Reintroduce camouflage in Tails

The Windows camouflage was dropped from Tails 2.0 (January 2016) because it was not ported to GNOME 3.14:

https://tails.boum.org/news/windows_camouflage_jessie/

https://tails.boum.org/news/version_2.0/

6 years later, we decided that we won't introduce a Windows camouflage again. We still understand the potential benefits but we don't think that they outweigh the costs and that our team should spend resources on this:

• The camouflage can help hiding that you are using Tails to people who can see your screen. It felt and looked super cool (that's important!) but it was not a reliable security solution either (that's important too!).

  For example, for it to avoid suspicion in a cybercafé or public library, or computer room in a school, it would have to mimic the particular version of Windows or macOS used in the neighboring computers. That's quite more complex than allowing users to switch to 1 alternative GNOME theme.

• Even external contributors want to help with a new camouflage, it will always require our core team to do some work for it to happen (answering questions, reviewing code, etc.). Plus, it becomes something that our team will have to maintain in the future. Lack of maintenance work is precisely what caused the original camouflage to be dropped.

• Before a user starts the session, it would be extra hard to make the Boot Loader or Welcome Screen look like Windows or macOS. Once the user started using Tails and have several windows open, the screen looks more like a "browser with a website on it" than Tails. At most it looks like a Linux distribution to people close enough to see the menus and window frames.

We should instead work on allowing a custom desktop background (that people could configure to look like Windows or macOS) and maybe a custom GNOME theme. (#7148)

It would be less initial work, less maintenance, and would cover more use cases.

Customization is an important emotional aspect of technology. For example, custom background was in the top 5 feature requests in the 2020 survey, along with Monero, VPN, and persistent Tor Browser settings:

https://gitlab.tails.boum.org/tails/blueprints/-/wikis/user_survey/openpgp_and_pidgin#general-purpose-questions

#10830 (closed) was rejected 4 years ago by lack of progress. I think that we should use this thread on tails-dev to clarify whether we might accept contributions to a new camouflage or not, and document this on #10830 (closed) for the future.

My opinion is "no", in favor of #7148 but I'd like to hear more from the rest of the team.