I am using 2.99.18 (non release, unstable build). Non destructive editing has landed. You can make adjustments through the usual menus and then enable/disable the adjustment under layer effects.
I’ve tried learning GIMP, and it sucks. I’m not saying GIMP sucks, but you have to be crazy to not see that it’s hard to learn.
I use GIMP for memes and here's my two favorite tips
Hit the forward slash key / to open a command palette and jump to any action
To remove backgrounds, use a layer mask. select around the object and paint a white/black section on the layer mask. Here comes the trick: use a Gaussian filter on the layer mask to create a transition from black to white and the crop job looks a lot less choppy.
My anti-tip
Adding text and shapes sucks and I never found a way to make it better. Export your image and finish the job in Krita, Pinta, Photopea, ...
Smart people use terms like gaussian blur to refer to a blur distribution rather than a non technical term such as feather. Feathers are what helps birds fly for example.
Let's try:
"You can feather the pedals on your car to make it drift."
"You can feather the pedals on your Ford to make it do the Tokyo style turns."
Any tricks on getting the fuzzy select tool to work? Even after adjusting the threshold, it is just garbage in my experience. Nothing close to Affinity/Photoshop. Unless I am selecting something that is in front of a very solid background, I just use a paint brush on a layer mask in order to "cut out" an object.
It has the same problem as most FOSS packages that are too wide in breadth and have multiple contributors with their own hobby horses pulling in all different directions, and to this day does not actually provide a feature-complete whole, nor an interface that actually makes sense. And it's not a matter of the workflow just being different -- it categorically fails to replicate functionality that is core to its commercial competitors. Numerous other "big" productivity packages have the same problem including FreeCAD (boy does it ever), LibreOffice, etc. I say this as a staunch supporter of FreeCAD, by the way. It's the only CAD software I use even though it's a pain in my ass.
The shining exception to this I see is Inkscape, but it is still significantly less powerful than even early versions of CorelDraw.
For 2D graphics work these days, I hold my nose and just use Corel. I use it for work. Like, actual commercial work. That I get paid for. It is at least a lesser evil than doing business with Adobe.
And if you want to stick it to the man, it is easily pirated.
I always wondered if I could contribute/volunteer to a FOSS somehow with some UIX stuff, but I don't even know where to start. Would you just draw a concept ui for the team to work out or something?
Not that I'm great at it, but man, we gotta start somewhere, right?
This is probably common. The people that work on UI often aren't the people who do pull requests. But I think if you want to contribute it would be best to get in touch with a maintainer on the chat of the project. Projects often have a matrix/irc/discord on the git page.
I think you can start a figma or other collaborative UI/UX as an idea first. If a developer is interested in implementing it, then you move on to the next feature
There's also two main plus one lesser issue that are less commonly discussed:
Lack of manpower. FOSS devs often doing it as a side project on top of some other and/or a full-time job, so that even lowers one's ability of concentrate on stuff like the UI, when you're already working hard on fixing bugs, looking up things (which is getting harder and harder thanks to AI slop - I once managed to destroy a Linux on my Raspberry Pi while trying to adjust the path variables).
Getting comfortable with the uncomfortable parts of your application. There are many times I haven't noticed a a very uncomfortable part of my GUI after months of use, then I had to refactor things, which obviously took time away from other things. This also affects the users already in the userbase.
Elitism is also a factor. A lot of people like the feeling of being part of a special group, and for them, the steep learning curve is a feature, not a bug. I've seen Blender users being angry at the devs for "spoonfeeding" the normies, and letting in all kinds of people. Also just look at OP's image.
My old best friend and I used to be a programming tag team that worked pretty well; he'd slap together w semi-functional version of the idea we had and then id go in and make the UI make sense and fix all the logic bugs and typos.
I'm not saying I'm some perfect UI guru or anything but the way he (and other people I've met) seem to have no internal base knowledge of shit like "similar settings probably shouldn't go on completely opposite sides of the screen under different menus" or "5-deep nested drop-down menus hurt people's souls"
Honestly I still struggle a lot with this. I can click around a UI and feel what might confuse a user, but building a UI from scratch feels like such a shot in the dark.
