In Road To Qt, Audacious Switches From GTK3 Back To GTK2

[dee.]

Hey, it's all good if someone wants to create a polished user experience. But if you start removing options because you want everyone to see all the cool stuff you made for them, if you purposely break customization options for the sake of "brand visibility", that's some highly fucked up shit right there.

I can tell an anecdote: I redesigned a whole new UI & theme for a certain software project, and worked hard (for free) and led/organized the work of several other people to create a functional UI and a consistent, clean and pleasing visual style for the application. But you know, at the same time, I also wrote a lot of code to add options for theming and customization, and whenever I changed something, I tried to make it in a way that is not hardcoded but instead definable in themes, because no matter how great a theme or UI, it's never going to satisfy everyone because people have differing tastes and opinions.

And during all those months spent working on this project, these two goals never ever contradicted each other. A project can have a consistent "brand", and a cool, modern, pleasing visual style and a functional UI, and ALSO offer the option to customize the look and feel for those who'd rather dress up the application the way they want. Go figure!

I have nothing against GNOME really. I'd really like to see GTK succeed, because it's good to have an alternative for Qt... sadly it looks like GTK is turning into a GNOME-only toolkit, which forces everyone to use Qt, and that's going to be kind of shitty for DE's that use GTK like Cinnamon...

[psychoticmeow]

I see three major problems here:

1. It shows gtk+ is happy to break API compatibility if it benefits the goals of Gnome, no matter what effect it may have on the community. So far this has mostly affected theme developers, but the general ideas that gtk+ developers have stated can be applied to anything in gtk+ (as recent issues with window decorations demonstrate).

Indeed, hence the discussion earlier about communication.

2. They apparently are happy to work with Red Hat theme developers, they just don't care about community themes. Why is only Red Hat allowed to break Gnome's precious "uniform presentation"?

What's this now? I need more information.

3. No matter how hard they try, they are not going to have a "uniform presentation". Distros will just patch their own stuff in no matter what gtk+ does. The only people who suffer are community members.

If a particular distro wants to make their community members suffer, how is that the fault of the Gnome project? Why would they patch Gnome to break its UI when alternatives exist?

I don't see any FUD. I see people upset with the direction a project is headed. I see people upset that the promises of API stability were not upheld. FUD implies that the information is false or somehow intentionally misleading. But nothing they said was false or misleading.

What about this "hardcoded" icons nonsense, the thing that I specifically called Luke on twice now, and then claiming that Gnome is a tablet-inspired interface, when it isn't, and never was. Isn't that false information?

You may not agree with them, and gtk+ obviously doesn't agree with them. But just disagreeing with something doesn't make it "FUD" or "horse shit". People are allowed to disagree with the decisions of a project. People are allowed to voice those disagreements.

You think dismissing anyone who dares disagree with the decisions of gtk+ devs as "FUD" and "horse shit" counts as "an actual discussion"?

If you think I've dismissed a valid argument somewhere, please let me know and I'll correct it. That said, before Luke got here there was actual interesting discussion happening, and I'd like to see it continue because if the community doesn't see fit to address these issues, they won't get addressed. Shocking I know.

[TheBlackCat]

While almost everyone here has been willing to have an actual discussion about what is broken, you've come in with the FUD guns blazing.

You think dismissing anyone who dares disagree with the decisions of gtk+ devs as "FUD" and "horse shit" counts as "an actual discussion"?

[TheBlackCat]

All I see in that blog article is a bunch of idiots unable to grasp that the Gnome project wanted to try and work towards having a somewhat uniform presentation and actually getting some spit and polish applied to their user

I see three major problems here:

1. It shows gtk+ is happy to break API compatibility if it benefits the goals of Gnome, no matter what effect it may have on the community. So far this has mostly affected theme developers, but the general ideas that gtk+ developers have stated can be applied to anything in gtk+ (as recent issues with window decorations demonstrate).

2. They apparently are happy to work with Red Hat theme developers, they just don't care about community themes. Why is only Red Hat allowed to break Gnome's precious "uniform presentation"?

3. No matter how hard they try, they are not going to have a "uniform presentation". Distros will just patch their own stuff in no matter what gtk+ does. The only people who suffer are community members.

It's fine if you don't want that in your desktop environment, but at least acknowledge that the goal is worth it, and not pull out the "Microsoft-like" FUD.

What a load of inflammatory horse shit.

