Hacker News new | past | comments | ask | show | jobs | submit login
GIMP 3.0 has a release schedule (librearts.org)
130 points by pabs3 on Nov 24, 2023 | hide | past | favorite | 141 comments



A while back I was working on an indie game dev project and I wanted to see if I could do it without using Photoshop and in my naïveté I downloaded gimp expecting it to be an easy switch over. It was probably one of the most frustrating experiences in software in my life, I knew what I wanted to do but it seemed like gimp was going out of its way to make it as difficult and unintuitive as possible. I was literally getting angry in front of my computer which never happens.

My solution in the end was to download Krita, without using any tutorials I was able to figure it out and was working on assets and textures with very little friction.

Now days Photopea, a web based editor is much more useful and less frustrating than gimp.

I still think of gimp as one of the worst UX experiences in my life and it makes blenders interface seem amazing in comparison which is sad because blender too has an awful unintuitive interface that keeps getting “fixed”.

In the end, what the gimp team should focus on is the UX of gimp and make using it not feel like pulling teeth.


I think it's just what you're used to. I mainly work with GIMP, but have to switch to Photoshop for CMYK every once in a while. For me using Photoshop feels like pulling teeth: everything is hidden, I can't find tools to do what I want, and the tools I can find work unintuitively.


> I think it's just what you're used to.

If most of the mainstream programs work the same way, there's an age old set of UI/UX axioms for developers of competitor programs:

1. The users aren't wrong, <<you are wrong>>.

2. If you think you aren't wrong, <<you're still most likely wrong>>.

3. And if you still think aren't most likely wrong, <<show me your usability studies>>.

Open source programs basically have 0 usability studies because they don't have money for them, so this argument basically dies at step 4:

4. <<Shut the f** up user, I'm giving this away for free>>.

:-)


> Open source programs basically have 0 usability studies because they don't have money for them

For a lot of projects it's not even the goal to be usable for random people: it's primarily something by the developers, for the developers. If it also happens to be useful for others: great. If not: that's okay too.

Of course other projects do have larger goals beyond just that.

GIMP developers have given kind of mixed signals over the years about what kind of project this is.


This is an important point. People often say "you need to do X to get a large user base", which is basically "capitalist thinking" where the goal is maximizing profit (via large user base).

But open source projects / devs don't have these incentives, they don't get rich out of a large user base. They often develop for fun, scratching their own itch, fulfilling some specific vision. Having user base is still important (testing, bug reports, ideas, PRs, validation/ego), but I think it's rarely the case that OSS's project's top priority is the largest user base possible.

Sometimes you even get sort of gatekeeping devs preferring a smaller, more advanced (dev) user base, since a larger user base composed of basic users bring more headache (support) in comparison with the benefits (quality code contributions).


> "But open source projects / devs ... don't get rich out of a large user base."

Nah. Usually all they often get from a large user base is the frustration of dealing with entitled users who don't understand the first thing about open source, and who sees the dev as nothing more than their personal "code monkey", there only to make the project an exact clone of some proprietary software that user is familiar with.


Photoshop users are not gimp users though.

You need to listen to you actual user base and make sure the data is representative.

As a Gimp user, I sure think that there is some stuff could be improved but "make it more like Photoshop" is not what I need.


I’m a Photoshop user who also runs Gimp (I typically use Linux but have a Windows partition and a Mac laptop). Gimp is totally fine for miscellaneous day-to-day basic stuff, but if I’m going to settle in and do any serious image work for more than 15m I’m switching to PS every time. There’s just enough friction in the Gimp interface that I feel more focused using Photoshop, and at this point I’m pretty confident it’s not a question of familiarity, as I’ve used both regularly for years.


See, that's the thing. People dismiss this as being "just Photoshop users". No, it's not, people listed Corel, Krita and 3-4 other photo editing programs here. And those are probably used by 95% of people editing photos.

So Gimp's user base, while existing, is minuscule.

Your is the other common defense, the Emacs defense. "What about the existing users?" (All 4 of them? :-p)

Yes, even the worst programs in the history of mankind (not saying Gimp is horrible, just going for reduction ad absurdum) have their loyal users. That doesn't make them right.

This argument is flawed, in my opinion.

Gimp receives a lot of TLC from its developers. That TLC would help more people if it were better guided in the UI/UX department. My 2 eurocents.

At the end of the day, the Gimp devs can do whatever they want.

But my bet is on this discussion happening in much the same form in 2050.


> "If majority is always right - let's eat shit... millions of flies can't be wrong" - Waldemar Łysiak

Measuring software quality based on number of users is such an insane take. Just imagine applying that to other fields. So Pop music charts represent the peak of musical artistry now? When deciding what laptop you gonna buy you google "Most sold laptop" or what? The most healthy food is the one most eaten?

People are different. They have different needs, different experience levels, different cultural backgrounds and so much more. There is not that one piece of software that is best for everyone.

If a piece of software does not work for you, that is fine. Not everything needs to be for you. For someone else it might be exactly the tool they need.

I absolutely get why people dislike using Gimp but I don't get why that means it needs to change for them. Like for me, yeah Gimp annoys me sometimes but we get along, other photo editing programs annoy me more.

You might not get much value out of Emacs but even if it is just the best tool ever for four people that is still pretty nice. (Honestly, there are dozens of us, dozens!)


There is a point where measuring the user base has merit.

If I wanted a laptop I would certainly buy the laptop bought by most people with the same needs.

For tools like gimp a large user base is beneficial since people wanting to learn how to do photo editing of graphic design will google how to do that and the tools used make a difference, if the majority of users are using Photoshop and Illustrator unfortunately most of thr tutorials and courses will be tought using that and the industry will continue using it.

If however your UI/UX isn't abysmal and you can effectively teach first principles using Gimp, courses and tutorials will come and so will the users.

There's a reason blender became as popular as it is, it was good enough in the beginning and with more users came more funding and more/better developers and now it wouldn't surprise me if it's the preffered tool in the industry.


