Red Hat Enterprise Linux 10 Is Eliminating GTK 2 Support

[Spike29]

Gtk+2.0 is simpler (and standard C99); Especially once combined into a monolithic library (currently it is a modular monolith that is common in the Linux ecosystem). Admittedly it wont ever be as light as FLTK (limiting to C++03 and a subset of STL was a good idea for portability), but I do predict that maintaining it won't be so bad.

I have already done half of the combining whilst maintaining a Gtk 2.0 Quake III level editor.

Out of curiosity, what Quake III level editor are you referring to?

I use NetRadiant to edit Xonotic maps, it's one the last GTK+2 applications installed on my system.

[kpedersen]

Out of curiosity, what Quake III level editor are you referring to?

I use NetRadiant to edit Xonotic maps, it's one the last GTK+2 applications installed on my system.

Ah nice. Admittedly NetRadiant does have more features than mine (OpenRadiant). I had to strip mine down a little because I plan to do things like curves and other engine specific features a little different in future.

https://www.thamessoftware.co.uk/openradiant.html

The advantage of my fork is that it works like a generic modeler like Blender. I.e exports directly to Wavefront .obj for any game engine (without needing to go to bsp and then converted), and grabs textures from the local directories rather than a Q3 base dir or PK3 files. Mine also does lightmap baking using some middleware I wrote a while back that can inject it into .OBJ wavefront (which doesn't traditionally support lightmaps).

Additionally, I am ~70% the way through a port of it to FLTK. Actually, I originally wrote FL_Flex to help me with the port which I am quite excited has been integrated upstream by the FLTK developers

https://github.com/fltk/fltk/blob/master/FL/Fl_Flex.H

But alas, day to day work (admittedly working on quite a similar tool), gets in the way of finishing this one off!

Ironically my map making skills are actually fairly basic. I just kind of find it relaxing. Do you have any screenshots of yours?

[birdie]

gpicview and lxdm are gtk3

Not on Fedora 36 for some reasons. No idea why.

[Anvil]

Any volunteers to port these apps to GTK3/4?

ddccontrol-gtk - use it daily
gkrellm - likewire
gpicview - use quite a lot
hexchat - use quite a lot
lxdm - my DM
pavumeter - use it often
unique - it has UI? Really? OMG

A weird decision on Redhat's part.

And these are GTK2 applications that I use. I imagine people have lots more.

you wouldnt bother porting Hexchat as its a Mess with rather old ancient stuff in it, easier just to rewrite a New Client

[kpedersen]

you wouldnt bother porting Hexchat as its a Mess with rather old ancient stuff in it, easier just to rewrite a New Client

This constant rewriting of software is not sustainable. How can a rewrite even every half a decade (forked from XChat right?) ever hope to catch up with classics such as mIRC on Windows?

[hamishmb]

I agree, people aren't being fair to birdie here.

[Anvil]

This constant rewriting of software is not sustainable. How can a rewrite even every half a decade (forked from XChat right?) ever hope to catch up with classics such as mIRC on Windows?

if one want's an IRC client like Mirc they can use KVIRC which is old again an doesnt look like it has any active development on it, xChat/Hexchat is old outdated , it needs a whole rewrite , not just the client itself, other things that the client uses. it'd take longer to port the current hexchat over to a New Toolkit than it would just to rewrite a whole new client that was based on gtk4

[fitzie]

Where the fuck did I complain? I invited people to port these applications and didn't welcome this decision. What's up with people today?

Also, you know what's wrong with Linux exactly and why Windows, Chrome and Mac OS are everywhere and Linux is only seen on supercomputers? These OS give a fuck about backward compability. Linux/Open Source developers are obsessed with developing something new and throwing away perfectly working solutions. This is not limited to the core Linux OS, NPM/Ruby/Python libraries have the same issue.

You do not break people's workflow. You do not make them look for replacements, oftentimes either missing features, working differently or not existing at all. Must be something very wrong with you if you think otherwise.

I appreciated your list. I will take a stab at porting gkrellm as that is the only app on the list I heavily use. I occasionally use hexchat, but as someone mentioned, the issues are massive and it is unlikely to get ported any time soon. That's sad, because it really is a friendly irc client.

[Anvil]

read this is to what would need to happen to port Hexchat etc https://github.com/hexchat/hexchat/issues/2047 ( TingPing ) has maintained Hexchat since he took the code from xchat i think

[DMJC]

Having to keep changing apps/workflow really is making Linux annoying to use. Mate exists for a reason. Having to switch app names from Gedit to Pluma and all the other mate apps was annoying as hell. Having to change my IRC client AGAIN is going to be pain. It was bad enough just finding out that Freenode.net had pushed everyone to other servers with their stupid account policies. Every time these big changes happen the communities fragment and we lose some people. Not to mention how much sound has been broken over the years by badly done distro changes. Bluetooth was especially painful in the pipewire migration.