That's a good idea. It's not hard to implement, you need to define two classes, List (for example) and Listview, one handles the link to the programming language, one handles the display on the screen.

For example, group.c defines the Group class, about 200 lines of code, most of it to implement save-as on groups of images:

https://github.com/jcupitt/nip2/blob/master/src/group.c

And Valueview controls the display, about 120 lines of code:

https://github.com/jcupitt/nip2/blob/master/src/value.c

I think the problem is that nip2 is still a gtk2 program. It really needs updating to gtk3 (gtk4 is coming soon, argh) before committing much more work to it.

I've done a nip3 branch here:

https://github.com/jcupitt/nip2/blob/nip3

I've ported everything, it builds, but it doesn't run. All of the image display stuff has changed for gtk3, so it needs reimplementing.

I've done a vips gtk3 image viewer here:

https://github.com/jcupitt/vipsdisp

That builds and runs, and looks quite nice, but it's not really finished, it's missing a few features, and it has an annoying flicker when you zoom.

So the nip2 roadmap is currently:

• finish gtk3 vipsdisp
• copy-paste that into the nip3 branch, replacing the old nip2 image display code
• a few more fixups to get a nip2 that basically works with gtk3

Then there are a load of things that need fixing in nip2:

• move to vips8 API, get rid of the nip2 cache
• rework the look n feel to fit the gtlk3 model

That would be enough to make something that could be released. The nip2 TODO has a long list of more blue-sky ideas.

Anyway, not too much can happen before the move to gtk3, unfortunately. I'm hoping that now vips8 is stabilising I'll have some time to fix up nip2.

I see.

As far as Nip2 goes, its working pretty nice. ie: no real showstoppers known at this side.
If the plan is to update nip2 to gtk3, I would give that more precedents over smoothing out nip2.

As far as gtk4 goes ... although its already in development. The 4.0 version seems to be planned for spring 2019. But before gtk4 moves into its so called stable state, I think you can probably add at least one or two additional years to that.

(skipping gtk3 all together could also be a option. It would kinda trash the work you already did on it. But would free-up the time to complete it, in favor of venturing into gtk4)

you need to define two classes ...

I appreciate the vote of confidence in my presumed coding and compiling skills, but I'm sorry to say that compared to the real deal/coders, I'm not even coming close to what you would call an apprentice. (Complete C noob to boot)

... But I'm of course not the only one that might read this, I hope. :)