[presentations] 04/04: GUUG: update images and numbers
Holger Levsen
holger at layer-acht.org
Thu Mar 23 16:59:47 UTC 2017
This is an automated email from the git hooks/post-receive script.
holger pushed a commit to branch master
in repository presentations.
commit ac1abfecc20008ad91bb20b2b22236cf34a36f99
Author: Holger Levsen <holger at layer-acht.org>
Date: Thu Mar 23 17:58:09 2017 +0100
GUUG: update images and numbers
Signed-off-by: Holger Levsen <holger at layer-acht.org>
---
2017-03-24-GUUG/2017-03-24-GUUG.tex | 36 ++++++++++-----------
2017-03-24-GUUG/TODO | 5 +--
.../images/stats_bugs_sin_ftbfs_state.png | Bin 28257 -> 28133 bytes
2017-03-24-GUUG/images/stats_pkg_state.png | Bin 30283 -> 0 bytes
2017-03-24-GUUG/images/stats_pkg_state_testing.png | Bin 32427 -> 32523 bytes
.../images/stats_pkg_state_unstable.png | Bin 33669 -> 34133 bytes
6 files changed, 19 insertions(+), 22 deletions(-)
diff --git a/2017-03-24-GUUG/2017-03-24-GUUG.tex b/2017-03-24-GUUG/2017-03-24-GUUG.tex
index bc8e010..774cb27 100644
--- a/2017-03-24-GUUG/2017-03-24-GUUG.tex
+++ b/2017-03-24-GUUG/2017-03-24-GUUG.tex
@@ -424,8 +424,8 @@ Darmstadt, Germany\\
\texttt{experimental}
\item Also testing: coreboot, OpenWrt, LEDE, NetBSD, FreeBSD,
Arch Linux, Fedora and soon F-Droid too
- \item 44 nodes (amd64/i386/arm64/armhf), 200 cores and 1 TB RAM
- \item 486 jenkins jobs running on jenkins.debian.net
+ \item 46 nodes (amd64/i386/arm64/armhf), >200 cores and >1 TB RAM
+ \item 502 jenkins jobs running on jenkins.debian.net
\item 43 scripts in Python and Bash, 283 lines of code in average
\item 37 contributors for \texttt{jenkins.debian.net.git}
\end{itemize}
@@ -450,15 +450,15 @@ domainname & \texttt{debian.net} & \texttt{i-capture-the-domainname} \\
uid & \texttt{1111} & \texttt{2222} \\
gid & \texttt{1111} & \texttt{2222} \\
UTS namespace & shared with the host & \textit{modified using \texttt{/usr/bin/unshare --uts}} \\
-kernel version & Linux 3.16 or 4.X & on amd64 always varied, on armhf
-sometimes \\
+kernel version & Linux 3.16 or 4.X & on amd64 and arm64 always varied, on armhf sometimes,
+on i386 32/64bit kernel variation instead \\
umask & 0022 & 0002 \\
CPU type & \multicolumn{2}{l}{varied on i386} \\
- & on armhf varied a bit, not on amd64 & \\
-filesystem & \multicolumn{2}{l}{same for both builds on amd64: (\texttt{tmpfs}), on armhf \texttt{ext3/4}} \\
+ & on armhf varied a bit, not on amd64 nor arm64 & \\
+filesystem & \multicolumn{2}{l}{same for both builds on amd64, i386 and arm64: (\texttt{tmpfs}), on armhf \texttt{ext3/4}} \\
& & \textit{(and we have} \texttt{disorderfs}\textit{, but the code is disabled)} \\
-year, month, date & \multicolumn{2}{l}{on amd64: 398 days variation, on armhf not yet} \\
-hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minute differs… } \\
+year, month, date & \multicolumn{2}{l}{on amd64, arm64 and i386: 398 days variation, on armhf not yet} \\
+hour, minute & \multicolumn{2}{l}{hour and minute deterministically and non-deterministically varied… } \\
\textit{everything else} & \multicolumn{2}{l}{\textit{is likely the same…}}
\end{tabular}
}
@@ -545,7 +545,7 @@ hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minut
};
\end{tikzpicture}
\begin{center}
- \footnotesize{23,405 (93.3\%) out of 25,067 source packages are reproducible \\
+ \footnotesize{23,378 (93.8\%) out of 24,909 source packages are reproducible \\
in our test framework on \texttt{amd64}}
\vfill
\end{center}
@@ -559,7 +559,7 @@ hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minut
};
\end{tikzpicture}
\begin{center}
- \footnotesize{20,309 (78.9\%) out of 25,734 source packages are reproducible \\
+ \footnotesize{20,597 (79.2\%) out of 25,982 source packages are reproducible \\
in our test framework on \texttt{amd64}} (difference due to build path variations)
\vfill
\end{center}
@@ -571,10 +571,10 @@ hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minut
\begin{itemize}
\item \url{https://reproducible.debian.net/$src}
\item 48 package sets
- \item 282 categorised distinct issues
- \item 7,413 notes
- \item 1,595 unreproducible packages in \texttt{stretch/amd64} (testing), but only
- 111 without a note (5,288 in \texttt{unstable} but also only 154 without a
+ \item 292 categorised distinct issues
+ \item 6,604 notes
+ \item 1,473 unreproducible packages in \texttt{stretch/amd64} (testing), but only
+ 90 without a note (5,253 in \texttt{unstable} but also only 149 without a
note)
\item maintained in \texttt{notes.git} by 49 contributors
\item currently Debian only, but cross distro notes are planned
@@ -628,8 +628,8 @@ hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minut
\begin{frame}
\frametitle{Debian summary / What's left to do}
\begin{itemize}
- \item This is/was a proof-of-concept, Debian is neither 93.3\% reproducible nor
- 78.9\%. (and 10\% > 2,500 sources packages!)
