Default applications review for 22.10 / 23.04 and beyond

For Ubuntu Budgie 22.10 the project team have embarked on a fundamental review of what applications we ship by default.

This is primarily needed due to GNOME core apps shipping libadwaita and the inability to have such apps looking the same across the whole distro in-accordance to a users wish. If you want to see the state of play now - just install Debian Testing into a VM and install the same applications as Ubuntu Budgie ships. Look to see how they look and work now.

We as a distro want to give that choice to users - workarounds such as “copying gtk4.0 into a users .config folder, or using a GTK_THEME=themename executable name” fundamentally break those GNOME application authors wishes - i.e. they have designed their apps look and work with the default GNOME Shell adwaita theme.

Shipping any application is a positive endorsement into the application author(s) creativity and zeal. We should be all proud of software creators and their vision.

We would like to have a consistent out-of-the box look and feel - but we are very cognisant that some GNOME Apps do not have alternatives. So this is going to be a multi release strategy all leading to the next LTS in 2024.

The review will be a game of two half’s … first look at apps in the repo.

Second where we cast our net over the wider non debian based ecosystem.

So how are we going to approach this?

  1. The list of packages we ship are called seeds. The jammy list is here .
  2. We are going to look at the upstream software authors source repository current issues and current code-base to see what their current and future intentions are (assuming this info is available)
  3. If the app is/will be libadwaita based this app will be a candidate for re-evaluation.
  4. We also will look at the list of current issues, last release date and how active the author/team are in responding to issues.

This is critically important. If the app is not maintained / or maintained only periodically then that app is part of the evaluation to be potentially replaced.

  1. We should consider how well any application is maintained in Debian / Ubuntu. Primarily, packages for evaluation need to exist in the repositories. By looking at the debian changelog for the package and seeing how the package maintainer deals with issues (timeliness / critical & stability patches) will guide us whether such a package should be part of the evaluation.

  2. If a potential candidate replacement application is not part of Debian we need to ask the fundamental question - why not? Does the application simply not work in a system wide install basis? Remember that’s where packages install stuff - for multiple users not just for a single user.

  3. Likewise, taking on the production of the package, looking for sponsors, working through package production issues and upkeep in Debian & Ubuntu is a long-term commitment. Who is going to-do the work? Volunteers who put their hands up are more than welcome.

  4. We also will look critically at the dependencies of the application. There is little point to include an application if it required hundreds of megabytes of dependent libraries i.e. No KDE based applications. Qt applications possibly but theming is key here.

We will maintain this current post with a list of apps that are going to be replaced … and with what.
Whilst we appreciate your insight here with posts - if you want to make a suggestion then please state clearly how all of the above is met with your suggestion. We will ignore anything that just says “why dont you choose X” with little or no reasoning.

Use the following PPA to install and test the following:

sudo add-apt-repository ppa:ubuntubudgie-dev/switch
sudo apt upgrade
sudo apt autoremove

GNOME-Calculator → Mate Calc
GNOME-Calendar → drop from the install
GNOME-Maps → drop from the install
GNOME System Monitor → Mate System Monitor
Evince → Atril
File Roller aka GNOME Archive Manager → leave as is and monitor
Deja-Dup backup tool → leave as is and monitor
GNOME Screenshot → drop from the install
GEdit text editor → undecided
GNOME Font Viewer → font-manager
Cheese → guvcview or webcamoid
Celluloid → Parole
gThumb → leave as is and monitor
GNOME Log → leave as is and monitor
GNOME Disks → leave as is and monitor
Rhythmbox → Lollypop + Goodvibes + gpodder - open question
GNOME Characters → leave as is
Transmission → leave as is and monitor


GNOME Calculator vs Mate-Calc


Mate-Calc was forked from GNOME Calculator in the early days of GNOME 3. It is maintained by the Mate Desktop team - and its a core-app for them. The issue tracker is well managed and the Debian Mate Team quickly packages their stuff.

It doesn’t install that many mate dependencies - basically translation stuff.

Mate Calc does NOT play nice with Pocillo - however with a bit of tweaking (in git master of the pocillo-gtk-theme repo) the calculator looks stylish in light styling and dark styling. Plays nice with Qogir and Arc theme. Tried WhiteSur as well and looks ok. Materia does not work that nicely though.

1 Like

Just my 2 cents here…

Don’t other distributions do the same already ? Mint and Mate mostly ? Maybe an opportunity to create a common effort ?

1 Like

GNOME-Calendar → drop this as a default install.

For budgie-desktop the clock drop-down will no longer have a link to open the default calendar.