I don't see any FUD. I see people upset with the direction a project is headed. I see people upset that the promises of API stability were not upheld. FUD implies that the information is false or somehow intentionally misleading. But nothing they said was false or misleading.

You may not agree with them, and gtk+ obviously doesn't agree with them. But just disagreeing with something doesn't make it "FUD" or "horse shit". People are allowed to disagree with the decisions of a project. People are allowed to voice those disagreements.

[psychoticmeow]

OOOhh that's a lot of personal attacks!

You're damn right.

Still, the hardcoded icons are in the filechooser's sidebar and I have yet to find any options to revert them. Since they are images they could surely be replaced in the source and the resulting modified source compiled. It would be beyond my skill to write support for changing them at runtime in GTK, so if they ever make the rest of the icons hardcoded I will be forced to replace and recompile.

GNOME can publish anything they want, but I do not have to use it, nor do the folks at Audacious, who kicked off this thread by quite sensibly dumping GTK3 for GTK2. So can Mint, which created Cinnamon to be everything a lot of desktop users feel GNOME 3 should have been. All GNOME would really have to do to remove the need for so many forks would be to stop breaking the extensions with every new release, and not hardcode things like those sidebar icons in the filechooser. At least this is FOSS and not Microsoft, so those of us who like to play with things really DO get a choice.

How are these any more hardcoded than other icons? If your theme doesn't provide the necessary icon file then it'll fall back to using those from Adwaita/the default. And what's this now about extensions breaking with every release? Gnome shell extensions, or something Gtk related?

[Luke]

gtk-file-chooser is where icons are hardcoded

What's this about hardcoded themes? Adwaita isn't hard coded. And why should you, or anybody not part of their project have say in how their project runs? If you want a say, then you need to be a contributor. But you've already said it is incompatible with what you want, then you should probably contribute to something else instead of spending your time dick waving about how 'awful' it is.

OOOhh that's a lot of personal attacks! Still, the hardcoded icons are in the filechooser's sidebar and I have yet to find any options to revert them. Since they are images they could surely be replaced in the source and the resulting modified source compiled. It would be beyond my skill to write support for changing them at runtime in GTK, so if they ever make the rest of the icons hardcoded I will be forced to replace and recompile.

GNOME can publish anything they want, but I do not have to use it, nor do the folks at Audacious, who kicked off this thread by quite sensibly dumping GTK3 for GTK2. So can Mint, which created Cinnamon to be everything a lot of desktop users feel GNOME 3 should have been. All GNOME would really have to do to remove the need for so many forks would be to stop breaking the extensions with every new release, and not hardcode things like those sidebar icons in the filechooser. At least this is FOSS and not Microsoft, so those of us who like to play with things really DO get a choice.

[psychoticmeow]

That "uniform presentation" is incompatable with what I want on my desktops, I've had to put a lot of work into reverting unwanted UI changes while using current libraries, kernels, and drivers. The only excuse I will accept for that filechooser is if it uses CONSIDERABLY less resources. At least as it is open source in a worst-case scenario I could build it myself with the icons replaced, I suspect we will be seeing such builds down the line, with various distros building their own versions of GTK3 to replace hardcoded GNOME themes with their own hardcoded themes. Ahh yes: fork city made necessary by hardcoding themes both distros and users want to change!

What's this about hardcoded themes? Adwaita isn't hard coded. And why should you, or anybody not part of their project have say in how their project runs? If you want a say, then you need to be a contributor. But you've already said it is incompatible with what you want, then you should probably contribute to something else instead of spending your time dick waving about how 'awful' it is.

GNOME should think long and hard about what they are doing. They won't drive anyone to Windoze, that is true, but it is simply amazing how distros like UbuntuStudio have had to ditch the tablet-inspired interfaces for DE's like XFCE that are actually compatable with what they are doing.

Blah blah, tablet-inspired, blah blah, not good for real work. Yawn. I'll keep using this desktop environment to do actual fucking work just fine.

The huge popularity of Cinnamon, MATE, XFCE, LXDE, and even GNOME having to release a classic interface all show how many desktop and laptop users don't what a tablet style UI. On top of that, tablet users have also declared that the new UI's don't actually work that well for a touch interface either.

So Gnome is a bad desktop because alternatives exist. Got it, brilliant argument. And Gnome, a desktop interface, doesn't live up to life as a tablet interface. Wow, I wonder if Gnome might not be a freaking tablet interface you moron.