You forgot a favorite reply to criticism of open source projects, namely "If you don't like it, fix it yourself, here's the code." In the past I felt that daring to criticise Gimp was equivalent to criticising the idea of Open Source itself.


Ah, that one, too. Sometimes even if you're:

A. A developer

B. That is familiar with the software stack used by the application

C. Has enough time and energy to investigate the root cause of the bug or a proper design for the feature

D. Has enough time and energy to design the fix/feature, implement it properly, write tests, jump through all the hoops

E. Send the pull request

... sometimes the PR still languishes for years or the nightmare scenario, is discarded during a big rewrite that doesn't actually cover this requirement...

So even the code being out there isn't a silver bullet.


Sometimes people really do that. There is a "UI reskinned" Blender fork called Bforartists.

Whether it's a good idea is debatable.


Sending a PR is the easier part of development, the real work is maintaing that code over the years / decades.

Connected to this is the fact that many PRs are kind of misaligned with the project vision, that's at least the main reason I'm rejecting PRs.


Then don't just send a PR: maintain your own fork.

I think that's the bigger part of open source. You can fork it, change it to your needs, and not give a damn what anyone else thinks about your changes.


Oh, that other open source trope, I should publish a book.

Because maintaining an entire fork of any non trivial software is... trivial :-)

Let's just admit that these are complex problems and frankly after watching the hype for almost 20 years, open source proved to be an alternative and a good refuge for many things but on the end user side the early hype until 2008 or so was 80% wrong.


If you truly believe it to be trivial, you should just do it rather than complain.

If you don't, then you can't justify putting more work onto the maintainers of upstream for changes they may not even care about.


Are we still talking about Photoshop where the combined transformation tool is absent from the toolbox and needs to be called from the main menu instead?


Uppercase is strongly discouraged at HN, suggest you switch to italics.

https://news.ycombinator.com/newsguidelines.html


Changed emphasis.


Fully agree. I was playing in the early 90ies with warez CDs and played with most photo editing packages at the time as a poor teenager.

Photoshop for me was actually the worst software package of the bunch from a casual user, and never understood why it was so popular. My impression was that without plugins it was next to worthless.

As I migrated to full-time linux early, gimp wasn't perfect to what I was used to, but it was pretty ok. The ui is a bit different, but nothing extreme. I really don't see the parallel with blender: blender today still requires a minimum of RTFM to get going, while you can do almost everything with gimp just by trial and error.

I assume there are nuances you encounter as you get more deep into photo editing. But as a newcomer, and my experience with greenfield colleagues to both, is that this is just an internet trope. You get used to one, and complain about that "the other is not the same". This is the dumbest form of complaining ever.

I've attempted multiple times to use krita for example after reading HN and seeing the praise, and while I agree it has a somewhat more intuitive ui, for actual photo editing (that is, not drawing), I still prefer gimp.


Blender's peers are products like Maya, 3ds Max, Modo. Go give them a spin and tell me about Blender.

The domain is much deeper than photo editing and it appears to defeat an entirely accessible UI. Blender has kept pace _with its peers_ as a world-class 3D package.

I don't think GIMP comes anywhere near to passing the same test. But, Blender has been showered with money and other support.

What surprises me the most is that there has been no big rally for a real FOSS answer to Photoshop. Animosity toward Adobe is _strong_ and has been for a long time. There's no reason there couldn't be a convergence of moneyed interests in a competitor.


I've been using blender as soon as the first OSS release, and I never considered the UI to be "bad". Learning curve was much steeper than it is today, but it was very efficient after being learned.

And that's the main point: I'm not pitting blender's UI choices against another product just on the merit of being accessible from the uninitiated or looking at another product. That's the kind of comparison you make if you want to make a clone of another product.

It would be the programmer's equivalent of saying "I don't understand why vi is not more like vscode".


Photoshop was not my first image editing program. I used to use Corel back in the day and moved to PS when it became industry standard. Unfortunately I think gimp baby duck syndrome (the is an argument gimp users like to use against other without irony) because why can I jump between a handful of image editors without issue except gimp? Gimp is the odd one out unfortunately.


I went from Photoshop to Gimp (and other image tools) a long time ago, and it wasn't that bad of a switch for me. Things were in other spots, but it was basically all the same idea. I've used both of those and others since then, and they're all basically the same.

So yeah, I 100% agree that people who hate 1 or the other are mostly just used to what they've been using, and it's not about how bad the software is.


> "Things were in other spots, but it was basically all the same idea. I've used both of those and others since then, and they're all basically the same."

I've read or watched PhotoShop tutorials and adapted them to Krita and GIMP before, so you're pretty on the nose with your assessment there. They're enough the same that I (who is not a professional graphic artist by any means) am able to learn from tutorials for other software which does a similar task and easily enough use that knowledge in whatever tool happens to be in front of me (once I find the related features in the menus or toolbars). Just gotta be adaptive to the differences, and aware of the similarities enough is all. It actually holds true for many other softwares as well, even all the way up (down?) to operating systems. There are more similarities than differences. Focus on what you instantly recognize as familiar, and start there, and then explore the other bits, and one tends to do just fine.


Same, I grew up on GIMP and it works great. Tried Photoshop and it was painful. I've had similar experiences with all complex GUI suites, like DAWs and IDEs, each one is different.


Bear with me and take a breath :)

If you think with a cold head about it, do you believe the frustration came from the fact that Gimp has made actually bad choices about placement and design of its features? (tools, menus, options, etc...)

Or would you say it's just that they took a different path, not worse nor better, just wholly different, and that makes things more confusing for people who are used to, and have the different thing made by Photoshop, ingrained in their minds?

(btw your mention about Krita was very spot-on, because otherwise conversation usually ends up revolving about how Gimp is FOSS and thus they cannot afford having a full time designer working on the best possible UI... but in this case it seems clear that Krita has been better for you, while also being FOSS)


As someone who's used gimp for more than 10 years because it covers all my needs without issues, and I can work reasonable efficiently because I've already memorised its UI and quirks...

