/make/ - Creative

World made of things

Want your event posted here? Requests accepted in this /meta/ thread.

Max message length: 5120

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB


(used to delete files and postings)

Open file (34.60 KB 1030x683 demopaint.png)
An image editor project Anonymous 05/09/2020 (Sat) 23:37:41 No.76
I often try to design something like the UI for a program or a game, or maybe typeset a manga, but struggle to do it because a program that is "decent", "free", and "image editor" at the same time doesn't exist. Gimp as a whole is a clumsy piece of shit, Krita devs seems to have no intention of ever fixing their broken ass text tool, MS Paint simply crashes if I try to draw a text box, many others have either no text tool at all or are severely lacking in other ways. I could write a book about all the ways that every single image editor is a pile of shit, but I can't pray and wait forever for one of them to get fixed so I think I just need to bite the bullet and prioritize making it myself. There's 2 I have already thought about in the past, one of which is just a simplified and stripped down version of the other, meant to replace MS Paint, pic related. I doubt I'll be properly making either one for a while, but I want to at least start on an image editor of some kind so I can design things. Making an image editor isn't actually hard at all, it's only hard to add certain features like magnetic selections and weird filters and support for all the file types, and making the rendering fast when you use large brushes on a large canvas and such. But for now I just need a simple editor that can draw basic shapes and text, paste images, and cut/paste selections.
>but I can't pray and wait forever for one of them to get fixed so I think I just need to bite the bullet and prioritize making it myself. Or find the midground: fix it for them. That saves you spending many years re-inventing the wheel, instead giving you near-immediate results that benefit everyone using those projects. Krita needs a new text tool, you say? Why not fix that issue? We don't need 'yet another program' unless you have a goal which has it's own significant values.
>>78 >Why not fix that issue? Because I would have to effectively rewrite the entire program if I want to be satisfied, and that means spending 17 years getting familiar with their almost certainly cancerous and bloated codebase along with all the gay as fuck libraries they use and basically taking over the entire project, and it's impossible to do that in a place where an arbitrary number of random internet people keep submitting changes and the people who made it so shit to begin with are running some dev group that is charge. I also don't want anything to do with the Krita dev team itself, they're clearly fucking retarded based on what I've seen in the program, heard them say, and seen them prioritize. There's many things I want that I think are obvious, but they've flat out refused to add because "it's not what the program is for". It would be easier and more enjoyable to just start from scratch. I'm not joking when I said I could write a book about all the ways these programs suck. It's hard to give an idea about what I mean without actually writing that book, but basically every button and every panel in the whole program has something wrong with it. Whether it's hard to read or scales badly or looks inconsistent or works weirdly or resets randomly or wastes space on the screen or has confusing organization or flickers visually or is lacking useful things or requires unnecessary fiddling with or is split between multiple places for no reason or interrupts the keyboard or is just flat out broken. And even if you ignore those, everything just takes too long. It shouldn't take 10 seconds to open the program, it shouldn't lag the whole program to death or load for 20 seconds if a file has 70 small layers. There's also ways in which I think all these programs just approach certain things wrong, including commercial programs like Photoshop. One of those things is shoving popup panels at you for every little thing instead of just having the options in the UI. Another is having multiple programs for very similar jobs even though there's no reason they can't be the same program. There's all kinds of little ways you use and interact with the program that I think would be better if done another way, but just like minimaps in videogames, people don't even consider any other option because they're so used to the way things are.
>Tried GIMP and MSPaint, so I'm going to write my own Hey nigger, maybe try more than the two worst options available.
>>81 Like bot.net, Pencil2D (no text tool), Mypaint (no text tool), FireAlpaca (no text tool), Milton (no text tool), Pinta (crashes or something, I forget)? I'm sure there's other I'm not remembering.
>>82 Paint.NET is honestly pretty solid.
