Now that Qt is LGPL, Should it become the standard toolkit?

[bufsabre666]

i am surprised how close this is 17 for, 16 against ATM. i was expecting it to be like 2 to 45

[SomeGuyDude]

I honestly don't see the argument. KDE won't explode if you run a GTK+ app on it or vice versa.

Leave the fanboy stuff with little kids with game consoles.

Unity. Drives me bonkers when I run something that doesn't use my GTK settings.

[cb951303]

I don't think 'looks' is an argument here. QT is 100% capable of simulating GTK (QGTKStyle) thus it is at least as customizable as GTK.

[Slug71]

i feel the same, they both serve a purpose. but if they wanted to make qt standard i wouldnt be against it for two reasons:

1) it seems to be in more active development. (either that or gtk just doesnt like to announce every little update)

2) its always nice to standardize behind something, there are several areas i wish linux would just pick one thing and standardize behind it so that the standard one would get the funnel of development and become great (coughfilesystemscough)

Have to agree with number 2) here. Some standardization so that certain things can really excell would be nice. I think we'll start seeing more of this though. Specially when Jaunty gets Packagekit which is already used by Fedora, Super Ubuntu and Foresight Linux and is now Supported by Mandriva and OpenSuse. Hopefully the same will happen when 9.10 gets Plymouth and Login Experience.

As for the Filesystems though, hopefully we'll see that get cleaned up when Btrfs is out.

[geoken]

Ok, so perhaps we should get rid of Gentoo, Fedora, Opensuse, Arch, Linux Mint, and all the other distros, and just make one super distro. Then take away all desktop environments and just make one super desktop environment. Same with Office, same with Web Browser, same with... oh wait, how are we going to keep people from forking? We can't! As long as people use GTK+, it won't die, as someone will develop it and you can't stop that.

We're talking about standardization, not killing.

You guys probably don't notice how lack of standardization hurts you because you obviously have no way of knowing when a developer considers Linux, then scratches that idea because he sees the landscape of api choices like a minefield.

For example, do you guys know that the creator of Braid wanted to make a Linux port. Here is one of the comments he made on his blog post about the matter;

"It’s hard to get good communication with Linux users though, because there are a lot of options and no reliable information about them anywhere. (For example you change your mind about which API provides minimal latency in these last two comments. So as a third party who has no experience with either API, am I supposed to believe one conclusion or the other now? Now multiply that by N thousand pundits and M APIs that play audio. It is a total mess.)"

As of now it seems like he won't be going through with the project because he feels like the task of choosing suitable API's is greater than the actual porting of the game.

[bufsabre666]

Have to agree with number 2) here. Some standardization so that certain things can really excell would be nice. I think we'll start seeing more of this though. Specially when Jaunty gets Packagekit which is already used by Fedora, Super Ubuntu and Foresight Linux and is now Supported by Mandriva and OpenSuse. Hopefully the same will happen when 9.10 gets Plymouth and Login Experience.

As for the Filesystems though, hopefully we'll see that get cleaned up when Btrfs is out.

thats pretty much what im aiming for, is univeral package installation, universal file system and universal sound system.

those are the things that definitely need to be combined

qt becoming standard would be nice but its not something that is a must, but if gnome was to do that now would be the time so theres time to transition before gnome3 even has a release schedule

[phrostbyte]

I think the people who keep making posts like this should go ahead and lead this development effort. Because I don't think they realize that going to Qt is basically like throwing away a decade year of progress on Gnome. It's not like there is some switch built into Gnome and a quick recompile does the trick here.

[cb951303]

I think the people who keep making posts like this should go ahead and lead this development effort. Because I don't think they realize that going to Qt is basically like throwing away a decade year of progress on Gnome. It's not like there is some switch built into Gnome and a quick recompile does the trick here.

And I think you should read about GNOME 3 and GTK 3
Gnome team already thinks about breaking backward compatibility which means a huge rewrite of lot of things. The question is, while at it, why not use QT4?

+1 for Geoken, very true statement.

[mangar]

I'll drop in another two cents:
1. In software development, Choice is good as long as the parts are interchangeable.

I'll explain:

If choosing a toolkit was purely a development choice, with no visible effect on the end user, than jolly good and all is well, and this whole argument is unnecessary.

Unfortunately, this is not the case. There are many incompatibilities between Qt and Gtk application interaction on almost any level (I'm mixing in Gnome and Kde, since those are the FOSS main application platforms): Incompatible HIGs, incompatible look and feel, incompatible soundsystems, incompatible OLE, drag and drop behavior, etc.

Using Qt and Gtk applications together does not work well.

Some may claim that freedesktop.org is making progress toward unified user experience. Unfortunately, the current rate of progress is extremely slow, without many results (with some highly notable exceptions, such as dbus), and generally speaking will probably never be enough to allow for seamless user interaction with the system.

2. Gtk and Gnome are extremely short on man/woman-power - for an illustration, Nautilus is currently maintained by a single person - and are currently using those limited resources to refactor the existing libraries and software. Since a large effort and and extensive rewrite are already planed and taking place, why not reduce the overhead of maintaining Gnome and it's applications by using a common, well maintained development platform? It can allow Gnome to focus on its strength, rather than try to rebuild the entire stack.

[cardinals_fan]

Currently, the duality of Gtk/ Qt is a real barrier to entry for Developing for Linux, since it forces the potential developer to peek a side in the "Desktop Environment wars", and thus fracturing the application base of Linux;

Since Qt is to be released as LGPL, ending (imho) the historical significance of Gtk, I believe that it is for the best interest of all the involved parties - Gnome, KDE, Qt Gtk, and the application developers, to transition away from Gtk to Qt, creating a strong foundation of standards from whom the FOSS community can draw.

In short - retire Gtk, standardize on Qt.

Write your next program in Qt then. If it is genuinely the better choice, developers will choose it and GTK will die out. Think of it as survival of the fittest.