Yes, the frustration most people have is because Gimp has made actually bad choices about placement and design of its features, not just that they took a different path.

At this point I've got muscle memory for many things, but I still sometimes find myself frustrated about the UI being so bad.


I'm in need of photo/image editing rather rarely. I used to pay for Photoshop years ago but I' was a total beginner with it and since I use it for 15 minutes per month, I can't justify the cost of a monthly subscription.

My needs are rather basic: 1) work related: combine some screenshots, crop some part of them, add arrows and text, add a rectangle here or there, get the pixel value of a specific part of the image etc. 2) personal usage: basic photo tuning (fiddle with contrast/sharpness/color balance) or doing some weird mashups of different pictures (cut out stuff from one image and add to another, add some layers, draw some pixel here, remove some from there..)

I've tried various different free alternatives and while other have their quirks and pros and cons, I can honestly say that gimp is _by far_ the worst in terms of UX. It's not even a single thing that I can point out that's the cause for this. It's a combination of many things. Unintuitive placement of features, simple actions that require a huge number of clicks and steps to complete, extremely slow vs every other app (on a M1 laptop), cumbersome UI controls, unexpected side-effects of performing some UI actions etc. If I would draw the usability of various apps on a graph, then every other app I've tried would be clustered together somewhere at one side of the axis and gimp would be on the other side, alone. The magnitude of the UX clumsiness is just so huge compared to everything else :(


This response is for all people who replied to my GP comment, not only for this one: thanks for sharing your experiences! I'm too in the camp of "15 mins a month" (even less) so I always assumed my struggles with GIMP were caused by mere lack of experience. It's enlightening to see that people consider some design decisions "objectively" bad.

I guess all this means that maybe GIMP should really focus on getting help with this. TBH I'd just try to copy Adobe here. They can and I'm sure have spent thousands just to have people thinking about the best placement for tools and widgets. All that can be leveraged by a FOSS tool and advantage from some of the well thought out decisions made by Adobe's professional designers.


> If you think with a cold head about it, do you believe the frustration came from the fact that Gimp has made actually bad choices about placement and design of its features? (tools, menus, options, etc...)

Not the GP.

GIMP was my first image editor (granted it was almost a decade ago so it might be much better today). And the answer is yes. Yes a thousand times.

Once I switched to CSP+PS (later Krita+PS) it's quite hard to imagine going back to GIMP.


There are some legitimately questionable design decisions in GIMP’s UI.

One of my favorite is the layer palette which contains what looks like a bog standard list/table widget like a user has encountered countless time across desktop platforms and software since the 90s.

That illusion breaks when you try to use it as a bog standard list widget — no, you can’t multi-select with Shift and drag the selection around, GIMP has its own unique process for that for no good reason.

Things like that will cause friction even for users who aren’t coming from Photoshop, being different from established desktop UI conventions.


“ do you believe the frustration came from the fact that Gimp has made actually bad choices about placement and design of its features? (tools, menus, options, etc...)”

Yes, yes absolutely without a doubt because I can use photoshop, photopea, krita, paint.net and maybe Corel if ever use it again without issues. Gimps interface is objectively bad.


I learned photoshop from nothing a lot faster and easier than I learned Gimp. Paths for the same process aside, I think Gino absolutely could learn from photoshop when it comes to making things intuitive. Even if doing the task itself is done differently.


I've been using GIMP for well over a decade, and the UI still deeply frustrating.

When I see what Blender has achieved over that time, I wish GIMP could have done the same.


I not only believe that (and I first used GIMP more than 2 decade agos), I believe that the entire project has a culture that is proud of being crude and inaccessible that is shown most strongly not by their bad UX but the fact that they still haven't changed the awful name.

The entire project gives intense vibes of "we don't care what you think" that have made it the poster child for many of the worst parts of OSS for a long time.


As someone who went through the same experience it'd take a lot for me to even think about touching GIMP again. I'm sure that it's a viable Photoshop alternative but any functionality is obscured under a web of riddles.

Blender's UI used to be nightmarish and backwards for any newcomer (why would they make right click select things instead of left??), but as someone who's tinkered with it over the last decade the UI changes introduced in 2.8 finally made things comprehensible. GIMP needs much the same, but I doubt it'll happen until a wider community (or some industry) incentivise it.


Agreed. I am willing to pay for a Photoshop license out of pocket than getting paid to use GIMP. Every time someone says "use GIMP" I wonder how much time they have spent designing/drawing anything and what level of experience they have. It is not a matter of different UX -- it is a matter of whether the UX makes sense at all. GIMP stands in your way of productivity.

And I notice UX is a big problem for GNU software from that era, and to some extent community driven open source software in general. Of course there are notable exceptions like IINA, but often it is like "we have this feature" and are done with it and nobody bothers to work on improving the UX.


Not suggesting your comment is not without merit. I can understand going from Photoshop to Gimp can be frustrating but I think we can pick many examples that follows the same pattern. I guess it is more about what you are use to than anything.

I am a software engineer, and familiar with Windows and GNU/Linux systems. There are a few good development software I use. However, If I was handed an Apple Mac Laptop for working on a new project, I am sure it will be a very frustrating experience for me. It is not because the 'apple experience' is bad. It seems like a well-though-out system but I am sure to be annoyed and wasting time trying to figure how to do things in the OS as well as the software I am using.

Now I do not call my self a graphic designer by any means. I do know enough of Gimp to do a number of things. I guess I am good with a working knowledge. I am sure if I was given Photoshop I would be frustrated trying to do X, Y and Z.

Maybe there is a bit of bias showing on my part. To have a package like Gimp (or Blender, Inkscape, etc) that is Free Software is just a huge win. Of course, maybe I can say that considering I am not an expert designer that needs to use these types of software 24/7.

