The screenshot shows a mix of styles. Some of it looks very old (eg the Finder analogue). Scrollbars are flat, traffic light window buttons are gel. Semi-puzzling so far in terms of UI design target.
I found this: 10.2-era, pre-Brushed Metal, seems to be the target: https://github.com/orgs/gershwin-desktop/discussions/1
Thankyou for the comment and polite interest, btw -- IMO an excellent way to encourage involvement in an open source project!
This is pretty exciting!
If this desktop takes off, maybe we’ll finally see an ecosystem of applications that use GNUstep instead of GTK or Qt. In my opinion, the traditional charm of the Mac isn’t just the desktop; it’s the entire ecosystem of applications that conform to the Macintosh Human Interface Guidelines. It would be cool to finally see this happen with a GNUstep/Gershwin ecosystem!
Sadly this barely exists anymore.
Cross-platform Electron apps have replaced native AppKit. Cloud-based apps like Linear, Slack and Figma cater to the lowest common denominator of desktops by shipping their web client in a wrapper.
The last real native Mac app that was truly successful was probably Sketch ten years ago, and Figma ate their lunch.
Meanwhile Apple themselves have given up on the HIG. In the Alan Dye era, it’s been form over function across all the Apple operating systems. Their own apps don’t follow any guidelines and the latest macOS 26 is a UI disaster – probably the most inconsistent Mac release since OS X early betas.
I hate how mainstream desktop computing has gone to crap in recent years. Thank goodness for free, open-source software.
The idea is to bring the UX of OSX Snow Leopard back, adjusted for today’s possibilities (better developer experience, AI, etc.). I’m developing a DE, SwiftUI/AppKit-equivalent, and a bunch of reference apps I‘m personally missing in terms of quality (e.g. Raycast/Spotlight, Mail).
You would want to adjust it for today's display and input technologies. A high resolution OLED display deserves a different UI design than a 6-bit low-contrast TN LCD display did.
Those principles survived CRTs, TN panels, Retina, touch, trackpads. They’re not tied to a specific technology.
Can you give me an example of a change in todays UI that was motivated by change in display quality?
There are a lot of places where I now see a miniature thumbnail preview of a file's contents, where in the 1990s you would only have seen an icon corresponding to the file type. Those previews are enabled partly by faster IO and processors making the preview rendering cheap, but also by higher resolution displays making the previews a lot more useful than they could have been at 32 pixels or smaller.
While it's not exactly a quality change as the driving force, the proliferation of dark mode UIs is a result of OLED displays that draw meaningfully less power with darker content, so pushing users toward darker UIs helps battery life. And it looks much better on a display with decent black levels than it would on a crappy LCD that washes out all the dark colors.
Going to launch a few apps powered by AppLib in the next few weeks and then continue with the DE.
Really, anything that can adhere to the old-school HIG well enough to offer A) consistent keyboard shortcuts across the apps that B) use a dedicated, thumb-actuated command key, I'm sold. (The `control` key, and thereby your pinkie finger, should both only be used for sending terminal escape sequences, as god intended [1].)
Might be just me, but that feels like an odd description. Assuming I'm understanding the intent correctly, perhaps something like
> Gershwin is a desktop environment based on GNUstep but made more intuitive for users coming from other desktops
might be easier to understand?
Notably the Hello System, by Simon Peter, AKA probonopd.
I looked at that, too.
https://www.theregister.com/2023/01/31/hellosystem_08/
Probonopd is now involved with Gershwin and was quite active in its Github when I looked.
When I started FuryBSD which was a livecd creator for FreeBSD that made it easy for others to spin up projects Probono noticed and started reaching out, helping me making some great contributions. It became the basis for the current GhostBSD LiveCD, HelloSystem, I believe RavynOS, and FyneDesk used it or at least were also using it in the past.
Probono blew me away with his work on LiveSTEP, and it just kind of stuck with me. I ended up silently carrying it forward, and getting back in touch after I realized what we could do with it. I recently gave Probono ownership for the GitHub org, and full creative control so I could focus on harder functional parts like a truly integrated WindowManager. It's all just been somewhat a miracle, and a matter of timing lining up I suppose. I am very much looking forward to seeing the cool things we can do together in 2026!
:-)
> which eventually brought me to making FuryBSD for a short time.
Ahaaaa. I did not realise that. Perhaps you should mention that in a FAQ or something? I think tying the different projects together like that would make it clear there is a quite considerable bit of history in here.
> I am very much looking forward to seeing the cool things we can do together in 2026!
Kudos for the positivity.
I left the GNUstep community a year or so back, after the admins got angry with me for daring to have opinions about the project that differ from theirs.
I think that as well as (1) a set of development libraries, it's also (2) a quite impressive set of apps, (3) an app packaging format, and perhaps most importantly (4) a quite complete desktop environment. They only seem to care about #1 and regard points 2-4 as annoying distractions.
For what it's worth, I know of two other active, current GNUstep-based desktop environments, which have slightly different focuses.
1. Ondrej Florian's GSDE: https://onflapp.github.io/gs-desktop/index.html
I've written a bit about this:
https://www.theregister.com/2023/07/06/two_new_debian_deskto...
I had great difficulty getting it to build on Debian 13, but I should try again at some point.
2. Sergii Stoian's NEXTSPACE: https://github.com/trunkmaster/nextspace
Slightly more mature but Stoian's been distracted recently by Russia invading his country, so it's not seen much work of late.
We do largely operate outside of GNUstep. Now we approach it like let us be the desktop, let them be the core libs. My take is I do think GNUstep should be marketed as more of a cross platform solution to build applications than anything else. You are more than welcome to come discuss ideas with us at Gershwin anytime.
GSDE (Screenshot.app), more so NextSpace a lot of things do not work with a lot of modifications on FreeBSD for example and I found the build systems unexpectedly difficult. I am a fan of the efforts otherwise and will try to make Gershwin components like WindowManager.app something they could use if they want to make use of in the future. I think each project has a place, and a role in promoting GNUstep. I wish they each had Live ISO's with installers. There is also agnostep now that looks promising by the way. https://github.com/pcardona34/agnostep
It's a good distro and I like it. It's easy, it has good accessibility, and as you say, it looks great. But I tried daily-driving it for a short time and found it too limiting for me. Once I'd manually installed Firefox, Thunderbird, Chrome, VLC, LibreOffice, Ferdium, Panwriter, VirtualBox and my other everyday tools, I wasn't using a lot of the distro's own tools any more, and suddenly the ones I was (settings, file manager, app launcher) became limiting.
I recently began porting Ladybird from QT to GNUstep and the results are a 20 minute build on low powered raspberry pi vs 3 hours using most of the same APIs from AppKit. I believe I can take the same code, and make it build on Windows. I suppose this is my elevator pitch as to why.
Mac users looking to switch want the whole package: the aesthetics, the functionality, the design philosophy, and the holistic approach.