+ \item This is/was a proof-of-concept, Debian is neither 93.8\% reproducible nor
+ 79.2\%. (and 10\% > 2,500 sources packages!)
\item<2-3> All our required changes are finally in Debian now!
\item<2-3> Debian 9, "stretch", will only be partially reproducible.
\item<2-3> Because, Debian does not (yet?) do full rebuilds before
@@ -838,8 +838,8 @@ hour, minute & \multicolumn{2}{l}{hour is usually the same… usually, the minut
\begin{itemize}
\item Individuelly signed checksums (think web of trust) could work in the
Debian case (we have a gpg web of trust), but IMO won't scale.
- \item { Another idea: rebuilders, run by large organisations
- (ACLU, BSI, CCC, Deutsche Bank, Greenpeace, GUUG, NASA, NSA).}
+ \item { Another idea: rebuilders, run by large organisations,
+ eg. ACLU, BSI, CCC, Deutsche Bank, Greenpeace, GUUG, NASA, NSA, etc…}
\item Fedora rebuilds Debian, Debian rebuilds openSUSE, openSUSE rebuilds
NetBSD, etc…
\item Big customers could just rebuild everything themselves.
diff --git a/2017-03-24-GUUG/TODO b/2017-03-24-GUUG/TODO
index 379e7ff..335accc 100644
--- a/2017-03-24-GUUG/TODO
+++ b/2017-03-24-GUUG/TODO
@@ -4,13 +4,10 @@ meta:
Progress
Details
summary
- variation misses arm64
- update graphs
+ variation: maybe go back to devconf version…
add slide: the *100th* RBWN issue!
♥ Lunar
add slide: netbsd vs freebsd, mention electrobsd
- add slide: fedora
- add slide: openSUSE
add slide: coreboot, LEDE/OpenWrt
add slide: S_D_P_M
add slide: debian now. coccia, stretch. aim for buster. stretch is boring.
diff --git a/2017-03-24-GUUG/images/stats_bugs_sin_ftbfs_state.png b/2017-03-24-GUUG/images/stats_bugs_sin_ftbfs_state.png
index 917903b..9a063db 100644
Binary files a/2017-03-24-GUUG/images/stats_bugs_sin_ftbfs_state.png and b/2017-03-24-GUUG/images/stats_bugs_sin_ftbfs_state.png differ
diff --git a/2017-03-24-GUUG/images/stats_pkg_state.png b/2017-03-24-GUUG/images/stats_pkg_state.png
deleted file mode 100644
index 37ccd22..0000000
Binary files a/2017-03-24-GUUG/images/stats_pkg_state.png and /dev/null differ
diff --git a/2017-03-24-GUUG/images/stats_pkg_state_testing.png b/2017-03-24-GUUG/images/stats_pkg_state_testing.png
index 865b46a..4cb9666 100644
Binary files a/2017-03-24-GUUG/images/stats_pkg_state_testing.png and b/2017-03-24-GUUG/images/stats_pkg_state_testing.png differ
diff --git a/2017-03-24-GUUG/images/stats_pkg_state_unstable.png b/2017-03-24-GUUG/images/stats_pkg_state_unstable.png
index 9b0e33b..067698d 100644
Binary files a/2017-03-24-GUUG/images/stats_pkg_state_unstable.png and b/2017-03-24-GUUG/images/stats_pkg_state_unstable.png differ
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/reproducible/presentations.git
More information about the Reproducible-commits
mailing list