With GIMP (and Blender) I am mostly using the keyboard shortcuts. Maybe this helps balance what is annoying UX? I dont know.


Going from photoshop to gimp is not like going from windows to mac os. Mac os is annoying in the fact that its restrictive because it wants to be safe and protect the user from themselves at all cost but it is still very very intuitive.


Maybe I'm the odd person out, but I don't find most Apple UIs all that intuitive. The iPod Classic was fantastic but we're a ways past that.

I've used Windows, and Linux on the desktop for a lot of years, and struggled with my new Mac this year. The last time I used a Mac was System 7.

Tons of normal options are hidden and undiscoverable, except by accident or tradition. Maybe my annoyances are just because they feel restrictive, as you say. I'm not sure.

Calculator on iPhone is an example. Vertical for standard, horizontal for scientific. Simple once you know, but I only found that because I put my phone down on my desk with it open, but back to the Mac.

Please let me stick windows to corners. Do not make me hover a UI element with my mouse to see more options. Mac's built in window management is less powerful than Windows 7's.

If it isn't what you're used to, a program being open but not having any windows is a strange way to work.

And how do I show hidden files in Finder? I was convinced I just can't find the menu item, but if you Google it, you get some suggestions for command line options, and running Automator tasks.


My point is not just about the difference between Windows and Mac.. but the software I need to use to get things done.

I use specific software to write programs.. to code in. Some are not available on Mac so I have to use whatever is "best" on that platform. I am sure to come across many issues.

Again, it's not that one software/tools is better than the other.. it is really because I am more familiar over one than the other.

That was kinda my point.

For someone that is good at graphic design - someone who has been a "photoshop guy" for many years and very good at what they do.. I can understand the frustration using something like Gimp.

Yes - despite not using Photoshop for many years, I am sure it is better than Gimp but I also think Gimp gets a lot of stick where it is not needed or deserved.


> I still think of gimp as one of the worst UX experiences in my life and it makes blenders interface seem amazing in comparison

I never get the comments like this I see about Blender. After the 2.8 redesign, on which the major versions after it have only improved, I felt that the program is really good to use. Well, I'm a Blenderer, I know the basic Blender keybinds. If I didn't, the experience would be much worse. But once you know about shift+a, tab, and x, using the program should be no problem. I feel that the UI is pretty intuitive and internally consistent.

But this is just my experience... Maybe some have a harder time forming intuitions.


I can use 3Ds max with pretty much one hand and when I zoom with the mouse wheel it zooms where the cursor is an not the center of the screen. The metaphor I like to use for blender is it is like riding a bike with the seat missing, you can get to where you need to but the ride is much less comfortable. Gimp is like a bike with no seat, you have to peddle in reverse and the handle bars also work in reverse.

I’ve used many other 3D programs, even some that don’t exist anymore like XSI so no baby duck syndrome here but I just find blender just not fun to use. Even milkshape 3D back in the early 2000s was way more intuitive than blender.


It is possible to configure Blender to work this way, it's one of the first things I do on a new install. If I'm remembering right it's under your Navigation preferences.


I hate Maya ever since I tried Blender 2.8


Your message intriguing... Can you point to some specific examples of frustration? I've never used Photoshop, but as an occasional user of gimp it seems quite natural and intuitive to me. I'd like to see what could be greatly improved about it.


I agree - I have used the gimp since 1998 and when using other apps (e.g. Affinity) I often return to Gimp and Inkscape because I understand their UIs best.


Photopea is an embarrassment to GIMP. I had a very similar experience trying to use GIMP once. Using Photopea was seamless and easy and it hasn't lacked any feature I've wanted yet.


I'm a happy user of GIMP whe also happened to have switched from Photoshop. I even managed to use keyboard shortcuts cuz it's so intuitive. I never make efforts to learn them, not in gimp not elsewhere, but the description was straightforward that I would have a hard time not learning it. (P for paint/pen, M for move, T for text, Shift+E for erasor)

Now I like Krita too, but mostly because of the pen library. Krita is actually less intuitive than GIMP imo (Why are there different moving tools for vector and raster layers? How do I enlarge my selection? How do I flip the canvas for correction reasons? I know there's a shortcut for that, I pressed it by accident.)

Would I call Krita bad UX? Not really, it has a place in my workflow. But I prefer GIMP for its power and since I can get around it well. They have been obviously stretched thin for a long time, I myself have been waiting for features and improvements in GIMP for years. But really, I think people underestimate the complexity of these tools and quick to label them unintuitive. I don't think you can overhaul the UI without removing some or many features GIMP currently has.


From what I understood about GIMPs development is that they have to purposefully make things "different" from photoshop so as to not get sued. Which makes going from PS to GIMP incredibly painful (I too, tried a few times).

I think if you come into GIMP fresh without using other tools it is probably beter.

I usually ended up relying on paint.net back in the day instead of GIMP on windows tho.


> From what I understood about GIMPs development is that they have to purposefully make things "different" from photoshop so as to not get sued.

I don't know how these rumours get started but that is simply not true.

GIMP does some things differently and then it does some things the same (because it's just how this type of features work). And then some things were purposefully verbatim-copied from Photoshop (think Heal tool and Perspective Clone tool) because they looked like a good idea.

Personally, I find it incredibly frustrating that this type of conversation always boils down to "If you are asking for explanation why copying Photoshop is good, then you are flat-out refusing to change UI and purposefully making things different".


Yeah GIMP has it's issues but GP is influenced by vendor lock-in.

Better example is Inkspace (because it is more comparable to closed-source alternatives): transition from Corel Draw to Inkspace is much easier then from Adobe Illustrator.

I feel GP's pain, I'm in same situation ...but there are different UX/UI paradigms.


> Yeah GIMP has it's issues but GP is influenced by vendor lock-in.

He mentioned Krita as being better 2 paragraphs in, how does "vendor lock-in" make any sense here?


