Xonotic Forums

During an IRC discussion some time ago, I found out that gtkglext (a GTK extension needed for netRadiant) is rumored to have died and no longer be maintained. Obviously this means that without a major rewrite, netRadiant is going to die too. Since I'm a mapper and netRadiant is my only and favorite editor (I couldn't imagine using anything else) this really worries me.

What's going to happen if that library dies and no one switches netRadiant to something else? Is it the end of mapping in Xonotic and even a danger to Xonotic as a whole? Or will mappers be forced to map in 3D modeling programs directly? I sure hope that never happens... if it does I can probably never map again. I always loved Radiant for being very easy and well optimized, and IMO any different map editor would be worse.

There has to be a way to save it! But divVerent said it would take rewrites so major he doesn't consider doing it. Maybe someone else can get Radiant to use another rendering backend, or find a way to keep it going as is. There are rumors that darkRadiant might try a way and we could switch to that... as long as it looks and works the same as gtkRadiant I don't mind personally. I also noticed that gtkRadiant itself is still in development (I thought it was dead for years), maybe they'll find a way too?

i thought netradiant used gtk+ 2 anyway?

Xonotic could use it's own easy-to-learn mapping tool, but that would be a huge undertaking.

(01-10-2013, 08:50 PM)s1lence Wrote: [ -> ]i thought netradiant used gtk+ 2 anyway?

Xonotic could use it's own easy-to-learn mapping tool, but that would be a huge undertaking.

netRadiant uses GTK. But in order for the viewports (wireframe top / side views and 3D view) to be rendered in a GTK window, gtkglext is currently needed. If that dies then the only way is to make it use some other extension or rendering technique, but what and how?

And Xonotic already has its own easy-to-learn mapping tool, which is netRadiant. Like I said I don't believe anything else could be as good. If it can't be replaced with another gtkRadiant fork (darkRadiant, or gtkRadiant itself) that would be really sad. I don't see myself using anything else since Radiant has been the best tool and anything different will certainly not be as easy, well optimized and functional. So I keep hoping one of the gtkRadiant forks can and will be saved, but have no info on that personally.

Well, Xonotic does maintain NetRadiant too in git? Then maybe put gtkglext in it and make it compile too when running the makefile?

(got no experience about this though)

Then our gtkglext will have to be included in the distribution instead of it being an external dependency. That means shipping gtkglext for all platforms. And if there's a gtk+ update that gtkglext doesn't like for some reason we'll have to include an old gtk+ too. It gets more and more annoying the older it gets, so it's best to stay up to date.

At least that's how I understand it.

I spoke with divVerent again about this since posting here. From what I understand, if gtkglext ever does break, netRadiant will include its own version of it which can be maintained against GTK. That doesn't sound too difficult from my understanding, so if it's ever the case I'm less worried now So far gtkglext still works and might not break years from now, therefore it's not a very actual concern like I thought.

Oh okay, and I'm sure he's not the only one who would be interested in a working gtkglext either. Good.