Skip to content
Commits on Source (2)
......@@ -159,44 +159,38 @@ https://reproducible-builds.org
* Once upon a time
#+ATTR_BEAMER: :overlay <+->
- a list mail in 1997, very few more in 2001 and 2003.
- A list mail in 1997, very few more in 2001 and 2003.
- Then, in 2011 and 2012, Bitcoin and Torbrowser were made reproducible.
- Wow.
* Debian
* Why unreproducibilities exist (historically)
#+ATTR_BEAMER: :overlay <+->
- then in 2013 Debian started to investigate this
- and kicked it off in 2014.
- then CII came in in 2015.
- since 2017 in Debian Policy, as a "should" directive, not "must".
- 2023 with "must"?
- Historically software was reproducible! Every bit counted.
- And every bit was known.
- Bit for bit reproducible GNU toolchain in the early 90s on 10(?) architectures.
- *And then we all forgot.*
* Four summits so far
* Debian
- Athens 2015
- Berlin 2016
- Berlin 2017
- Paris 2018
#+ATTR_BEAMER: :overlay <+->
- Marrakesh 2019
* Collaboration is great
FIXME: Show jelle's light
- Then in 2013 some people in Debian started to investigate this.
- And kicked it off in 2014 by introducing systematic testing, classifications and weekly blogs.
- Since 2017 in Debian Policy, as a "should" directive, not "must".
- 2023 with "must"?
* Collaboration is great
* Debian main unstable/amd64, since October 2014
#+ATTR_BEAMER: :overlay <+->
- Debian, SuSE, Archlinux, OpenWrt, Guix, Nix, FreeBSD, NetBSD, 'random' 'strangers', ...
- we welcome you.
- we welcome free software.
[[./images/stats_pkg_state.png]]
* Why unreproducibilities exist
* Shared research and developments / WIP
#+ATTR_BEAMER: :overlay <+->
- Historically software was reproducible! Every bit counted.
- GNU toolchain in the early 90s on 10(?) architectures
- Test/research setup for many but not all projects.
- Since end of 2018 shared db for some of those.
- Sharing issues, patches and upstreaming them.
- Shared public blog, now called monthly report.
- More collaboration is possible!
* Why unreproducibilities exist
......@@ -206,26 +200,50 @@ FIXME: Show jelle's light
- timestamps
- build paths
- timezones, locales
- hundries different other reasons
- hundries different classes of causes !
- It's fun to discover these! Well, mostly.
* A light at the end of the forest?
Diffoscope!
- Diffoscope!
- FIXME. explain. pics. urls.
* Theory vs Praxis
#+ATTR_BEAMER: :overlay <+->
- 93% is a lie
- 93% is a lie.
- Getting software reproducible in theory is 33% of the way.
- The next 33% is to do it in practice.
- And the last mile is on the users machines. "Enable everyone to independently..."
- The next 33% is to do it in practice, to change the distros tools and workflows. Not always easy.
- And the last 33% are about distributing trust and/or "Enable everyone to independently..." for real. And that again is different for each distro.
* The end
* Four summits so far
#+ATTR_BEAMER: :overlay <+->
- Athens 2015
- Berlin 2016
- Berlin 2017
- Paris 2018
- Marrakesh 2019
* Collaboration is so great, again.
[[./images/reprobuilds-display.jpeg]]
* Collaboration, again.
#+ATTR_BEAMER: :overlay <+->
- We stand on the shoulders of giants.
- And elves and dwarfs, unicorns, women and men, and some witches and magicans.
- And we welcome you.
- We welcome Free Software.
* The end / summary
#+ATTR_BEAMER: :overlay <+->
- We made 93% of the first 33%.
- Currently. There's new software every hour.
- And for some distros, half or a third of the 2nd 33%.
- in-toto or Benjamin's Hof stuff or a combination / something new got us partly on the last 33%, but we can't go on that path without the 2nd 33%
- There are ideas and even code for the last 33%, but we can't go on that path without the 2nd 33%...
- *There is a lot to do. Please. Help.*
* Questions
......@@ -276,3 +294,7 @@ Thank you for your time and contributions.
https://creativecommons.org/licenses/by-sa/4.0/deed.en
https://en.wikipedia.org/wiki/File:Ring-eye-sauron.gif
reprobuilds-display from Jelle is under MIT:
https://github.com/jelly/reproduciblebuilds-display