I also can’t stand the gimp ui/ux. I haven’t tried it myself but I just googled this[0].

[0] https://github.com/Diolinux/PhotoGIMP


For me, the relentlessly changing gimp gui is what drives me nuts.

I need to use it, every so often (once a year, but for a week or more), and every couple of years, options move, settings change, everything.

And it always does seem to do things in a weird way.


I'd be very curious to learn which parts of the GUI you mean, I think in 15 years of using GIMP, I've never had this problem, though granted, I never used it at a professional level.

The only things that I can think of is some more tools appearing on the same buttons in the interface, or more options in the Tools/Filters menus.


Used Photoshop since its really early days, but since I moved to Linux as an OS, it's really hard to find an alternative. Eventually, I fall back to Photopea, as Gimp is just a PITA to work in.


For a quick edit I am using Affinity Suite in bottles/wine: https://forum.affinity.serif.com/index.php?/topic/182758-aff...

If it gets more involved I boot into windows.


> In the end, what the gimp team should focus on is the UX of gimp

Oh I hate to break it to you, but you are part of the problem ;)

Please Gimp, leave the things exactly where they are. Thank you.

Learning new UIs just sucks. It is the way it is.


I really wonder how many people still believe Blender should have stayed in 2.79 UI paradigm today.


Certainly not I, because nothing of real value was lost during the improvement phase. Everything that made Blender good is still there today. Only genuine improvements and new features have been added.


Because of people like you gimp will never reach its full potential and wallow in mediocrity. Gimp can have a blender style redemption arc but people like you will fight tooth and nail to prevent that from happening because five guys like the UI as it is the thousands who say it is horrible be damned.

Learning bad UIs sucks and thats not the way it should. I relearned the unreal engine ui three time udk/ue4/ue5 and it never sucked, in fact it was a lot of fun.


Sometimes it feels like people want to keep GIMP unintuitive so as to keep their little club as exclusive as possible. How else would they be able to look down on people and say "well you're just not clever enough to understand it".

I'm very happy to see blender fix itself and become so popular as a result. Tons of cool stuff coming out of that community these days.


I don't see how catering to the latest UI fads would improve anything. Like, it has to be way better to justify learning a new UI.

I might be an outlier but I more or less can't relearn a new UI after it got into muscle memory. It is stuck for ever.


It's not about fads. Good UI is evergreen, even if it looks a little dated. GIMP UI is everbrown. You could dress it up in the latest colors and styles and it'll still be awful to use because the foundation is warped and misshapen.


Honest question. Am I missing something? Is it that bad?

I make drawings for posters sometimes with the aim to "look pro" for a local party. I am quite bad at it, but it looks ok. Like, do high level users have different needs?

I just fired GIMP up to see if I am missing something and I don't really see what.


It's less that high level users have specific needs and more that the workflow doesn't exist for high level users to worl efficiently. You see this a lot when you listen to what people actually do with GIMP. They either only use it for small, quick jobs or they'll start with GIMP and migrate to another software when it grows complex.

That said, GIMP is lacking in two particular feature areas, its selection tools and its completely non-existent nondestructive editing tools. The transform tools are also a lot less wieldy, and GIMP's text tools, on top of being unwieldy again, have rendering glitches in their antialiasing code that gives everything a green tint on my machines. There's many pain points like that where things are lacking, more difficult to use, or subtly broken. Fine if you don't dive deep, but increasingly frustrating if you do.


> GIMP is lacking in ... its selection tools

Could you please elaborate? Is this about fast selection of complex and semi-transparent objects (e.g. hair)?

> have rendering glitches in their antialiasing code that gives everything a green tint on my machines

This is a Cairo issue though. See here and further: https://gitlab.gnome.org/GNOME/gimp/-/issues/8455#note_15219...


Yes, hair specifically is understandable but other complex objects are just more finicky to select in GIMP, and using those selection tools has a higher friction. I often find myself fiddling with tool settings in GIMP, and those tool options panels are painful to adjust.

Regarding rendering, wherever the issue is in the stack, it's been over a decade and it's killing GIMP's usability for an extremely basic task, regardless of whose fault it is.


The Paint Select tool in 2.99 was one of the attempts to make selecting objects simple. Unfortunately, it's unfinished and the original developer is no longer active.


Blender's always been great. Just different. However, it's only been getting better over time, and that thrills me to death, because they didn't kill the things that made it great in the process of improving it. They just made various things better, and added new stuff that made the whole thing overall better.


Yeah, that kind of attitude is surely going to help the development of GIMP.

There are lots of alternatives to Photoshop these days, free or paid, open source or not, and most of them are vastly better than GIMP in many ways. Yes I have used almost all of them. I won't be surprised if GIMP is forgotten to oblivion in a few years and only mentioned in forums like hacker news by amateurs who for some reason loves this abusive relationship.


Thanks for mentioning it. Gimp is okay, except for the missing CMKY. There was once a CMKY support plug-in, wasn't there?

Do you know what I miss? The inkscape competitor Xara Xtreme (XL). Gone..


The separate+ plugin isn't maintained, but GIMP 2.99/3.0 has late binding for CMYK (basically, work in RGB, export in CMYK).


100% with you i kept hearing about. gimp being the defacto FOSS standard for image editing/creation, but holy crap its like you need a degree to get started with it


I agree. I also think this about Figma though. I’ve been using it a fair amount for a few months now, and while it’s powerful, a lot of it is very unintuitive.


Should have gone with Krita


Krita is a great piece of software for certain tasks, for sure. I use it alongside Blender, GIMP, and Inkscape. Well worth keeping it installed and ready for those times it's just the best tool for the job at hand.


Usually when there is a post on HN about some open source project with a GUI there is a deluge of comments complaining about it for one reason or another. Even though it's free. I don't understand why.

Gimp is a wonderful piece of free software that I use often, and I greatly appreciate the work and time the maintainers are putting into it.