So, why do UI people not use and contribute to FOSS then? Are they all on Mac? Then go complain to them or contribute your desired UI improvements. FOSS isn't an all you can eat buffet.
Personally, I think UI people are less idealistic and I do look down on them for that.
It's super hard to get involved as a UI person. If you're a developer, you can just rock up to a project and fix bugs, and if you follow the coding style they'll probably get accepted.
If you want to successfully contribute as a UI person you have to convince a bunch of developers that you know what they should be doing better than they do. It basically never happens.
It's super hard to get involved as a UI person. If you're a developer, you can just rock up to a project and fix bugs, and if you follow the coding style they'll probably get accepted.
If you want to successfully contribute as a UI person you have to convince a bunch of developers that you know what they should be doing better than they do. It basically never happens.
It's super hard to get involved as a UI person. If you're a developer, you can just rock up to a project and fix bugs, and if you follow the coding style they'll probably get accepted.
If you want to successfully contribute as a UI person you have to convince a bunch of developers that you know what they should be doing better than they do. It basically never happens.
KiCAD has also improved greatly over the last few years. It still has an opinion on how the work flow should be, but that work flow moves pretty well. It's gotten easier to find pre-made footprints, too.
The Autodesk forums are 40% this, 20% "just learn to program, spend a few years getting good at it, then write yourself a custom script to do what you are struggling with", 20% "you are wrong for wanting that in the first place" or "you are wrong for having this issue", 15% "this has been brought up once at some point in the past two decades, try searching", 4% "OMG yes I have this issue too!"...
...and 1% split between actual helpful answers, and confirmation that it's a known issue.
Yeah, I recently found a post there where a person wanted to modify a downloaded mesh. The first comment was telling them they would need years of experience to do it well. The OP responded that they had figured out a solution that they were happy with, to which someone told him that his results were shitty and then explained a way to do it better. When the OP got upset at this back to back dismissal, everyone unanimously decided they were an asshole.
It's infuriating trying to find solutions to issues with Autodesk.
But you did forget a classic one: "Hello I'm X from Autodesk support, you should open a support ticket so we can discuss this issue in a more one on one manner." And then the thread is closed without a solution.
I stuggled with GIMP at first, it was super frustrating because it does UI things differently than other image tools.
i.e. in other tools your active layer masks your drag selection, and in GIMP I would constantly be grabbing lmages from another layer, till I realized the pixel under pointer determines what image is moved. That function can make you highly productive since you don't need to preset layer, but god was it enraging at first
I've been using Gimp for years. It's the only way I know. If I tried to use photoshop I would have a hard time getting anything done too. I'm really good with gimp though.
I've used Gimp all through my teenage years. And I used it a LOT. It was quite a difficult transition to Photoshop (which my workplace uses). But once I got the hang of photoshop, I realized how convoluted Gimp really is.
Half the time spent in Gimp is making backups before making an edit. A third of your layers will be backup layers in case you change your mind about a design decision. The whole design process is super inflexible and therefor kills creativity.
Want to use an effect like gaussian blur or drop shadow? Make sure you backup your layer!
Want to edit text after you stretched it all out? I hope you made a backup of that layer!
Want to work with large files with many layers? You better hit ctrl S after every edit, because the program just might crash on you if you make a difficult selection!
To be fair, I haven't used much Gimp since 2.8, so if stuff is different now: awesome!
And I admire all volunteers that work to make stuff better. But for now, I'll stay away from it if I need to do heavy editing.
Photoshop is also hard to learn. What's your point? Just because it's different to what you used to does not means it's more or less difficult to learn.
Not having Paint.NET sucked when I switched to Linux. I got very used to it and that was the one I missed most... it took a few years bouncing between programs but I'm happy with Krita now. GIMP just never clicked for me unfortunately.
I sometimes think about making a Paint.NET clone for linux but i have too many other projects and hobbies i wanna do instead yk