Re: TODOS

Choose a license (I used MIT because Nixpkgs uses it, but we could also use something else)

For inclusion in this repo MIT is best.

Determine if we want to fallback to other files than /etc/os-release

Can be done later. Should not delay this PR.

Also, just support the standard lines.

Verify the output (LGTM, but I might have missed something)

This script has an additional newline at the end of the output.

Except that identical to debian and ubuntu. 👍

[davidak@nixos:~]$ /nix/store/slm41d7vfqc6i429d0hn5341li3pi0m6-lsb_release/bin/lsb_release -a
No LSB modules are available.
Distributor ID:	NixOS
Description:	NixOS 19.03.172866.4649b6ef4b5 (Koi)
Release:	19.03.172866.4649b6ef4b5
Codename:	koi

root@d646ab96b21c:/# lsb_release -a
No LSB modules are available.
Distributor ID:	Debian
Description:	Debian GNU/Linux 9.5 (stretch)
Release:	9.5
Codename:	stretch

root@ab88485573f0:/# lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 18.04.2 LTS
Release:	18.04
Codename:	bionic

[root@5205a23e9de4 /]# lsb_release -a
LSB Version:	:core-4.1-amd64:core-4.1-noarch:cxx-4.1-amd64:cxx-4.1-noarch:desktop-4.1-amd64:desktop-4.1-noarch:languages-4.1-amd64:languages-4.1-noarch:printing-4.1-amd64:printing-4.1-noarch
Distributor ID:	Fedora
Description:	Fedora release 30 (Thirty)
Release:	30
Codename:	Thirty

[root@ba72babf3e28 /]# lsb_release -a
LSB Version:	1.4
Distributor ID:	Arch
Description:	Arch Linux
Release:	rolling
Codename:	n/a

This script has an additional newline at the end of the output.

That also tricked me but this is due to the configuration of nix-shell, which prints an additional newline after each command (if you pipe to vim, xxd, etc. there's only one newline like on Debian, etc.).
Not sure where this is configured.

But I just noticed that the script from Debian prints No LSB modules are available. to stderr, I'll change that in the next commit.

Thanks for pasting the outputs on other operating systems, that's really helpful :) (I only compared it to the output on Debian stretch)

I think the newline is no problem. It's only different.

Thanks for pasting the outputs on other operating systems, that's really helpful :) (I only compared it to the output on Debian stretch)

I just used docker to quickly spin up an instance. Perfectly for such tests :)

I think the newline is no problem. It's only different.

What I meant is that this is caused by the shell, not by the program (lsb_release):

$ lsb_release -a
No LSB modules are available.
Distributor ID:	NixOS
Description:	NixOS 19.03.git.50e5e3a (Koi)
Release:	19.03.git.50e5e3a
Codename:	koi

$ PS1='\[\033[1;32m\][nix-shell:\w]\$\[\033[0m\] '
$ lsb_release -a
No LSB modules are available.
Distributor ID:	NixOS
Description:	NixOS 19.03.git.50e5e3a (Koi)
Release:	19.03.git.50e5e3a
Codename:	koi
$ 

The prompt is configured to print an additional newline (via PS1, I did initially miss that, when I had a look).

So the output should actually be identical now (apart from the actual values of course).

@primeos is this ready for merging or do you want to change anything before?

@primeos is this ready for merging or do you want to change anything before?

@davidak Sorry for the delay, I ran out of time (the time of this PR was a bit suboptimal).
I just checked the TODOs and it should be ready to go :)

cc @dtzWill @rardiol (previous contributors to lsb-release): Any objections? If not I would merge this PR in a few days.

This pull request has been mentioned on Nix community. There might be relevant details there:

https://discourse.nixos.org/t/prs-already-reviewed/2617/38

If not I would merge this PR in a few days.

Unfortunately I was pretty busy and lost track of this PR...
I just ran the rebuilds and unfortunately this broke the build for radiotray-ng.
Turns out that the output wasn't backwards compatible if /etc/os-release is missing (e.g. in the Nix build sandbox).

I did try my best to make the output as backwards compatible as possible (at least without /etc/os-release as the current binary doesn't work correctly outside of the sandbox and is therefore a bad example).
If the rebuilds succeed I'll merge this PR - let's hope there won't be any problematic regressions... (but in that case we should still be able to modify the script accordingly).

Unfortunately the code isn't really nice anymore but I guess it'll have to do for now... (we can still refactor it later if it works as intended). Also: We might want to change the default values if we're in a Nix build sandbox to increase the usefulness of the new script (currently it only helps at run-time).