Tails

#4 Updated by intrigeri 2019-03-10 15:44:12

@hefee, thanks for creating this ticket :)

#5 Updated by intrigeri 2019-03-23 18:16:35

> Please add your comments and the start of the list of packages etc.

So I’ve read a private discussion from 1.5 year ago, extracted the relevant bits, and compiled the following list (on some of those I’ve changed my mind since ;)

What I think is definitely in:

• Tools required for general use cases that we want to support (e.g. Browser, email client)
• Adapt Tails Installer or the Persistence assistant to a new version of GTK, iff. we have to (e.g. the old version of GTK is going away, or becomes poorly supported in current GNOME, or integrates poorly in current GNOME)
• Automatic upgrades, e.g.:
  • Adapt Tails Upgrader to a change in underlying technology, iff. that change is being dictated to us (e.g. UDisks API change)
  • Switch to a different upgrade system if/once there’s a better realistic alternative to our in-house thing that only one person can touch
• Adapt WhisperBack to a change in Onion Services, because it’s a tool that’s beneficial for the project (and not only to our users): we need bug reports
• Adapt to a new KeePassX upstream version (or new upstream entirely)
• Avoid regressions in accessibility technologies and input methods
• Additional Software: we base more and more decisions on the fact we have this feature, and I think it’s good
• XMPP+OTR, until we support other acceptable IM solutions that are compatible with the comms many more people use nowadays
• AppArmor for GUI apps, until we have better sandboxing
• Sandboxing of non-GUI key software such as tor

Unclear, needs some user research and/or strategic decisions:

• OpenPGP Applet and more generally, OpenPGP support apart of the basics (to be defined): see discussion on Bug #8310
• VeraCrypt support (once it’s fully upstreamed it’ll be less maintenance work but what if it’s never fully upstreamed? thankfully we’ll soon have survey results that should help us assess how important this is in the end)
• Memory erasure on shutdown and on boot medium removal: it’s one of our “signature” features but OTOH, we’re not even sure it has made a practical difference in 1+ cases.

Not in:

• Electrum
• “Niche” utilities: MAT, Onion Circuits
• IRC support: mostly used by power-users who’ll have no problem installing their preferred client themselves (which, incidentally, might not be Pidgin, and might run on a remote server)
• RSS/Atom feed support

Finally, wrt. whether brand new features developed outside of the FT become part of the core Tails system that the FT is supposed to maintain: in theory they do, but we need to take into account our skill set, and whenever the new thing is way outside of our comfort zone, either pick another technology (that we are in a position to maintain) or budget maintenance work by external contractors; better think about this before the feature is ready.

#9 Updated by intrigeri 2019-05-18 06:16:36

intrigeri wrote:
> Unclear, needs some user research and/or strategic decisions:
>
> * OpenPGP Applet and more generally, OpenPGP support apart of the basics (to be defined): see discussion on Bug #8310

sajolida’s conclusion on Bug #16709 is “we should […] consider the possibility of doing OpenPGP operations from the GUI a core feature of Tails”.

#16 Updated by intrigeri 2019-09-06 14:12:56

So far, folks who volunteered to be part of the working group are: intrigeri, sajolida, segfault, u.

#18 Updated by hefee 2019-09-16 16:39:29

me is interessed too in defining the core code base.

#24 Updated by emmapeel 2020-02-21 08:50:55

Bug Bug #17183 is giving hard times to our privacy minded newbie users, we have many users writing to frontdesk about this problem.