When you don't want to do something, you look for thousands of excuses to grab for not carrying out the task, and honestly, I think this is the case of the developers of Xfce with the theme of carrying gtk3 la version 4.12 This Desktop Environment.
The thing start hereWhere Nick schermer begins by saying:
Friends,
We have ported some parts of Xfce to Gtk3 and it is better to stick with Gtk2 in version 4.12. Along with the fact that it is a tremendous amount of work, the resulting applications are visibly slower, they consume more resources and there are almost no advantages (API technically). So for me porting the Xfce kernel to Gtk3 in the current state is, for me a: no-go.
That said, you can still prepare for this to make the change easier in the future.
- Depends on Gtk 2.24.
- Use the new API (compilation with GSEAL) as much as possible and deprecated API.
- In a number of places to use Gtk widgets instead of our own versions.
- Libraries (libxfce4ui and exo) can provide a gtk2 and gtk3 version of the lib.
- Some basic modules (xfce4-appfinder for example) you want, you can compile against gtk3. This allows us to track progress and possibly Gtk3
help finding topics.This will give us more time to move the parts to Gtk3, due to the complexity involved, it will be time consuming, leaving not much space to add new features. Not to mention the fact of carrying (and only port) everything needed for fun outside of the project, at least for me.
It might even be possible to run Gtk3 panel plugins on a Gtk2 panel (externally), as well for freebies that will give you one more migration path.
We've already talked about this a bit on IRC / private, and Jannis and Peter agree with this route. Hopefully others do too, because then we can map the portability parts and work on roadmap 4.12.
More or less this has been the translation of the original message in English. In part I share the idea that, if gtk3 It is not yet as optimized as it should be, and this leads to a problem for the performance of the applications, the development of Xfce should follow on gtk2, but the question is for how long?
gtk2 it will gradually lose support and yes, although it may be a long time, in the end they have to carry Xfce a gtk3. But it is not only the Desktop Environment. Most of the applications used in GNU / Linux and are developed on gtk, they have also started to implement gkt3.
Therefore, I think that the most correct thing would be in some way, to give support or compatibility to gtk3 en Xfce 4.12, even more so considering that by the time this new version comes out, it is possible that gtk is even more advanced.
What do you think?
I think sooner or later they will have to accept that they cannot continue in gtk2 forever, and give in at some point, for the sake of users.
It seems to me that at some point the change is going to take place, it is logical and necessary for the development of XFCE, in addition, it must be considered that the success of this environment today is no longer just because it is light (which it is no longer so much), but because it is a real and complete alternative for those who want to continue using the «classic» desktop. In addition, the applications already ported to GTK3 that I use in my XFCE (such as Audacious or Transmission) have not seen their performance affected at all since they were ported and, beyond looking horrible with certain themes, no tangible difference with applications is noticed. GTK2.
I tell you, for me they are nothing more than excuses. The developers of Xfce they are stubborn. Same with Thunar, they use performance and simplicity excuses not to add tabs or an extra panel. You have to accept that, however it may be, as much as I like that Xfce consume little, if you can't, then you can't.
If the truth is that something stubborn is because the fact that they do not put tabs on Thunar and Mousepad has no excuse, the tabs do not have to affect performance and if they are so concerned they can remove the application with an option that activates the tabs and so each one has it as they want.
There have been several times that I read that GTK3 is heavy, but if it is so heavy then what is the advantage? And I share with the author, if you have to do it sooner or later, then do it and that's it.
I think that the insignificance of memory that consumes xfce with gtk3 is not perceptible by a human
i even think i read something about porting lxde to gtk3 so i think all gtk based desktops have to change or be stubborn and outdated
I don't know what Gtk2 or Gtk3 is but don't start messing around a lot now that I've found my perfect distro on Xubuntu.
As "normal" users we won't even notice the change when it's done.
Not in lightness? Ah, then let them play whatever they want 😉
Gtk are the libraries and what the desktop is made of ... originally the libraries are from the GNOME project but are used by several others.
I agree with the xfce developers, that you need to enter gtk3, it is better to let the library become more stable, even if they say that they are not good for it, opt for another and if this takes more work, reduce consumption from memory because you can extend the change to gtk3, including having the two alternatives, that way people decide.
I think that as long as they integrate well visually (that's it), you don't have to do the trick until gtk3 provides some useful advantage for the environment.
and if developers so choose, they must have good reasons
For GTK2 to be what it is, it took more than 6 years, with GTK3, we are barely going 2. It may be true about performance to some extent, but like everything in this world of computing, the principles of Moore's Law affect these things directly.
In a few years, an average PC will have no problem with a GTK3 program weighing a bit more than its GTK2 equivalent (With 8GB of ram, who will care about that?).
I do agree that they make transitory packages and not make the change to dry for the next version ... That is, because simply doing so would affect reputation ... we all know that it does not take but a handful of disgruntled trolls to damage any image ^^ U
Moore's Law has its days numbered (one or two decades :-P).
Moore's Law was created in the 50s and is still in effect
But the laws of physics will end it (says Stephen Hawking and Moore himself).
They will undoubtedly have to give in to gtk3…. but that xfce is no longer the lightweight desktop par excellence is debatable.
Xfce is light, what is not so light is what I call the "xfcedesktops" like Xubuntu (xubuntu-desktop) Do not confuse pure xfce4 with distributions like Xubuntu that use Xfce and add much more.
This possible delay in moving to GTK3 is not really that worrisome either. What happens is that when they do portability it is possible that they lose some of the stability achieved so far. The problems lie in the fact that GTK3 is not quite as mature as the Xfce developers would like (for example, the memory consumption problems referred to by Nick Schermer) and therefore they value an intermediate solution:
Develop 4.12 in GTK2 but "orienting" the design to GTK3 (I don't know how to explain it better), in this way when the library change is made, it is carried out more quickly and efficiently (as fast as possible). go Xfce development). This would avoid a transitional version that had the change made "halfway" with the associated problems.
And the change of version of Xfce to 5 has been mentioned again ... but maintaining the traditional resistance to changes from the developers, it seems that it has remained a proposal.
They should wait or in any case make a testing version of Xfce that works with gtk3 while the stable one continues with gtk2 so whoever wants to can risk installing the test version and whoever wants to can continue with the stable version they use gtk2. What is not logical that as soon as they get involved with gtk3 which will surely lack many things to polish.
You may have to wait a bit, to ensure the good performance of xfce and thus also improve the support of gtk3, while further improving xfce.
Greetings.