Blender has made some big UI gains. I used it a long time ago, and the hidden menus and backwards control scheme made it unusable for anyone with basic computing knowledge (right click was select? Insanity) and doubly so for anyone with 3D knowledge. They revamped it a few years back and it’s usable now. I went from a blender hater to happily dropping paid 3D software and preaching about how good blender is.
Companies could throw money at gimp. But that doesn’t mean it’ll ever have a control scheme that’s usable to the 99.9% of people with prior digital art experience who go in expecting something similar to 99.999% of other software, only to find it’s completely backwards.
Literally just putting aside all other development plans and going all in for a year on making the UI usable would launch the project to success.
And comments like this always get a “well, it’s open source so you could fix it yourself if it’s that bad!” I could and so could others. But there’s usable software out there, and instead of dedicating loads of time to trying to fix a decades-old mess, we’d all rather just use a better product even if it costs money and get to work, or make something different from the ground up.
I love the GIMP, use it all the time, find it super-intuitive, and last time I had to use Photoshop I felt like a fresh vim user who couldnt wait to figure out how to get the hell out of there.
I suspect if you surveyed 1000 art professionals, you'd get very few who like or use GIMP. You're going to get a very different result with something like Blender and 3D designers though, which really should make the GIMP devs pause and think what it is that Blender does well that GIMP doesn't.
But I have found Gimp to be wildly inconsistent in the UI. I never remember how to draw a circle, for example, when I just want to add something to a screenshot because you have to use the circle cut tool and then choose a different option which I never remember because it is unintuitive. Yet its easy to have a paintbrush that draws one of those texture patterns that I can't imagine anyone ever using.
As such I then don't use it.
There just isn't a task that Gimp fits.
Most people don’t even think about photoshop UI. Same with Microsoft paint: users can get in and start figuring out what they want to do just by clicking around.
There's hundreds of thousands downloads every month. I'm sure you have better facts at your disposal and will share them easily.
I'm afraid you are dismissing quite a lot of work to improve the UX/UI.
Could more be done? Absolutely! Was there zero attempt? Not really, no.
It’s like: if you’re a large manufacturer, would you prefer investing in a city where government officials pocket the cash and roads don’t exist but with handwavy excuses that something might improve with a little more cash, or invest literally anywhere else where the roads improve every year and things are getting cleaner?
Gimp seems to be spinning its tires for infinity. I’ve tried it once a year or so for 15 years. It hasn’t changed much. It’s incredible. Almost no company is going to throw them cash and it’s obvious why.
The right kind of an analogy would be an investor stumbling upon a tiny community of developers that has been serving hundreds of thousands of users for the past 25+ years without pay.
IMO, one of the core reasons why UI/UX in open source applications (that aren't backed or actively contributed by dedicated staff in big companies that keeps them on a salary for the sole purpose of contributing) is because it's more widely accepted for programmers to contribute to open source than UI/UX designers.
In this case the best one can do is design mockups, but then who implements it? Who even has to be convinced that the mockup is a step forward or backwards, the moment every design decision is treated as having the same value as the next one, on the grounds of "it's just an opinion"?
My point is, we don't lack programmers/developers we lack UI/UX people willing to contribute, and we also lack the former to actually listen to what they have these people to say instead of dismissing their suggestions as "not data driven enough", "too opinionated" etc.
It’s why as a UX designer I don’t contribute to open source projects any more as the fight just isn’t worth it.
GIMP is peak that attitude. It could be a top-quality mainstream open source image editor if they wanted it to, but they don’t.
Photopea is the work of one guy, Affinity Photo was made by a small team, so it’s clear it’s not manpower but attitude.
They’ve got things just how they like it and the last thing they want is to make things easier and have a bunch of n00bs running around and spoiling things, thank you very much.
Except the GIMP team worked with a UX architect for several years. That's how you got single-window mode, convenient rectangular/ellipse selection and cropping tool, polygonal selection tool built into lasso, the unified transform tool, and more.
And even after Peter (the UX guy) departed, the team still made a bunch of UX improvements (and continues making them, time-permitting).
> Affinity Photo was made by a small team, so it’s clear it’s not manpower but attitude.
Serif Ltd. is 200 to 500 employees. They have paid developers working full time. They had the funds to start from scratch and build on their previous experience creating similar applications.
In case of GIMP, the existing team is already stretched thin and has to deal with too many reports and requests to handle. In the past 4 years, the amount of reports in the issue tracker more than doubled, but the team hasn't grown accordingly. I don't think it's realistic to expect that the team would be able to maintain the current code base while starting anew. If the current progress is already commonly referred to as glacial, what do you think will happen if they start reimplementing GIMP from scratch?
And I get it why you are referring to Photopea here. I think Photopea is a formidable effort. But it's been under development for 10 years already. That's quite a long time. And Ivan didn't have another image editor to maintain while working on that project. AFAIK, he originally intended the write a web app to merely open and show a PSD file in the browser. That's a quite different development trajectory.
There’s perfectly good software out there. I’d rather get work done and use that, than spend forever trying to make sense of some mess of an ancient project’s code and trying to fix it. There are even free projects with usable UI that would be a better use of time if I were to want to contribute to something. Krita’s been mentioned here a few times and it seems to have a brighter future.
