, are identical to starting Windows for the first time.
Every year I look how I could give a few hours a week, but I leave bewildered (seems like every projects wants you to fork ans submit changes). Do you or anyone know of arrangement more like “here’s a ticket, punch in, do it, punch out”.
seems like every projects wants you to fork ans submit changes
I think that’s just how software devs tend to work: a main “trunk” of code, from which they branch off for a feature or change, then when they’re done writing the new code, merge it back into the trunk. That’s the mode we’re used to, the only tool we know for managing parallel code development.
How else would we do things? Or rather: how would you do things? This isn’t a rhetorical question: I genuinely admit I don’t know how you’d prefer to operate, or what you mean by punching in and out.
If you’re asking for specific tasks, projects may have an issue tracker of some form where you can find stuff other people found, but other than that, FOSS often is a “see something that needs fixing, go and fix it” thing. Particularly if the devs themselves don’t notice a problem, because they know what that confusing icon means or how that tool works, they won’t create a task they don’t even know needs doing. Even if they know something is bad, they might not know how to make it better.
If you’re using or trying to use a program and notice something has been done poorly, is too complex for an entry-level user, whatever it is you think could or should be improved, talk to the people making it, or look for a contribution guide.
Even if you can’t contribute code, it might be helpful to share your skills and knowledge by opening issues, explaining specific problems and offering ways to solve them.
Becsuse you get someone in on the creation thst can say things like.
“99% of all software uses Control+s to save, yes W stands for Write Out, but literally no one thinks that way and Control W is unintuitive.”
“People prefer larger icons with meaningful symbols.”
“This setting is only accessible by using a CLI to edit a .conf file in /etc/software/config/current and thats a terrible design.”
“Your pop up dialogs should all be consistently sized with Yes/Cancel buttons in the same space so users can more instinctively know where to click.”
“The window design looks like Windows 95 and makes people think they are in prison and has driven at least one user to depression and suicide, maybe consider modernizing the look with some color and roundness.”
Yeah, the real “Year of the Linux Desktop” only occurs when devs actually get input from UX designers and stop designing clunky as fuck UIs.
The UX designers I knew had a bit of a circular problem: the poor UX put them off too much to develop that passion for FOSS that drives devs to work on it for free. They’d first have to puzzle out what things are supposed to do and how they’re supposed to go about achieving whatever they’re looking to do, basically reverse-engineering user flows that were never properly engineered.
I don’t have a solution for this and I don’t know if they’re representative, but acknowledging the importance of UX is a start for sure.
I strongly feel the only reasons Linux is actually gaining traction right now are
A) proton/wine getting good enough interfaces for regular people to start using them (steam, heroic, hopefully bottles someday gets easier to use)
B) KDE making strides with their desktop envoronment.
C) flatpaks being easy to use and fairly reliable once someone learns to prioritize them
All still have a LOT of work to do with UX, but they are close enough to good that its less of a hurdle for new people to pick them up.
Honestly there are some things I feel Microsoft and Apple do still do better and since they are UI-first considerations, Linux will likely never keep up.
For example: Apple has app packages. Imagine a flat pack to install by just dragging into a folder. Thats it all dependencies are there in the package. When you double click, the default action is to run a predefined binary within the package. It’s basically a folder whose default action is launch a binary within rather than open and browse. If you want to uninstall, just drag to trash. The OS keeps receipts and when I sees a package is sent to the trash, it handles the cleanup for you.
Flatpaks kind of do this but not really. It’s really clever and there’s no reason Linux can’t do this other than it just doesn’t face the same design pressures Apple did when they made it. Apple was GUI first, Linux had a GUI as an afterthought.
Microsoft has the plus of turn it on and it’s already there. Anymore, that’s pretty much all it has. Linux is getting much better with easier to use graphical installers and pushing themselves into a system much faster than windows does. But if a user doesn’t understand how a computer boots, that they can tell the computer to just boot something else, it’s overwhelming to get to the Linux-is-easy part. I feel like there should be a single click installer that creates a recovery partition, tells the uEFI to boot into it, wipes the original windows partitions out and installs to the computer, then removes itself, lets Linux reboot and reclaim the entire drive with a partition resize.
Most linux installs are easier than Windows. It’s not a Linux issue, it’s an “already installed” issue. I wouldn’t want installations to be any easier than they are because the steps you’re involved with are steps that someone needs to give input or you risk destroying everything you had. Any OS installation that isn’t OEM, linux or Windows, starts with the assumption that there might be something important that it shouldn’t touch. Linux has OEM installations that from the perspective of the user receiving the computer, are identical to starting Windows for the first time.
How the hell do UX designers contribute?!
Every year I look how I could give a few hours a week, but I leave bewildered (seems like every projects wants you to fork ans submit changes). Do you or anyone know of arrangement more like “here’s a ticket, punch in, do it, punch out”.
I think that’s just how software devs tend to work: a main “trunk” of code, from which they branch off for a feature or change, then when they’re done writing the new code, merge it back into the trunk. That’s the mode we’re used to, the only tool we know for managing parallel code development.
How else would we do things? Or rather: how would you do things? This isn’t a rhetorical question: I genuinely admit I don’t know how you’d prefer to operate, or what you mean by punching in and out.
If you’re asking for specific tasks, projects may have an issue tracker of some form where you can find stuff other people found, but other than that, FOSS often is a “see something that needs fixing, go and fix it” thing. Particularly if the devs themselves don’t notice a problem, because they know what that confusing icon means or how that tool works, they won’t create a task they don’t even know needs doing. Even if they know something is bad, they might not know how to make it better.
If you’re using or trying to use a program and notice something has been done poorly, is too complex for an entry-level user, whatever it is you think could or should be improved, talk to the people making it, or look for a contribution guide.
Even if you can’t contribute code, it might be helpful to share your skills and knowledge by opening issues, explaining specific problems and offering ways to solve them.
Great points.
The way I’ve worked is usually: onboarding, identify work to be done, tickets (part of that forking), review/done.
So if the onboarding, identifying wtbd and then the work is done by me independently… that seems like weeks of work to submit a suggestion… you know?
UX design work needs a lot of research, understanding and alignment to stick.
Am I missing something?
Becsuse you get someone in on the creation thst can say things like.
“99% of all software uses Control+s to save, yes W stands for Write Out, but literally no one thinks that way and Control W is unintuitive.”
“People prefer larger icons with meaningful symbols.”
“This setting is only accessible by using a CLI to edit a .conf file in /etc/software/config/current and thats a terrible design.”
“Your pop up dialogs should all be consistently sized with Yes/Cancel buttons in the same space so users can more instinctively know where to click.”
“The window design looks like Windows 95 and makes people think they are in prison and has driven at least one user to depression and suicide, maybe consider modernizing the look with some color and roundness.”