>>83 I don't remember why since it was years ago, but I wrote it off as being too jewish/suspicious or something. Like it just wouldn't let me install it without going through all kinds of things or something.
You should look into Nichimen N-World and its spiritual successor Wings 3D. You say you could write a book about all the ways that every single image editor is a pile of shit, and I say you probably should because it would be more productive in the end than trying to roll your own. At least this way some people would get the memo even if nothing ever changes. See the peanut gallery comments you received.
I second the motion for OP to write the book on how all image editors suck ass, then draft a design document for a non-shitty FOSS image editor that could eclipse everything else. Extended tooltips (explaining the limitations of a tool) would be a huge asset in just about every program where there are multiple complicated ways to do something that should can be explained in a straightforward manner. I get by with GIMP but the learning curve was annoying.
>>87 >>88 To write something comprehensive enough to be of use to the world, I'd have to start using various image editors a lot, and I don't really want to spend my time on that. Plus it's hard to communicate certain things without giving an example that you can feel, especially when the world is full of people who think Gimp is better than Photoshop and generally see no problems with FOSS programs. >Extended tooltips (explaining the limitations of a tool) I'm not sure what you mean by limitations, but making the program easier to understand is one of the things I'd improve as well. Mostly by making the program simpler and more consistent, but also by guiding the user somehow, like changing the cursor based on what's under it and giving better feedback when you hold shift/ctrl/alt so you can tell how they change the tool behaviors. And also by having a manual built into the program that explains all the ins and outs of the various features and how you're expected to use them. I'd also like to have explanations for filters and blending modes that explains what exactly it does and how you might want to use it.
Open file (79.76 KB 1137x1008 imgui_v167-misc.png)
>>76 >UI for >a game will this do, OP? > You'll have to be able to write code, ofc. https://github.com/ocornut/imgui
Open file (159.25 KB 1715x833 explorer_yoc8UlHTQO.png)
To avoid UI being unusable abomination start with snappy/floaty panels pixel-perfect well-proportioned DPI-scalable GUI toolkit similar to Adobe's or Affinity's that would blow out GTK, QT and other freetard trash out of the water. Good luck, I believ e in u.
>>91 >just spend millions of dollars developing complete, proprietary systems before you make your first babby-tier steps <WE BELIEVE IN YOU ANONYMOOSE!11 you glow like a thousand suns tbh.
>>92 yeah, put ui building blocks for a graphical editor aside. every time. most freetard application software ends up with perpetual downs syndrome look and usability no matter how cool tech may be underneath because ui is scary/unimportant/another excuse.
Open file (154.56 KB 1437x1042 g.png)
>>90 I don't need the code for a UI system, I just need a design/graphics program where I can design the way the UI looks before I actually start building it. >>91 >>93 I don't plan on using a premade GUI library like GTK since I want full control over the way everything looks and works and I want it to work simply and logically. I've worked on a UI system in the past, though it's not finished so I can only use it for very simplistic things for now. I'm not sure about making a "replacement" for those things though, I think generally speaking making a "do it all" system like that just ends up in a mess that's more complicated and slow than it needs to be because it has to account for every possible scenario. Who knows though, maybe it'll end up being something like that. >DPI-scalable I plan to show settings that lets you scale the UI (and do other tweaks and changes) when you first run it. In terms of design my #1 priority by far is clarity and ease of use, I'd rather make it ugly than make it hard to read and inconsistent and hide settings just for the sake of cleanness. Pic related is the latest UI mockup for the bigger program I mentioned, though there's a lot it doesn't show.
Not exactly on-topic, but I thought these might be fun to look at: https://github.com/yaronn/blessed-contrib https://jesseross.com/clients/gnustep/ui/concepts/
>>97 not the OP, but that's some pretty neat stuff anon thanks.
Krita has a decent text tool nowadays. Krita is only getting better.
Open file (370.03 KB 1280x1080 1591675617511.jpg)
>>112 >"we're gonna remake the text tool it'll be vectors and shit!" >tear it apart and break it, making it effectively unusable >give up and forget about it like a true open source team >hold fundraisers but make no mention of finishing the text tool with any of that money >still not a peep over a year later

Report/Delete/Moderation Forms

Captcha (required for reports and bans by board staff)

no cookies?