Cropping Into A Debate

My post about moving to Linux last week didn’t necessarily set the world ablaze—at this point, Linux users are less like Never Nudes and more like loose seals—but it did draw a little bit of conversation.

One of the conversation points I spotted about it was focused on my blanket dismissal of GIMP, the most popular open-source image editor out there. It has this unusual status in the software world—it is immensely popular among Linux users, but it has often been dismissed by people with design backgrounds like myself, despite being highly capable.

I think a big part of this is that it finds itself in the middle of a weird dichotomy. These are people who build visuals and user interfaces, and when they end up using a UI tool that doesn’t meet their muster, they tend to complain about it. For whatever reason, GIMP hasn’t met this muster. (I remember when I tried to give it the old college try on my Spectre, a problem that kept cropping up was that for years, it struggled with HiDPI monitors—not something you want to deal with on a graphical editor that presumably is going to get used in high-resolution settings.)

If you read threads trying to describe GIMP’s UI issues, they inevitably go in one direction: “Well, can you actually cite any issues, or are you just used to Photoshop?” That sort of standoffish approach to legitimate concerns about an interface tool that is buried in decades of cruft, that deserves a degree of modernization, just ticks people off.

(Here’s a secret: Photoshop also deserves this sort of reinvention. Both suffer from the same problem: ancient code bases full of complicated features, all of which long-time users will complain about if you change. People coming from the commercial software realm are just more used to Photoshop’s version of this cruft, and Adobe can spend money on market research to fix some of it.)

This layout, if reconsidered, would be an obvious opportunity.

What could be done to improve this state of affairs? One place to start: Currently, if you press the “/” key in GIMP, it pulls up a menu that lists literally every command you can do to an image. (Photoshop does not do this, but many other apps, notably Google Docs, do similar things.) If I were GIMP, I would not bury this in some random menu, but make it more central in the interface, using that as a way to make graphical features more accessible to users who aren’t very familiar with the GIMP interface, with maybe a handful of basic commands as a starting point. Odds are, the next thing you’re going to want to do is add a layer, for example—make that easy! To some degree, it can feel like training wheels—but the thing is, training wheels can offer a larger portion of the population work a lot faster, and it would be an easy way to make the interface more accessible without changing a lot of existing features.

I’m not the only person to suggest this might be a UX path forward for GIMP—the site Libre Arts suggested borrowing from Dune 3D, a FOSS CAD application. Another useful idea author Alexandre Prokoudine has is to offer more obvious hints in the interface.

You could make arguments about whether doing certain things would make a larger portion of the audience more comfortable—for example, I think the decision to emphasize brushes in the top-right of the default interface doesn’t make much sense because that is often not a feature that people editing photos heavily use. The contextual information is buried on the left side, in a less prominent place, meaning that important details are not in an obvious place. This is easily changed, but a simple, clean rethink of the defaults could help ensure that you’re not scaring people away out of the gate.

A more recent open-source graphical tool that I’ve been using lately is Penpot, which is web-based and is essentially an open-source take on Figma. It isn’t perfect, either, but I think its pacing and willingness to experiment with interface could be a North Star for projects like GIMP, just as Figma itself has been a disruptor to Adobe. I think Penpot’s interface is also friendlier for newer users, in part because of its lean on context in ways similar to Dune 3D.

Another challenge that these tools often face comes down to scope. GIMP, having been active for a quarter-century at this point, is at a point in its history where it has to not only consider the potential growth of its audience, but what it can do to protect its existing user base—on top of what it can do to keep up with modern trends in usability. It’s sort of in a damned-if-you-do-damned-if-you-don’t lifecycle stage. Krita, perhaps the primary competitor to GIMP as FOSS image tools go, has neatly sidestepped this issue by focusing on its value to digital artists, rather than trying to be all things to all people.

For me, I think part of my frustration is that a legitimate attempt to modernize the interface, Glimpse, didn’t really get the community support it deserved, in part because of a name controversy that alienated portions of the community despite being fairly well-intentioned. The name debate is very obviously the third rail in this discussion, so I’m going to leave it alone. But that said, it is disappointing that Glimpse never got its time in the sun, because GIMP needed something disruptive like it.

GIMP is full of menus and options that are buried in nooks and crannies. It is a do-everything tool, and like numerous other do-everything tools, if you get used to it, it’s fine. But if you’re new to the tool or trying to jump in from something else, it can be a real challenge.

I think, if I were coding an image editor from scratch in 2024, I might take inspiration from sources outside the GIMP/Photoshop/Affinity trifecta, as those approaches have been done already. There are some interesting image editors out there that don’t get the attention of the primary three: Pixelmator, for MacOS, recently gained video-editing capabilities, for example, making it an excellent partner to Final Cut Pro (and a probable acquisition target for Apple, if it was so interested). And the success of Canva, despite being built for a more simplistic approach to image editing, is such that I feel that there is probably an opportunity to see if its approach could work in more complex image tools.

But I want to make clear the opportunity if someone can get this right: Adobe’s dominance in the image-editing space is at a very precarious point in the wake of the regulatory decision that prevented it from acquiring Figma. With their dominance in the design space now limited, they are likely to see their Creative Cloud moat dissipate over time, at least among small-scale users who don’t want to spend a full week’s paycheck on a software license every single year. If someone can figure out a way to build some real competition for this legacy tool, it could be beneficial to computing at large.

If GIMP is a mature tool that doesn’t make sense for this, fine. But I hope we see someone take the opportunity.