Updated to latest version, and issue was fixed.

This will require changes in Laniakea, which is currently stuck in a PostgreSQL transition. It has been on my todo list already though to revise the synchronization process.
The likeliest chance is that these issues are from an older, broken sync (the more annoying option would be that this is an issue that still exists).
In any case, it's quite high up on the task list (I was aware of the general issue so far, but didn't know that it affected so many packages).

Dropped from Debian testing (and consequently from PureOS), and unlikely to reappear without addressing the bug.

Fixed in Debian.

Same issue here

I knew about this, it just didn't have a very hight priority yet due to the focus being on the OEM install (it's on my todo list now though).

In T140#3063, @ezs777 wrote:

ii purebrowser 46.2.0esr+re amd64 Pureos web browser

Desired=Unknown/Install/Remove/Purge/Hold

+++-==============-============-============-=================================
ii purebrowser 46.2.0esr+re amd64 Pureos web browser

In T140#3059, @ezs777 wrote:

I've updated Purebrowser to the latest version

I've updated Purebrowser to the latest version (and restarted both Purebrowser and my computer), and I still cannot access sites like youtube.com, without setting security.ssl.enable_alpn to "False".

no worries guys, we will do something :)

An idea for a quick workaround:

gsettings set org.gnome.desktop.lockdown disable-lock-screen true

maybe something like this (not tested, but you get "the spirit") can be executed at same time as launching the distro-installer.

I think that the password is "live".

46.2.0esr+really45.4.0esr+pureos-1pureos1 <- epic :D

This is upstream bug, it probably needs correct metadata.

Fixed in

(as a sidenote I also realize that a web browser can execute code from alien sources and similarly cannot be ensured to only process Freely licensed code - I have no idea by which logic FSF can approve operating systems containing web browsers...)

I don't understand BOINC well enough to say how exactly the distributed code gets fed into the system, only (believe that I) understand that it does happen and that we cannot ensure that code fed into the system is freely licensed.

Is it that the programs sent by the individual projects for the clients to calculate are often nonfree?

Fixed in purebrowser 46.2.0esr-1pureos2.

adding @theodotos.andreou so he knows about the need of service similar to debian.net subdomain. Coordinate that together.

Plan: Establish a web service as a debian.net subdomain:

In most of the World, copyright is automatic (i.e. there is an implicit "all rights reserved"), and as a consequence licensing must be explicit too (and arguably copyright statement must be explicit too, to be able to validate if the licensing was granted by the copyright holder and not bogusly from someone else not in possesion of rights granted).

no, by default things without license are considered proprietary unless someone did investigation and put it in pd if it suits that. If you want you could try contact authors of those addons and ask them to specify license

Yes, but if a plugin does not have a license tag, can we consider it public domain?

Not sure what you are asking.

What about those which don't have license field?

Upstream repository feed:

More sustainable than subscribing to a non-default repository - whether maintained by us or picking a third-party repository - is to setup a proxy for the feed which filters the repository feed and lets through only those with a license known to be Free.

Parabola patch is therefore overzealous: Only the three lines changing URL are relevant.

Parabola issue 1035 really tracks two different issues: a) contains non-free decompression engine for RAR archives, and b) support for non-free addons.

I did not remove any package using bleachbit.

This actually looks like the decryption worked okay and LVM is good.
Did you get this message *after* you entered your password to decrypt the device?

much better :) sorry

Since you apparently are using a fully encrypted disk, it would mean that any data would be lost when you reinstall.

Thanks for your help, highly appreciated !!!
Newbie with linux, my question may seem very stupid, but if I'm reinstalling PureOs do I keep my home directory folder datas safe (like with Mac system) or do they get 'crashed' (like with windows system) ?

Do not use Bleachbit - ever. See https://blog.martin-graesslin.com/blog/2012/07/do-not-use-bleachbit/
I can look into this later this week, to maybe figure out what's wrong, but if you can, reinstalling the OS is probably the best idea.

Hopefully that will help...

I agree with the conclusion done by Trisquel: While Debian arguably tolerate Free clients pulling non-free code from external services, PureOS cannot do that when seeking FSF endorsement.

Debian package does a fair job of stripping non-free sources, including the convenience code copy of glut code.

Upstream is dead, and alternatives exist. Seems just a matter of time before it is dropped in Debian.

Only Matthias can block packages.

GeoGebra licenses its documentation as CC-BY-SA-3.0, since at least version 3.2.40.0 packaged prior to initial release in Debian in 2010.

I am not trying to be fair nor compare them as technical solution - the idea of Tox was to have some default for PureOS secure chat/videoconferencing app but now that it seems that Matrix with its clients is going to end up being default for PureOS, I think we can safely disregard tox for now (at least until it gets properly packaged into Debian and is more end-user usable than Matrix).

As I understand it, QTox is different from Matrix:

Hm, looking the bug report (which is already couple of years old and the last msg was from 2 years ago), it seems this is stuck. I do wonder what is the current state of tox itself (especially now that matrix and riot are on rise).

oh well - it _is_ browser agnostic already (se Debian bugreport).

copyright referenced at the Github page seems to not cover documentation, and also what seems to be copyright portected in the documentation package is only the _packaging_ of it: Manoj is a Debian developer, and at least none of the few documentation files I randomly looked at were authored by Manoj.

Bogus issue: Does not affect freedoms.

I recommend *against* pulling in third-party code repositories.

This recent Debian thread looks related: http://lists.alioth.debian.org/pipermail/pkg-javascript-devel/2017-September/020574.html

Better to be browser-agnostic by use of xdg-open

Both cups-filters and inkscape recommend imagemagick.

Removed from Debian testing (for different reason, but expected to not re-enter without addressing this bug first).

it is only matter of updated package in this case.

Good to know, we will maybe need to check some diff in Fedora packages but I assume that it is only matter of updated package in this case.

Thank you for the answer.

Ok, I'll retest in a month, it will probably work because meanwhile, for professional needs (I'll teach Krita course in Paris in a week), I installed a Fedora 26 and didn't ran into this issues on same hardware.