This really removes the need for UB to ship a calendar. Since there doesn’t seem to be suitable replacements that I’m aware of my suggestion here is simply to drop GNOME Calendar from the default install.

1 Like

GNOME-Maps is currently a libhandy app.

There is an issue raised to convert to Gtk4 and libadwaita.

As such i dont see any further reason to ship by default since you would assume it migration will occur by the time of the next LTS.

There are no replacement apps. Someone could investigate providing a web app for google maps. Raise a hand if you can do this.

Lets just drop this as a default install.

webapp-manager from Mint might help doing this

1 Like

For Calendar:

For bluetooth: GitHub - blueman-project/blueman

As for Maps:

My suggestion is:

To check more map viewers there is the following page

Gnome System Monitor replacement


For Cheese:

Is GNOME-Maps primarily used for the Linux phones, because it would make more sense to not have it on the desktop version of Ubuntu Budgie, because most people unless they have laptops with mobile data wouldn’t be really utilizing the maps application.

The image is libadwaita on Endevour OS Budgie . What are termed legacy apps do have a custom theme applied, but not system settings or calculator. You can also see the latest default appearance setting for upcoming Gnome versions.

mate-system-monitor was forked from GNOME System Monitor in the early days of GNOME 3. It is maintained by the Mate Desktop team. The issue tracker is well managed and the Debian Mate Team quickly packages their stuff.

It doesn’t install that many mate dependencies - basically translation stuff.

Mate System Monitor does NOT play nice with Pocillo - however with a bit of tweaking (in git master of the pocillo-gtk-theme repo) the treeview headerbar is fixed in light styling and dark styling. Plays nice with Qogir and Arc theme. Materia does not work that nicely though.


My thoughts… We live/work in a web-centric world, do we really need a desktop maps and (stand-alone) calendar app? I’ve never opened the maps app, I head straight for the browser for Google Maps or Open Street Map and have never seen anyone do otherwise. The same goes for a calendar - people using desktop calendars are likely using them from applications like Thunderbird or Evolution, otherwise they’re accessing their (very likely) Google or Outlook calendar from the browser.

I use the calculator a fair bit, as long as it works with common themes (usually Pocillo or Qogir for me), I’m good with whatever.

For system monitor, just include Htop :laughing: (joking, sort of).

1 Like

Document Viewer - i.e. PDF reader is currently managed by Evince “Document Viewer”.

In progress upstream is a GTK4 port and it will be a libadwaita based app. Unknown when it will land - but currently has GNOME 43 milestone associated with it - so Ubuntu 22.10.

Thus looking at alternatives in the archive - most PDF readers seem to be Qt based - okular which is well maintained is a KDE based app - so not suitable.

Thus atril - which is managed by the Mate Desktop team is perhaps the best alternative (?) - disappointingly there is a huge backlog of issues and progress is relatively slow unfortunately.

So open to ideas here - I’ll mark atril as a replacement for evince for the moment

Sorry to repeat but…
…wouldn’t be an easier path ?

thats irrelevant - those are not in the archive.

That’s not irrelevant : an effort already exists to provide predictable app’s in GTK, why not use / share it ?

…wondering if it’s also Mate’s approach in that matter ?

I made it very clear my view on this - this is extremely important - and without volunteers any “what about this” is very irrelevant.

An effort already exists on that matter to avoid wasting your time…

Budgie’s not alone with this « problematic » I guess Mate and Xfce are concerned too. And Mint.

file-roller i.e. the GNOME Archive Manager is a Gtk3 libhandy app.

It is autoinstalled for Nemo - Nemo has a recommendation for nemo-file-roller which is its file-roller integration i.e. on a right click you can compress a file etc.

Looking upstream there doesn’t seem to be any current move to Gtk4 / libadwaita.

The Mate equivalent is called engrampa and basically does the same thing.

Given that there doesn’t seem to be a roadmap for file-roller and due to the nemo dependencies think we should just leave this be for the moment and just monitor future stuff.

If anyone is interested - I’ve created a nemo-engrampa extension if they which to use engrampa rather than file-roller - Add engrampa support - closes #356 by fossfreedom · Pull Request #451 · linuxmint/nemo-extensions · GitHub

Deja-Dup is our backup tool - it is well maintained by GNOME. The next version is libadwaita though.

There seems to be a real lack of choices in the repo - need something that covers both adhoc and scheduled backups - to both local and cloud providers.

… and TimeShift is not a backup tool for local stuff.

This is one area we need to carefully consider - for the moment this is one libadwaita app that we may carry forward.

I’d forgotten this tool until I started looking around . It may be way off the mark though.