It's a pain point for people who want to see it succeed. It's like the mole guy from Austin powers. Just because it's free doesn't make it good enough. It's not hard to pirate Photoshop, which is also free. Krita is a better Photoshop replacement than gimp is.


Krita is new, GIMP is legacy.

It's a hard and very time consuming tax to replace UI within a legacy application.

GIMP has existed long before the entitled generation of Linuxers who wish for everything to be handed to them on a golden plate.

However sometimes that plate is covered in muck that you need to clean yourself.


> GIMP has existed long before the entitled generation of Linuxers who wish for everything to be handed to them on a golden plate.

I've been there before the golden platter, and the Gimp UI was mediocre even back then. And even underneath, Gimp is at best bronze, I wouldn't call it gold :-)


Krita isn't that new. Between using another program and fixing the gimp UI, most graphics people aren't going to learn to program because they love GIMP so much. It's nice that people still try GIMP, once everyone 'knows' GIMP's UI sucks they'll just use krita for everything and it'll be forgotten.


Krita is old, Blender is old, both have improved themselves with age and thougtful change. GIMP has done nothing but fossilize with age.


> Even though it's free. I don't understand why.

Because if no one would ever complain, there is a lower chance it will change.


If complaints were enough to fuel change, GIMP would have had a truly amazing UI by now.


How a designer can revamp its interface? I mean that’s not a web page where you can basically invent anything you want. Should one need to learn basics of GTK, or whatever it’s made with? If so, how to do it then?


You don't understand why, when software has a bad interface, and said software is under discussion on a web forum, people discuss the bad UI? I find that very hard to believe.


Yes the common argument is about the UI and which one where that Blender and Krita are both the new kids on the block. They were started from scratch with the ability to have a clean UI.

GIMP not so. Its an ancient carved releic from the days of GTK2. If you've ever worked with old X you'll know.

Back then UI was in a different world than to what we have now. It's unfortunate that the UI of GIMP just never has been rekindled to the newer century, but that's a you problem not a them problem.

However saying that, nothing is stopping you from spear driving the new GUI design. It's open source after all.


> Yes the common argument is about the UI and which one where that Blender and Krita are both the new kids on the block. They were started from scratch with the ability to have a clean UI.

Yeah... talk about defending a lost cause.

Krita was launched 18 years ago. That's still a lifetime in software years.

Gimp was launched... 25 years ago. 7 years is a reasonably long time in software years, but it does <<not>> excuse a bad UI.


> Krita was launched 18 years ago.

The project started under a different name and team in 1998. But it wasn't as active as GIMP until the current project lead discovered and pretty much saved it from slow death around 2002.


I'm not sure what your point is (other than the requisite "it's OSS so fix it yourself" brain-dead take). Like I said, the UI is bad. We all know the UI is bad. Whats the point in GP expressing fake surprise when people say "the UI is bad"?


[flagged]


Nothing, I just keep using Photoshop. No skin off my nose.

You're*


Exactly. It's no skin off your nose. Which is why I don't understand the drive of commenters who try to take skin off the nose of the maintainers by bashing their free product.


[flagged]


You're*


I'm no Gimp user but it is tiresome when there is post about Gimp - people are just bashing it. You have found Gimp with bad UI and posting about it? Welcome on the internet!


>GIMP not so. Its an ancient carved releic from the days of GTK2.

If memory serves, GTK stands for GIMP Tool Kit.

That is to say, GTK exists specifically to give GIMP a GUI.

https://en.wikipedia.org/wiki/GTK#Linux


Even so, the "new kid on the block" Krita that's mentioned there is... 18 years old. The comment makes it sound like it's an Electron/React app from 2021.

Krita is old enough to drive, to drink (in civilized places), to vote.


Not sure if you replied to the wrong comment.

I actually agree with y'all that GIMP's GUI is god damn horrendous, and that kind of extends to anything that GTK is involved with or used in. Something about GTK at a philosophical level does not lend itself to a good, human design.

Anyway, I'm just pointing out that GIMP came before GTK and that GTK exists first and foremost to serve GIMP. I hate when history is altered, whether out of simple ignorance or more ulterior motives.


I was adding to your reply to the GP comment.

Also, for Gimp - Gtk, that was true for Gtk 1. Gtk 2 went beyond this to serve Gnome and random apps and Gtk 3 is quite far from the Gimp world.

Gimp 3 is about a port to Gtk 3... 10+ years after the release of Gtk 3.


Why won’t they port to GTK-4? Why port to already obsolete thing? If you’re so incredibly slow to implement, why not change the direction midways? I believe it’s not like porting to QT, so it would be dramatically different (if it is, I don’t know), it’s just different versions of the same library after all.


> Blender and Krita are both the new kids on the block. They were started from scratch with the ability to have a clean UI.

> Blender

What? Blender's codebase is about as old as GIMP's is. Up until recently (IIRC the 2.80 release in 2019), it was also constantly criticized for it's horrible UI/UX compared to its proprietary competitors.

If anything, Blender is a great example of UI refactoring over time gone right, and opens up the question why GIMP isn't able to pull off the same.


Blender is much older than you think.


>Yes the common argument is about the UI and which one where that Blender and Krita are both the new kids on the block.

Maybe it's a common argument where you come from, but it's a factually false argument.

https://en.wikipedia.org/wiki/GIMP

>GIMP: Initial release: 2 June 1998; 25 years ago

https://en.wikipedia.org/wiki/Blender_(software)

>Blender Initial release: January 2, 1994; 29 years ago

https://en.wikipedia.org/wiki/New_Kids_on_the_Block

>New Kids On The Block: Years active: 1984–1994, 2008–present

So Blender is actually four years older than GIMP, and a decade younger than New Kids On The Block.

And over all those years, Blender has continuously improved and evolved a huge amount, especially recently. And the developers listen to and respond to the needs of their users.

For in all those years, GIMP has ALWAYS stagnated, and the developers have contempt for their users.

