I love Patrick how you have a law that explains absolutely anything ever discussed.
As long as there’s no rush to decide, I myself have more testing to do but I’d be fine to present such a poll with comparative screenshots if I finalized a working smooth solution for a custom theme after I ruled out any bugs with it.
I still have the Arc bugs I described earlier, but that’s in my Whonix 14, not 15 (need to test that), for example, and I will get to it some time soon-ish.
@AnonymousUser that shows you how decisions are made here no polls. And really that’s the realistic way to do things.
If developers are happy with security level and no increased risks, I am content with every decision. I really hated KDE but endured with it because that was the only option at the time. Getting XFCE was like a breath of fresh air again. I am worried that piling more and more flashy features may be bringing us closer to bloated feel again.
Last remark on the subject: during the last year users moved from Whonix 13 to Whonix 14 to Whonix 14 XCFE. Debian buster stable is expected in a few months if I understand correctly. It’s really good and encouraging to see the progress in development here but please also consider that too frequent changes can be confusing. I will be happy with minimal GUI changes from Whonix 14 XFCE to Whonix 15.
This is the current XFCE default theme. It’s not just ugliness. Look at this window. Is ‘Application Shortcuts’ two tabs? Oh wait, it’s not, it’s one, but the dark theme is so dumb that there’s no way to even tell without clicking on it.
These are usability and accessibility issues. Not just how it looks, and whether it’s pretty or not.
@micky I do understand what you’re saying, and that makes sense. But we have to balance more than just that concern. We also have to think about completely new users joining the Whonix user base who are seeing it for the first-time. And that’s a long-term consideration.
I’d like to know: has there been a noticeable increase in Whonix OVA downloads in the last 18 months? Because if so, I am quite sure that I would have contributed to that due to my Whonix advocacy. I’m not trying to pointlessly brag, just demonstrating that I care and contribute to this aspect of the Whonix project a great deal.
And @micky a theme change isn’t comparable to an entire desktop change like KDE to XFCE (which had dozens of default app changes like Dolphin -> Thunar, KWrite -> Mousepad, and Qt -> GTK-based, resulting in the need to change to a different GPG app, and much more.)
Perhaps we should consider the theming of both Whonix 15 and Whonix Host together (handling each item in both systems) to make sure we properly differentiate between them, instead of doing one, then the other.
For example, when thinking of desktop, how should it look in the VM vs in the host.
How should the terminal look in the VM vs in the host. How will it look when VM terminal and host terminal are next to each other? if they look the same or similar, and both are debian based, it will be very easy to make mistakes there.
How should taskbar look in either of them. How will it look when one task bar is above the other?
I think the host should be very different than the VMs. Just a different desktop image isn’t enough. Perhaps use similar color themes as with Qubes. Host windows: all have gray frames and headers. Workstation windows all say dark blue or green frame, Gateway windows all red.
I agree that they definitely need to look different to avoid potential confusion. This would be the default case if we apply some theming to the Workstation only: as a result they would look quite different, as the Gateway would keep its stock appearance.
We don’t necessarily need all 3 of them. Maybe just the host package.
Once there are packages [or other ways to express what belongs in which file if git is too daunting] with files in /etc/xdg/xfce4, I can sort out the packaging.
[I.e. I can sort out any issues of whonix-xfce-desktop-config fighting for ownership with same file for XFCE thinking that using config-package-dev displace.]
Sooner or later we will have requests to apply some features / themes to the host as well. Defaults are rarely what people are interested in.
It’s good to think about what we want to have in each in advance rather than react later.
I am more concerned about host theme vs VM theme than about workstation vs gateway, since the gateway is (should be) used only for non frequent configuration changes or monitoring, while the host can and will entertain a broad range of activities.
OK I finally have been able to return to some theme testing for Whonix 15 Workstation consideration.
Testing arc-theme and arc-icons in Whonix 15, this bug happens when Linux Mint is the host, but not in other hosts (such as Tails).
I think this bug is kind of a deal-breaker unfortunately, as we have to think of compatibility with widely used host OSes for Whonix. Am I right? If yes, should I move on to finding an alternative superior theme to test?
Or are hacks to the theme welcome, and we could develop our own forked theme that fixes arc’s bugs / shortcomings?
OK, I’m personally fine with that. Debian closer to stock (e.g. Tails) is what I’m moving towards anyway, away from Mint. So, should I continue doing my theme testing on W15 on Debian host and ignore that bug with Mint?
Is using files from outside of Debian somehow avoidable? Can we pretty please only use stuff which already comes from packages.debian.org?
(Background: security, difficulty to make this work in build scripts on all platforms in all network configurations, signature verification, licensing, theoretically potentially malicious icons specifically crafted only for Whonix, Whonix Default Application Policy )
Here is the list of all “icon-theme” packages in Debian: https://packages.debian.org/search?keywords=icon-theme
Personally, I don’t like most of them.
If I had to choose, I would just use the default gnome theme (already installed) or Adwaita. I just dislike the current default one - Gnome Brave - which looks childish to me.