For what it's worth, GIMP does indeed have about a million dollars sitting in their donation account. So money isn't their issue, project management is.
Infact I find the MDI interface to be quite ergonomic.
Also i think Photoshop is intuitive mainly because everybody is so used to it. Gimp just has very bad ux because they cant colectivelly agree and unify the ux. If the ux was very different but very good it might be even advantage but now its lot worse PS.
This happened because Adobe sued Macromedia over palettes.
Annoyingly enough, the interface has not substantially changed in the intervening years despite Adobe eating MM.
I guess they are only intuitive if you use a graphics tablet?
I think people who consider GIMP a good substitute for Photoshop probably haven't ever used Photoshop seriously, or have photo editing needs that fall in a very narrow range.
Companies which don't want to pay for Creative Cloud can more easily pay for one of the various alternatives--the particular alternative that works for you will be different.
I am more a PSP than a PS guy and I really tried to use GIMP for my image editing needs on kubuntu. I even ran a pinned patched version of it fixing the save/export "bug" (Another boneheaded ideological dev decision).
A few months back I wanted to add a few straight lines to an image. In PSP these would be vector object that can be adjusted in size, color and position. In GIMP not so much. It has only the pencil tool with an alt mode holding shift that does straight lines.
However I needed to adjust the lines distance from each other slightly - What followed was a lot of cursing and repeated attempts, switching to the path tool followed by discarding that idea and switching to making a layer per line. After finishing I deleted what I had done and did the same by starting PSP 9 in WINE in about 30 seconds.
I did some googling and found the reason why GIMP lacks any useable Vector functions is that vector functions belong in a Drawing program not a Manipulation program. And GIMP is the latter. This is a paraphrased from memory quote by a GIMP dev.
That probably means that not putting any vector function in (even though stuff like the Text function really benefits from vector layers) was a decision long ago that passed into ideological canon by now and is probably a huge task to implement.
So I am currently back to using a PSP version from 1998 for editing images on ubuntu. Even though its buggy as hell. Newer version after the Corel acquisition sadly come with a DRM service that balks at wine. I have licences for a few of those I would love to use.
Vector layers was a 2006 Google Summer of Code project. It was too ambitious, it wasn't completed, whatever good bits were done got merged years later so that the code wouldn't be lost completely, and there is still no user interface for any of that.
But the sole reason this project was assigned was because people behind GIMP were quite open to having better drawing tools in the program. They just lacked the time and the motivation to work on that.
When your program is primarily intended for image manipulation and you have a crapton of requests from users to add image manipulation tools, filters, and whatnot, how else do you prioritize your work? Especially when you program's architecture turns out to be inadequate for complex features and you have to rearchitect and repurpose 800K LOC, not to mention updating it for the next generation of the UI toolkit that is a moving target of its own?
For every user complaining about not having shape tools you'll get three complaining about the UI, another two or three complaining about not having non-destructive editing, another two demanding layer styles, then an assorted group of individuals wanting CMYK image mode, LAB image mode, more use of GPU, AI features, more filters, better HiDPI support, and the list goes on. All the while you have essentially one person doing 50% of all work and a small team of people all working on small items. All the while there are free/libre applications like Inkscape that have been providing better drawing tools for years.
So I'll ask again: given all that, how else do you prioritize your work?
Sure, but when I look at the development of image editing tools in the 90ies then vector shapes and layer pretty much was widespread by 1998. If I were to guess then the functionality just came with support for TT Fonts and became a standard feature for image manipulation applications.
Again I tried really hard to make GIMP work for me - I still found myself hunting for the right tool icon time and time again (I think something is fundamentally wrong there but I cannot with certainty say what) This might be where the UI criticism comes from. I bet you would learn the symbols in the end and . I still got so annoyed by the save function insisting to save as some weird format I didn't want to use that I had to patch that out. But then in the end what killed it for me was the vector lines thing.
CMYK image mode being pretty much the only missing feature from Photoshop 3.0?
I think people tend to forget all the hoops they had to jump through years ago.
Do you remember how you couldn't edit text layers until version 4.0 or 5.0?
Or how most filters and some selection tools didn't work in 32-bit mode until fairly far in the CS release series?
So GIMP not having CMYK, or any color space support besides RGB, is actually sort of amusing. Just through version 7 Photoshop added vector shapes, history brush, adjustment layers, and a bunch of other stuff GIMP still lacks. It's an FOSS project so it's not like I'm saying the GIMP developers are bad, they clearly are quite talented.
What I'm saying is GIMP today is not as capable as Photoshop from 25 years ago. It's good enough for a lot of people but useless for others. I've been hearing your arguments about GIMP for as long as it has existed. It has a lot of capability and works for your use case so that must mean it just works for everyone. It's great that it works for some people. But you can't pretend Photoshop hasn't improved faster than GIMP. If you need anything beyond basic destructive editing GIMP is just not appropriate.
Must it mean that? Really? I had no idea! :)
> But you can't pretend Photoshop hasn't improved faster than GIMP.
