Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
GIMP 3.0 Roadmap (gimp.org)
69 points by pmoriarty on Feb 2, 2022 | hide | past | favorite | 23 comments


I still do not understand how they can construct a roadmap that, in 2022, does not lead to the imminent implementation of adjustment layers.

I mean, forget live filter layers (they are fun and incredibly useful in Affinity Photo). But adjustment layers?

You can make an argument that GIMP users have learned to get by without them, but it's hard to make the argument that the wider graphics editing community doesn't see the need.

It is quietly delusional.


Interesting to read GIMP should add live adjustment layers from Photo. Using Affinity for 4 years I’m getting more and more to the point of dropping it completely. So many nice features in theory, put in practice the number of never-got-fixed bugs piles up every day, making it a nice toy app but useless if you depend on one of the broken functions, or lost your work in one of the frequent crashes of file corruptions, or unable to save situations.

I always advocate to deliver a dependable app (fixing bugs, keep it in maintainable state) before adding nice functionality which always raises the risk of breaking something.

Affinity products are dirt cheap (for the average hacker news reader), so choosing GIMP or Photo is mostly no matter of cost, but of other factors. As even Affinity struggles with resources (to fix bugs) despite selling tons of licenses, how should the GIMP developers stem such an challanging feature?


The thing is, Photoshop has had non-destructive adjustment layers since version 4. 1997 I think. They are common to basically all other photo editors and it's really an essential part of most photographers' workflow.

And this isn't just "nice functionality" -- it's absolutely core to the compositing model. This sort of thing is built into operating system graphics layers now, and there are other, newer apps on Linux (Krita for one, I think) that can do it.

This is all the stuff GIMP users are promised for the future when the new GEGL gets integrated. But that's just like Infrastructure Week: it never arrives.

I'm not saying it is easy to implement; I don't know where I would personally start on much of this stuff. But I do think that it is indicative of the GIMP project's peculiar focus on scripting that something as crucial as the compositing engine in a photo editing application is routinely outsourced to another project.

It's such a basic function in other applications that it has to be holding back donations from interested users.

(I'm not sure where you're using Affinity Photo but on the Mac it has crashed on me about a dozen times since 2015. It does not strike me as more bug-ridden than GIMP!)


So the focus is moving to GTK+3, which came out >10 years ago...

GTK4 is already out. GIMP is massively underfunded, but I suppose that isn't really surprising.


The changes from 2 to 3 were massive. The changes from 3 to 4 are much more incremental. Not to say GIMP isn't massively underfunded (it 100% is), but once the transition to GTK+3 is done, the transition to GTK4 should be much more straightforward.


Then they need to move to GTK5 or GTK6. It's funny that G in GTK stand for Gimp but now i think it stands for Gnome.


Yep, all the open source hope has been put into Krita recently, rather than gimp.


They're for different things.

Krita is primarily for painting and drawing, while GIMP is for photo editing.


No that is not accurate. Krita is excellent at editing graphics and photos.


The features may be there, but they don't always seem obvious. I've had multiple tasks where I started with Krita and just could not find the obvious flow - reverted to gimp, done in under a minute. The last one I believe was to add transparency and remove background. I'm not the only one either apparently: https://krita-artists.org/t/how-to-make-background-transpare...

Some things are just much simpler to achieve in gimp it seems.


Whenever I have a workflow question I turn to a few YouTube guides and then I get my answers. That is a tip.


GIMP needs to have a GUI redesign and modernization. I hope it gets enough support to take it to Blender's level.


It's not the GUI that needs redesigning and modernising.

It's the fundamental goal of the project. I know it is a depressingly overused expression but if any open source project could be said to be "skating to where the puck was", it's GIMP. And they've made a lifetime's work of it.

As a scriptable editing platform for producing graphics, it's probably usable. But interactively, non-destructive editing is now part of the basic mindset of the graphics or photography worker. And this stuff just gets kicked further and further down the road.


I thought that many industry professionals didn't like blender's ui and thought it was confusing and needlessly complex?


All 3d packages have confusing and complex UIs. Blender’s UI has improved significantly over the past few years - it is now around the middle of the pack. Users care much more about features and functionality.


Blender was already an industrial product before it became open source.


Didn't Cinema 4D more or less clone Blender's UI?!


So far the comments are all resource related. So ... let's all throw them some cash.


I thought this is a good idea and actually tried.

After pressing the Donate button I'm able to donate to two sub projects.

But then on either one I can only chose a "monthly plan".

Guys, please don't do that! Just allow me to pop 20$ or 50$ to your project (which I love and sometimes use) but don't force me into a subscription like thing, which I really detest.

Maybe it's me and I was just too dumb to find the correct button. If somebody has a link to a one-time donation for the GIMP project please feel free to post a reply.


I have the same pet peeve. It irks me that Patreon takes pride in not accepting one-off payments.


A $5 monthly donation will in 12 months' time turn into $60, so it's not a great mystery why they want the recurring income.

In addition, it's more stable. If you just get money once in a while, planning is harder. If you have so and so many subscribers, you can make better guesses on what your budget will be.


Sure, I understand that. Everybody and his dog wants you to subscribe and provide a steady income stream.

The problem is that a lot of people, not just me, will violently resist such a scheme even while they'd like to donate.

So you just forgo about every donation coming from somebody who is just hell bent not to enter an additional recurring obligation and I'd wager those are a lot.

So in the end (and unless they change that) you get nothing.

From a fundraising perspective this really doesn't make any sense.


The move to GTK3 and Wayland can’t come soon enough for me. :)




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

Search: