Hacker News Reader

Comment by Kye | Hacker News Reader

So what's actually holding back development? GEGL is 22 years in and still isn't fully implemented. I've heard different explanations. The main one being that the people working on GIMP are (allegedly) hard to work with, so it's hard to get people with the needed skills to stick around.

View on HN

GIMP is a huge code base primarily in C and GObject. This is not a popular combination.

The UI toolkit is GTK, it's not exemplarily maintained for operating systems other than Linux.

A lot of potential new stuff used to be held back by the lack of a complete GEGL port. I think the public perception is that it is still the case (it is not).

Long development cycles used to demotivate contributors. You could write a new cool feature, and most users wouldn't see it until 3-4 years later. This has been resolved as well. Since 2.10, stable releases get new features backported from the unstable branch.

Organizationally, there is no structure like a foundation in place to help financing development. It's being talked about in the project for the past few years, no action has been taken yet.

Are you considering adding Rust to the mix?

There was some talk about making it possible to write plugins with Rust. Beyond that — no, I don't think so. I really hope not. If you start rewriting GIMP in Rust, you can pretty much kiss non-destructive editing goodbye. There will simply be no time for anything else.

I didn’t mean rewriting, not initially at least, but rather building new parts in Rust, as that would remove large parts of what is currently done manually or with macro magic (inside GLib or elsewhere).

But I can see how not having enough people can make even that very difficult.

Reply on news.ycombinator.com