There is one point here involving Microsoft: I can't believe how dumb MS must have been to not see the hornet's nest stirred up by the GNOME and Unity controversies, and bring out their own tablet-inspired interface in Windows 8. Unlike GNOME, MS has lost buckets and buckets of money because of "Metro" or "Modern" to the point of also costing computer makers big bucks as people decide to forget buying anything that comes with Windows 8. GNOME need not worry about pushing anyone to Windows 8!

While almost everyone here has been willing to have an actual discussion about what is broken, you've come in with the FUD guns blazing. Maybe next time you post, you could put some thought into it instead of recycling this trash?

And you still have not answered my earlier question about how a failing on your custom theme to support symbolic icons in a manner that suits you became the failure of Gnome to do... what exactly?

[kigurai]

all show how many desktop and laptop users don't what a tablet style UI. On top of that, tablet users have also declared that the new UI's don't actually work that well for a touch interface either.

As far as I know, NO ONE in the Gnome camp has claimed that G3 is a tablet interface, or was designed for that. I believe you can find quotes about "playing better with touch screens" or something, but for anyone who actually use Gnome 3 it is very obvious that the design is made for keyboard and mouse.

[Luke]

Lots of us don't WANT to be forced to use someone else's theme

All I see in that blog article is a bunch of idiots unable to grasp that the Gnome project wanted to try and work towards having a somewhat uniform presentation and actually getting some spit and polish applied to their user interfaces.

It's fine if you don't want that in your desktop environment, but at least acknowledge that the goal is worth it, and not pull out the "Microsoft-like" FUD

That "uniform presentation" is incompatable with what I want on my desktops, I've had to put a lot of work into reverting unwanted UI changes while using current libraries, kernels, and drivers. The only excuse I will accept for that filechooser is if it uses CONSIDERABLY less resources. At least as it is open source in a worst-case scenario I could build it myself with the icons replaced, I suspect we will be seeing such builds down the line, with various distros building their own versions of GTK3 to replace hardcoded GNOME themes with their own hardcoded themes. Ahh yes: fork city made necessary by hardcoding themes both distros and users want to change!

Proper theming support should be considered a prerequiste to the release of a finished version of something like GTK for inclusion in released versions of distros. Hardcoded themes are OK in alpha versions, just as I might hardcode things in early versions of a program but make them configurable for the finished version. I regard the program as unfinished as long as anything someone else might need to change is hardcoded-even in a shell script.

GNOME's advice NOT to use GTK3 in applications seems to be good at this point, I just hope the MATE team is sucessful in porting GTK2 to Wayland. That will then become the go-to version of GTK, and with luck will be the one GTK4 evolves from.

GNOME should think long and hard about what they are doing. They won't drive anyone to Windoze, that is true, but it is simply amazing how distros like UbuntuStudio have had to ditch the tablet-inspired interfaces for DE's like XFCE that are actually compatable with what they are doing. The huge popularity of Cinnamon, MATE, XFCE, LXDE, and even GNOME having to release a classic interface all show how many desktop and laptop users don't what a tablet style UI. On top of that, tablet users have also declared that the new UI's don't actually work that well for a touch interface either.

There is one point here involving Microsoft: I can't believe how dumb MS must have been to not see the hornet's nest stirred up by the GNOME and Unity controversies, and bring out their own tablet-inspired interface in Windows 8. Unlike GNOME, MS has lost buckets and buckets of money because of "Metro" or "Modern" to the point of also costing computer makers big bucks as people decide to forget buying anything that comes with Windows 8. GNOME need not worry about pushing anyone to Windows 8!

[ssokolow]

What is the point of another qt based music player? At least Clementine and Amarok should cover all the bases (the former for traditional music playing with all the network features, the latter for music browsing). It is kind of like how I'm not a fan of OpenShot porting to Qt when Kdenlive exists. It is completely redundant.

Last I checked, neither of those players could understand my playlist containing AAC, HSC, LDS, MIDI, SPC, GYM, NSF, GBS, MOD, IT, XM, S3M, STM, MP3, PSF1, PSF2, WavPack, FLAC, Ogg Vorbis, SID, Wave, VOC, AU/SND, and WMA files. (Heck, Audacious has got my back if I ever decide to add in some 2SF or VTX files too.)

I'm still mad at the UADE devs for threatening to break the UADE API if the Audacious UADE plugin was kept in-tree, then letting it bitrot as the plugin API changed. Because of that bratty decision, I can only play Gobliiins32.dum without converting it by using the uade123 command-line tool.