>However saying that, nothing is stopping you from spear driving the new GUI design. It's open source after all.

That's easy for you to say, but what have YOU actually contributed to open source software? How's that new free web browser you wanted to write from scratch going? Stop complaining about Chrome, and finish what you called for other people to do for you for free, yourself!

When you trot out that glib old argument (see what I did there: https://en.wikipedia.org/wiki/GLib ;), you're just admitting you're wrong, have no better arguments, so you're not arguing in good faith (after you post easily fact-checked misinformation like claiming GIMP is older than Blender).

https://news.ycombinator.com/item?id=38238274

DonHopkins 12 days ago | parent | context | favorite | on: Is this radical redesign of GIMP possible now?

The one worst thing in the universe that you can do to piss off the GIMP developers is to tell them it should be more like Photoshop. It's like telling RMS that he should develop Open Software for Linux, instead of Free Software for GNU/Linux.

prokoudine 10 days ago [–]

"Make it work like in Photoshop because {argument about better usability}" is appreciated.

"Make it work like in Photoshop because Photoshop" is not appreciated.

It's really _that_ simple.

DonHopkins 6 days ago | parent [–]

Not at all, it's much simpler and much worse than that:

The GIMP developers have heard both kinds of arguments MANY times before, as well all the arguments for using pie menus based on usability and scientific measurements of controlled experiments and decades of experience developing and using user interface toolkits and applications, from me and many other people, and at this point they are just cutting off noses to spite their face, not listening to any arguments about usability or anything else, because they have decided that somebody else once hurt their feelings by asking for the exact same thing for the wrong reasons (or as you literally say, simply without extensive justification), so they got permanently huffy with their panties in a twist, and always say no to everyone who ever asks the same question ever again no matter what their arguments about usability or experience with user interface design.

[...]

https://news.ycombinator.com/item?id=38332975

DonHopkins 4 days ago | parent | context | favorite | on: GTK: Introducing Graphics Offload

The problem with the GIMP team is not that they're enormously opinionated, but that they're WRONGLY opinionated.

It's not that they're the evil party, it's just that they should stop feeling so sorry for themselves that so few people want to use their image editor because of its terrible user interface caused by the fact that they refuse to listen to their users, and it has a terribly offensive name that they refuse to change.

At least they still have a fanatical following of MAGA incel edgelords and ESR sycophants who love it BECAUSE it has an offensive name, so they still have that hard core fanbase to appeal to.

They're as self-sabotaging as RMS himself, and they don't deserve to play the victim or to have a pity party, especially when they try to throw it for themselves.

https://news.ycombinator.com/item?id=38233793

luqtas 12 days ago | parent | context | flag | favorite | on: Is this radical redesign of GIMP possible now?

NO! WE SHOULD USE RE-DO WITH PIE-MENUS ONLY :P Universal Search? Like F3 in Blender? Sure! just do it...

i studied for a year advertising and marketing degree at an university and i was at the best grades on photograph and any other class that required some creative stuff. all done in Gimp... i don't know how the late game is but certainly some company wanting the proprietary .blob of your work in Photoshop lang binaries exists. but i also think that if you are highly creative, there's so much to-do with simple tools

edit: not that Gimp can't do complex stuff but i remember clear sizing the boobs of a woman once at Photoshop and having some trouble on how-to with Gimp at home...

DonHopkins 12 days ago [–]

Agreed! ;)

I know you're joking about "ONLY", but actually, linear drop-down menus are just an edge case of pie menus with multiple items in only one direction: down. So you can also make drop-down, -up, -left, -right, and other direction menus with a decent pie menu editor, like the Blender pie menu editor add-on.

Pie menus are much more useful if users can edit and create their own, especially in feature-rich extensible configurable applications that different people use in different ways like Gimp and Blender.

Blender has great pie menu support, and there's a nice pie menu editor add-on, but it really needs a built-in WYSIWYG pie menu editor, supporting on-the-fly direct manipulation WYSIWYG drag-and-drop pie menu editing, like Simon Schneegans's brilliant Gnome-Pie and Fly-Pie.

By "direct manipulation WYSIWYG drag-and-drop" I mean that ideally you should be able to put any existing menu into edit mode on the fly, and edit the circular pie, linear, or hybrid layout directly by dragging items around to different slices, instead of with an indirect linear scrolling list or outline in a separate window.

You need to be able directly and immediately edit them as they will appear to the user, not as some abstract linear tree outline (or god forbid, raw xml).

[...]

https://news.ycombinator.com/item?id=38237328

DonHopkins 12 days ago | parent | context | favorite | on: Is this radical redesign of GIMP possible now?

Sorry, but the number of people who have seen Pulp Fiction, plus the number of people who know derogatory terms for disabled people, is much greater than the number of people who know technical biological terms of art.

If the GIMP developers really want to score an edgy rhetorical point about how society should get over its uptight wokeness and let them use any word they want whenever they want, and that's the hill they choose to die on, then how about they go all in, and try convincing society to re-signify the n-word by using it IN ALL UPPER CASE as the name of a hard-to-use paint program with an overly complex incomprehensible user interface for TempleOS, then come back to me after a few years and let me know how well that went.

Prejudice by Tim Minchin:

https://www.youtube.com/watch?v=KVN_0qvuhhw

At least the Blender developers finally listened to reason, admitted they made a mistake, and switched the left and right mouse button behavior, which wasn't nearly as offensive to as many people as "GIMP", whose name makes it kind of hard to evangelize around the school or office without coming off like a flaming MAGA asshole.

Donald Trump appears to mock a reporter's disability:

https://www.youtube.com/watch?v=mdLfkhxIH5Q

By stubbornly refusing to change the name, the Gimp developers have lost the right to whine and feel sorry for themselves about how unpopular it is and how nobody takes them seriously. Because in the intervening 25 years since 1998, 4chan and GamerGate and MAGA and Q-Anon and January 6 and Elon Musk have kind of spoiled the coolness and originality of that rebellious "edgelord" attitude.