That's not what I'm doing here :)
GIMP development started in 1995. First version of Adobe Photoshop with adjustment layers was released a year later, in 1996. The notion just wasn't there.
If you still want talking about notions and time, GEGL as the non-destructive backend for GIMP was started by Rhythm&Hues developers in early 2000.
We could have a long conversation about priorities and availability of developers and refactoring and all the reasons resulting in non-destructive editing not yet available, if you like.
Layer Effects are the only 5.0 feature from Photoshop that GIMP doesn't yet have.
There's also a million things GIMP does better or does at all compared to all the versions of Photoshop you mentioned. I can give you examples or you can spare me the time and just admit you've seen Photoshop 3/4/5 so long ago that you don't quite remember what it was like :)
I could clearly see the image I wanted in the red, green and blue channels. But because the alpha channel was filled with full transparent, all I could see outside of individual color channel mode was the checkered background. Deleting the alpha channel or converting the image format from RGBA to RGB just erased the entire image to black.
I eventually found the solution in the eraser tool, which had an erase alpha mode. I had to set it to a really large size and go over the entire image.
Would probably have been faster to just write a script in python.
With that in mind seems believable that companies where long term impact of investing in tools is clearer are more likely to consider investing in open source tools for the sake of long term benefits. In comparison with 2d drawing industry where investment in tooling is mostly limited to skill and experience of individual persons working with specific software.
There is also fact that final output of 2d drawing software is plain image which have relatively simple widely used standard formats (for practical reasons the Photoshop files are still commonly exchanged but sooner or later result is treated as plain image). With 3D software the complexity of output and integration with other software is much bigger. This encourages other big software companies to invest in Blender for the sake of better integration with their tools. Unity, Unreal consuming 3d models produced in blender, there is interest for both good workflow and integration, as well general accessibility to tools required during game development because more game developers means more potential game developers buying their software. Nvidia and AMD interested in better support for their hardware, so that end users are more likely to buy it.
GIMP was always a toy for programmers, without any vision behind for how to actually make it a product that commercial users would want. It’s a big job to move a ship like that.
GIMP is not competitive with Adobe Photoshop from a capabilities or user experience perspective.
Here’s a non-biased article comparing GIMP to Photoshop: https://thegimptutorials.com/gimp-vs-photoshop/
Also blender was regularly used to make content with input from the community. Multiple shorts that dog fooded it and provided improvements through the process, while inspiring many to see what’s possible.
GIMP doesn’t have any of that. Photoshop can be had now for $10/mo or something cheap like that. It’s not thousands of dollars a year like for DCCs. It doesn’t have productions driving features in it , or creating content regularly to inspire people.
Krita does better here because they do small but more engaged community outreach to create content with Krita and have it feed back into the app
The Blender Foundation and community made that happen! No glib “if you don’t like it, fix it!” type comments. I’m not suggesting that you personally have said that, but look at the context and process of Blender’s development and success. The people at the Foundation, along with a buoyant community have worked hard and listened to the criticisms and transparently worked on them. It’s been a long slog.
They’re not only developing it, they’re trying it in actual productions and using that to improve it
> The Fuzzy Select tool in GIMP took about 20 seconds to decide what I had clicked on and isolate it for me
That's uncommon enough to be an isolated incident.
Would it be nice to have? Sure. But there are tons of alternatives to Photoshop that just don’t compete on market share, despite photoshop lagging behind in many areas.
If companies were backing an OSS editor though, Krita is by far more favoured by professional creatives. It’s more familiar coming from Photoshop, has better extensibility+Python scripting+Qt which is great for pipelines and has a much better handling of color science and other features important for professionals.
Really? How? :)
You are ignoring non-profits and countries foreign to you.
In the context of this discussion, the question was why companies aren’t funding GIMP development, and the ones that can’t afford a $20/mo aren’t the ones who would be funding development anyway.
Yes, of course there’s a market for free stuff. But the price of admission into the existing paid for products is low enough that there’s little motivation to fund it.
GIMP 2.99.x (future 3.0) is shipping with support for Python, Script-Fu, Lua, JavaScript, and Vala as languages for extensions.
The last time I used Gimp, the Gimp Python extension was an optional module that wasn’t included in the build I had But it’s been a while clearly and I missed it. I’ll edit my prior comment
I think a lot of things in Gimp are not broken, there is a worthy foundation to build upon. The text UI is bad but it's not the cornerstone of the product, it can be fixed without rearchitecting it. The library approach to graphical ops is more important, and it looks promising to me.
Look how much has Inkscape's UI improved, to say nothing about Blender.
Blender is, after Linux, the golden child of FOSS for professional level tooling and I’d argue the model for running an open source project.
If you're going to write a software for artists, design the UI for artists, not developers... "just learn the UI" doesn't cut it. The problem isn't resource oriented, it's "we don't want to change anything because we think there is nothing wrong with the UI".
While every software is different, there are rules to UX/UI and GIMP is breaking too many of them. Also the bar to contribute is C and GObject knowledge which is quite high. Blender UI can be modified with Python so it's much easier for a designer to make prototypes.
That seems to be a popular myth.