If you have to explain to people, "I'm not really ableist, but I am simply participating in performance art to resignify a derogatory slang term for handicapped people or submissive S&M sex slaves as the name of a paint program!" you have already lost them.


> That's easy for you to say, but what have YOU actually contributed to open source software? How's that new free web browser you wanted to write from scratch going? Stop complaining about Chrome, and finish what you called for other people to do for you for free, yourself!

Slow pace but surprisingly going well, thank you. Taking tkhtml as the basis and going for a non-js approach.

> what have YOU actually contributed to open source software?

What does this mayter, are we to have a dick measuring competition? Most stuff that isn't for the normal user. A library that controls my hexa robot is a cool lib. among other things. I'm not a programmer nor developer.

All i know is that we live in a shitty world with shitty internet and from many browsers we are now down to two. Which are both as shite.


You don't understand the difference between complaining and discussion? Seems even harder to believe.


I remember being really excited about the first GIMP release that partly used GEGL [1], when a fully GEGL-based GIMP 3.0 seemed right around the corner. That was GIMP 2.6 in 2008. I've been using GIMP for over 20 years at this point.

[1] https://en.wikipedia.org/wiki/GEGL


> Once 3.0 is released, we can have a long conversation about the way the team plans their work, which is not great, and the way they finance it, which is even worse. But right now the community needs to finish off this SOB of a release.

Gimp would benefit from faster pace of development and more funding. Blender is a good example how it can work.


Krita also seems to be doing better in that regard, despite not having the massive industry backing that Blender enjoys.


GIMP ui is hard for new users and because of the extremely hard learning curve most people give up on it. People overreact about the gui complains but people that have been using it for years have blinders same is the case for most of its developers.

Similar is the case for calibre another popular open source project I have been using it from almost its inception so I don't see the problems that new users face that have grown up with a new gui rules and interfaces.

Both of these have pain points for new users but the legacy behind the code makes it impossible for the current devs to change it. So when people say don't like it then you should fork it is actually the only real way that it can move forward. And then if it is good enough change it might get back into the main fork otherwise it won't.


Reading that if there is a project that needs more hands it has to be GIMP... That being said without introducing more workload than it fixes. I am not sure how long the lead up time is to being useful. Seems to me just fixing bugs would really help them - if can be done - without causing further issues.


Is there a link to an official announcement please, not to some suspicious site hosted in Muscovy? Thank you.


Both Patreon (original announcement by lead dev), GNOME's gitlab instance (where the roadmap is) and Libre Arts (linked post) are hosted in the US.

What's causing your suspicion?


I'm continuously surprised it doesn't get renamed to GNU Imp.


Am I reading correctly that non-destructive editing, advanced colourspace support, and reliable hidpi display support are all unlikey to make it into 3.0…? Does anybody know what is expected to make it into 3.0? (IIRC the switch to gtk3 is happening at last, which is nice… even though gtk4 was released in 2020)


> Am I reading correctly that non-destructive editing, advanced colourspace support, and reliable hidpi display support are all unlikey to make it into 3.0…?

GIMP 2.99 already handles color spaces better than 2.10. E.g. you already have late binding for CMYK (but not early binding, that's post-3.0 stuff).

There's no telling if layer filters will make it to 3.0, I guess we'll see how much is done at feature freeze time. So I wouldn't be too fast to jump at conclusions, Even if it's missing in 3.0.0, 3.0.2 would have it.

As for HiDPI, the three way switch for canvas zooming suggested by Jehan here — https://gitlab.gnome.org/GNOME/gimp/-/issues/5206 — is something I hope they will reconsider. There should be a better way to handle it.

> Does anybody know what is expected to make it into 3.0?

Everything you see in 2.10 point releases (much was backported from 2.99) plus some extras like GTK3 port, multi-layer selection, complete rewrite of linked layers concept, Lua and JS bindings, and various other things. Because of backporting, separating 3.0-specific changes from the rest is not an easy task.


I used to use PhotoShop heavily and still have a lot of the UI muscle-memorized. I switched to GIMP + https://github.com/Diolinux/PhotoGIMP and found it mostly fits my needs for basic image editing.


Looking forward to the stable diffusion tools (I'm guessing they will be in there).


Krita has a number of Stable Diffusion plugins (all OSS).



>This plugin is no longer actively developed.


and no mention of wayland support :-(



This is what I came here for too. In the past I got an impression that Gimp 3 will run on Wayland as it's based on GTK 3, source: https://www.gimp.org/news/2020/11/06/gimp-2-99-2-released/


Wayland has been working for a long time on the 3.x development releases. I run a wayland only (no xwayland) system, and I have been using GIMP development releases without issue for years.


Beta builds are working on Wayland for me.


There's a difference between Wayland support and working on Wayland. They're asking about actual wayland support, not xwayland, which presumably was never and issue


The beta builds don't use XWayland.


Where did I say Xwayland? It's 100% using Wayland, works perfectly for me.


Have you tried not using a monolithic display server?


GIMP is great!


i am not sure what i heard. i am sorry if the narrator is in this thread/reads this but i could not differentiate between the narration being made by a bad robot or someone stuttering. i tried to make out the difference between the two but simply couldn't


This is because that's one of the developers themselves and not some marketing team. I understood pretty well despite English not being my primary language.


I'm not a developer really, and I'm not part of the team for close to two years already :)


I grew up speaking English in the US and didn't have trouble understanding him. The accent is thick, but it's understandable.

Anyway all of the content is also in the text of the page.


no no. i understood everything normally. the sound felt weird thats all


Uncanney valley probably comes from the sounds either being speech or complete 100% silence, dead space being cut out. I think it is understandable and clearly not a robot though.


Oh, I'm probably bad, but definitely not a robot :) Bleep.

I think there's some distortion added by the Noise Repellent plugin, will look into tweaking settings.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: