[med-svn] r2388 - trunk/community/talks/200808_debconf8
tille at alioth.debian.org
tille at alioth.debian.org
Mon Aug 4 19:01:31 UTC 2008
Author: tille
Date: 2008-08-04 19:01:30 +0000 (Mon, 04 Aug 2008)
New Revision: 2388
Added:
trunk/community/talks/200808_debconf8/0fix_ralf_edu
trunk/community/talks/200808_debconf8/author_stats
trunk/community/talks/200808_debconf8/debian-edu.png
trunk/community/talks/200808_debconf8/debian-enterprise.png
trunk/community/talks/200808_debconf8/debian-jr.png
trunk/community/talks/200808_debconf8/debian-lex.png
trunk/community/talks/200808_debconf8/debian-med.png
Removed:
trunk/community/talks/200808_debconf8/author_stats_cdd
trunk/community/talks/200808_debconf8/author_stats_edu
trunk/community/talks/200808_debconf8/author_stats_med
Modified:
trunk/community/talks/200808_debconf8/debian-med.tex
trunk/community/talks/200808_debconf8/get-archive-pages
Log:
Commit work from traveling in the plane
Added: trunk/community/talks/200808_debconf8/0fix_ralf_edu
===================================================================
--- trunk/community/talks/200808_debconf8/0fix_ralf_edu (rev 0)
+++ trunk/community/talks/200808_debconf8/0fix_ralf_edu 2008-08-04 19:01:30 UTC (rev 2388)
@@ -0,0 +1,11 @@
+#!/bin/sh -e
+# I have no idea why the same author Ralf Gesellensetter has
+# so many variants of spelling in the index - but tis disturbs
+# the stats and so it is fixed here
+
+psql cddlistarchives << EOT
+begin;
+update listarchive set author = 'Ralf Gesellensetter' where project = 'edu' and author like 'Ralf%setter' ;
+commit;
+EOT
+
Property changes on: trunk/community/talks/200808_debconf8/0fix_ralf_edu
___________________________________________________________________
Name: svn:executable
+ *
Added: trunk/community/talks/200808_debconf8/author_stats
===================================================================
--- trunk/community/talks/200808_debconf8/author_stats (rev 0)
+++ trunk/community/talks/200808_debconf8/author_stats 2008-08-04 19:01:30 UTC (rev 2388)
@@ -0,0 +1,43 @@
+#!/bin/sh -e
+# This script draws a plot to compare mailing list activities on
+# CDDs and related projects.
+
+if [ "$1" = "" ] ; then
+ echo "Usage: `basename $0` <cddname>"
+ exit 1
+fi
+
+
+NAME=authorstat_"$1"
+DATFILE="${NAME}_year.dat"
+psql -t cddlistarchives -c "SELECT BuildQueryAuthorsYear('$1', 10) ;" | \
+ psql cddlistarchives \
+ >"$DATFILE"
+
+sed -i -e '/^[-+]\+$/d' -e '/^([0-9]\+ [A-Za-z]\+)$/d' \
+ -e 's/[[:space:]]*|[[:space:]]*/\t/g' \
+ -e 's/è/è/g' -e 's/ö/ö/g' \
+ -e 's/®//g' -e 's/é/e/' \
+ "$DATFILE"
+
+R --no-save <<EOT
+library(plotrix)
+dmstats <- read.table(file='$DATFILE', sep = '\t', fill=TRUE, header=TRUE )
+# png("dmstats.png", width = 800, height = 600)
+textcolor="yellow"
+pdf("${NAME}.pdf", fg=textcolor)
+par(col.axis=textcolor,col.main=textcolor)
+dmstats.mat <- as.matrix(dmstats)[,2:9]
+rownames(dmstats.mat) <- dmstats[['year']]
+dmstats.mat <- t(dmstats.mat)
+
+# barplot(dmstats.mat,beside=TRUE,col=rainbow(3),legend.text=TRUE)
+barplot(dmstats.mat,beside=TRUE,col=rainbow(8))
+# plot(dmstats.mat,col=rainbow(8))
+legend(x="topleft", colnames(dmstats[,2:9]),fill=rainbow(8), inset=0.05,
+ text.col=textcolor
+ )
+
+EOT
+
+rm $DATFILE
Property changes on: trunk/community/talks/200808_debconf8/author_stats
___________________________________________________________________
Name: svn:executable
+ *
Deleted: trunk/community/talks/200808_debconf8/author_stats_cdd
===================================================================
--- trunk/community/talks/200808_debconf8/author_stats_cdd 2008-08-04 13:49:42 UTC (rev 2387)
+++ trunk/community/talks/200808_debconf8/author_stats_cdd 2008-08-04 19:01:30 UTC (rev 2388)
@@ -1,33 +0,0 @@
-#!/bin/sh
-# This script draws a plot to compare mailing list activities on
-# CDDs and related projects.
-
-NAME=authorstat_cdd
-DATFILE="${NAME}_year.dat"
-psql -t cddlistarchives -c "SELECT BuildQueryAuthorsYear('custom', 10) ;" | \
- psql cddlistarchives \
- >"$DATFILE"
-
-sed -i -e '/^[-+]\+$/d' -e '/^([0-9]\+ [A-Za-z]\+)$/d' -e 's/[[:space:]]*|[[:space:]]*/\t/g' "$DATFILE"
-
-R --no-save <<EOT
-library(plotrix)
-dmstats <- read.table(file='$DATFILE', sep = '\t', fill=TRUE, header=TRUE )
-# png("dmstats.png", width = 800, height = 600)
-textcolor="yellow"
-pdf("${NAME}.pdf", fg=textcolor)
-par(col.axis=textcolor,col.main=textcolor)
-dmstats.mat <- as.matrix(dmstats)[,2:9]
-rownames(dmstats.mat) <- dmstats[['year']]
-dmstats.mat <- t(dmstats.mat)
-
-# barplot(dmstats.mat,beside=TRUE,col=rainbow(3),legend.text=TRUE)
-barplot(dmstats.mat,beside=TRUE,col=rainbow(8))
-# plot(dmstats.mat,col=rainbow(8))
-legend(x="topleft", colnames(dmstats[,2:9]),fill=rainbow(8), inset=0.05,
- text.col=textcolor
- )
-
-EOT
-
-rm $DATFILE
Deleted: trunk/community/talks/200808_debconf8/author_stats_edu
===================================================================
--- trunk/community/talks/200808_debconf8/author_stats_edu 2008-08-04 13:49:42 UTC (rev 2387)
+++ trunk/community/talks/200808_debconf8/author_stats_edu 2008-08-04 19:01:30 UTC (rev 2388)
@@ -1,33 +0,0 @@
-#!/bin/sh
-# This script draws a plot to compare mailing list activities on
-# CDDs and related projects.
-
-NAME=authorstat_edu
-DATFILE="${NAME}_year.dat"
-psql -t cddlistarchives -c "SELECT BuildQueryAuthorsYear('edu', 10) ;" | \
- psql cddlistarchives \
- >"$DATFILE"
-
-sed -i -e '/^[-+]\+$/d' -e '/^([0-9]\+ [A-Za-z]\+)$/d' -e 's/[[:space:]]*|[[:space:]]*/\t/g' "$DATFILE"
-
-R --no-save <<EOT
-library(plotrix)
-dmstats <- read.table(file='$DATFILE', sep = '\t', fill=TRUE, header=TRUE )
-# png("dmstats.png", width = 800, height = 600)
-textcolor="yellow"
-pdf("${NAME}.pdf", fg=textcolor)
-par(col.axis=textcolor,col.main=textcolor)
-dmstats.mat <- as.matrix(dmstats)[,2:9]
-rownames(dmstats.mat) <- dmstats[['year']]
-dmstats.mat <- t(dmstats.mat)
-
-# barplot(dmstats.mat,beside=TRUE,col=rainbow(3),legend.text=TRUE)
-barplot(dmstats.mat,beside=TRUE,col=rainbow(8))
-# plot(dmstats.mat,col=rainbow(8))
-legend(x="topleft", colnames(dmstats[,2:9]),fill=rainbow(8), inset=0.05,
- text.col=textcolor
- )
-
-EOT
-
-rm $DATFILE
Deleted: trunk/community/talks/200808_debconf8/author_stats_med
===================================================================
--- trunk/community/talks/200808_debconf8/author_stats_med 2008-08-04 13:49:42 UTC (rev 2387)
+++ trunk/community/talks/200808_debconf8/author_stats_med 2008-08-04 19:01:30 UTC (rev 2388)
@@ -1,33 +0,0 @@
-#!/bin/sh
-# This script draws a plot to compare mailing list activities on
-# CDDs and related projects.
-
-NAME=authorstat_med
-DATFILE="${NAME}_year.dat"
-psql -t cddlistarchives -c "SELECT BuildQueryAuthorsYear('med', 10) ;" | \
- psql cddlistarchives \
- >"$DATFILE"
-
-sed -i -e '/^[-+]\+$/d' -e '/^([0-9]\+ [A-Za-z]\+)$/d' -e 's/[[:space:]]*|[[:space:]]*/\t/g' "$DATFILE"
-
-R --no-save <<EOT
-library(plotrix)
-dmstats <- read.table(file='$DATFILE', sep = '\t', fill=TRUE, header=TRUE )
-# png("dmstats.png", width = 800, height = 600)
-textcolor="yellow"
-pdf("${NAME}.pdf", fg=textcolor)
-par(col.axis=textcolor,col.main=textcolor)
-dmstats.mat <- as.matrix(dmstats)[,2:9]
-rownames(dmstats.mat) <- dmstats[['year']]
-dmstats.mat <- t(dmstats.mat)
-
-# barplot(dmstats.mat,beside=TRUE,col=rainbow(3),legend.text=TRUE)
-barplot(dmstats.mat,beside=TRUE,col=rainbow(8))
-# plot(dmstats.mat,col=rainbow(8))
-legend(x="topleft", colnames(dmstats[,2:9]),fill=rainbow(8), inset=0.05,
- text.col=textcolor
- )
-
-EOT
-
-rm $DATFILE
Added: trunk/community/talks/200808_debconf8/debian-edu.png
===================================================================
(Binary files differ)
Property changes on: trunk/community/talks/200808_debconf8/debian-edu.png
___________________________________________________________________
Name: svn:mime-type
+ application/octet-stream
Added: trunk/community/talks/200808_debconf8/debian-enterprise.png
===================================================================
(Binary files differ)
Property changes on: trunk/community/talks/200808_debconf8/debian-enterprise.png
___________________________________________________________________
Name: svn:mime-type
+ application/octet-stream
Added: trunk/community/talks/200808_debconf8/debian-jr.png
===================================================================
(Binary files differ)
Property changes on: trunk/community/talks/200808_debconf8/debian-jr.png
___________________________________________________________________
Name: svn:mime-type
+ application/octet-stream
Added: trunk/community/talks/200808_debconf8/debian-lex.png
===================================================================
(Binary files differ)
Property changes on: trunk/community/talks/200808_debconf8/debian-lex.png
___________________________________________________________________
Name: svn:mime-type
+ application/octet-stream
Added: trunk/community/talks/200808_debconf8/debian-med.png
===================================================================
(Binary files differ)
Property changes on: trunk/community/talks/200808_debconf8/debian-med.png
___________________________________________________________________
Name: svn:mime-type
+ application/octet-stream
Modified: trunk/community/talks/200808_debconf8/debian-med.tex
===================================================================
--- trunk/community/talks/200808_debconf8/debian-med.tex 2008-08-04 13:49:42 UTC (rev 2387)
+++ trunk/community/talks/200808_debconf8/debian-med.tex 2008-08-04 19:01:30 UTC (rev 2388)
@@ -13,6 +13,8 @@
\usepackage{debian-at}
+\newcommand{\bs}{{$\mathtt\backslash$}}
+
\title{Healthy CDDs}
\subtitle{Strategies for building a Custom Debian Distribution}
@@ -57,53 +59,256 @@
\end{frame}
\begin{frame}
- \frametitle{Groth}
+ \frametitle{Who are the others?}
+ \begin{itemize}
+ \item \DebianJr (2000)
+ \item \DebianMed (2002)
+ \item \DebianEdu (2002)
+ \item [\MyBigRightArrow] DebConf 3 Oslo ``Custom Debian Distributions''
+ \item \Debian Accessibility (2003)
+ \item \Debian Desktop (2003)
+ \item \Debian Lex (2003)
+ \item \Debian Nonprofit (2003)
+ \item \DebianGIS (2004)
+ \item \DebiChem (2004)
+ \item \Debian Enterprise (2007)
+ \end{itemize}
+\end{frame}
+\section{Development}
+
+\subsection{Measures for development}
+
+\begin{frame}
+ \frametitle{Considering some numbers ...}
+
+ \begin{itemize}
+ \item Number of users?
+ \begin{itemize}
+ \item Popularity contest?
+ \item Specific software has usually low popcon numbers
+ \item Usage of metapackages might give some hints but popcon
+ works bad for these - the ``content'' is not actually used
+ \end{itemize}
+ \item Number of developers
+ \begin{itemize}
+ \item Number of Vcs commits $\rightarrow$ not every CDD uses common Vcs
+ \item Number of mailing list subscribers $\rightarrow$ people
+ tend to stay subscribed
+ \end{itemize}
+ \end{itemize}
+\end{frame}
+
+\begin{frame}
+ \begin{tabular}{c@{ }c}
+ Debian Med & Debian Edu \\
+ \resizebox{56mm}{!}{\includegraphics{debian-med}} &
+ \resizebox{56mm}{!}{\includegraphics{debian-edu}} \\[0.95ex]
+ Debian Junior & Debian Lex \\
+ \resizebox{56mm}{!}{\includegraphics{debian-jr}} &
+ \resizebox{56mm}{!}{\includegraphics{debian-lex}}
+\end{tabular}
+\end{frame}
+
+\begin{frame}
+ \frametitle{Signal, noise, artefacts}
+ \begin{itemize}
+ \item Not only SPAM is noise
+ \item (Un)subscribe
+ \item Debian Junior list peak in October 2003:
+ \begin{itemize}
+ \item Replacement for Abiword: LyX? Openoffice?
+ $\rightarrow$ 43 postings
+ \item Philosophy (was Re: Replacement for Abiword: LyX? Openoffice?)
+ $\rightarrow$ 33 postings
+ \item Future replacement for Abiword: Tux Writer?
+ $\rightarrow$ 3 postings
+ \item 79 postings about specific topic out of 89 non-SPAM
+ \end{itemize}
+ \item Intended and unintended robots
+ \begin{itemize}
+ \item December 2005: 92 messages by
+ \texttt{gentoo-\bs{}w+\bs{}+help at gentoo.org} about subscription status to
+ \link{http://lists.debian.org/debian-desktop}{debian-desktop@lists.debian.org}
+ \item Messages by \texttt{Debian Installer},
+ \texttt{bugzilla-skolelinux}, \texttt{Archive
+ Administrator}, \dots
+ \end{itemize}
+ \end{itemize}
+\end{frame}
+
+\subsection{Growth of Metapackages}
+
+\begin{frame}
+ \frametitle{Dependency of metapackages}
+
+ \begin{itemize}
+ \item Technical output - number of metapackages and their dependencies
+ \item Probably good measure
+ \item Hard to obtain because not everything is in Vcs
+ \item Querying \link{http://snapshot.debian.net}{snapshot.debian.net}
+ possible but really time consuming
+ \end{itemize}
+
+\end{frame}
+
+\begin{frame}
+ \frametitle{Selected metapackages of Debian Med}
+
\begin{center}
\resizebox{90mm}{!}{\includegraphics{dmstats}}
\end{center}
\end{frame}
+\subsection{Activity of members}
+
\begin{frame}
-% \frametitle{Groth}
+ \frametitle{Number of subscribers?}
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{debian-enterprise}}
+ \end{center}
+
+\end{frame}
+
+\begin{frame}
+ \frametitle{Activity of communication measure}
+
\begin{itemize}
- \item {\em ``secret society''} ;-)
- \item We think we are everything but secret
- \item At least one feature of secrecy: concealment
- \begin{itemize}
- \item Concealment inside advertising noise of proprietary
- products
- \item Concealment by disunity
- \end{itemize}
+ \item Intensity and quality of communication
+ \item Every CDD has a mailing list
+ \item Who are the active posters (except robots)
+ \item Influence of SPAM, noise, flames, etc. is reduced
+ \item Mind the run-over-by-bus factor
\end{itemize}
\end{frame}
+
+\section{Comparison of CDDs}
+
+\subsection{Graphing activity}
+% Debian Med
\begin{frame}
- \frametitle{Motivation}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-med}{debian-med@lists.debian.org}}
- \begin{enumerate}
- \item Support of target users with common profile:
- \begin{itemize}
- \item Less technical competence
- \item Unable to install upstream programs with acceptable effort
- \item No interest in administration
- \item \emph{Specific subset} of available Free Software
- \item Need for easy usage
- \item Defined security profile
- \item Language barrier
- \end{itemize}
- \item Support of administrators with common profile:
- \begin{itemize}
- \item Limited time frame
- \item Automation of often repeated tasks
- \item Lack of specialist knowledge
- \end{itemize}
- \end{enumerate}
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_med}}
+ \end{center}
+
\end{frame}
+% Debian Edu
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-edu}{debian-edu@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_edu}}
+ \end{center}
+
+\end{frame}
+
+% Debian Jr
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-jr}{debian-jr@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_jr}}
+ \end{center}
+
+\end{frame}
+
+% Debian Accessibility
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-accessibility}{debian-accessibility@lists.d.o}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_accessibility}}
+ \end{center}
+
+\end{frame}
+
+% Debian Desktop
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-desktop}{debian-desktop@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_desktop}}
+ \end{center}
+
+\end{frame}
+
+% Debian Lex
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-lex}{debian-lex@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_lex}}
+ \end{center}
+
+\end{frame}
+
+\begin{frame}
+ \begin{itemize}
+ \item Successfull SPAM protection: in 2008 this list gathered
+ only SPAM
+ \item Peak in November 2007:
+
+\medskip
+\begin{tabular}{lr}
+ Debian-Lex Rebirth? & 23 \\
+ Debian-Lex Rebirth? - python/perl/php? & 25 \\
+ Debian-Lex Rebirth? - wiki & 8 \\
+ Rebirth & 1 \\
+ Rebirth of Debian-lex & 2 \\
+\end{tabular}
+ \item Good example for ``vanished leader''
+ \end{itemize}
+\end{frame}
+
+% Debian Nonprofit
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-nonprofit}{debian-nonprofit@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_nonprofit}}
+ \end{center}
+
+\end{frame}
+
+% Debian Enterprise
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-enterprise}{debian-enterprise@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_enterprise}}
+ \end{center}
+
+\end{frame}
+
+% Debian Custom
+\begin{frame}
+ \frametitle{Top 10 posters on \link{http://lists.debian.org/debian-custom}{debian-custom@lists.debian.org}}
+
+ \begin{center}
+ \resizebox{90mm}{!}{\includegraphics{authorstat_custom}}
+ \end{center}
+
+\end{frame}
+
+\subsection{Conclusion}
+
+\begin{frame}
+ \frametitle{Most CDDs depend from single person}
+ \begin{itemize}
+ \item Debian Edu has a really healthy crew
+ \item Debian Med has at least run-over-by-bus factor two
+ \item Others seem to depend from single person
+ \item Need of better management
+ \end{itemize}
+\end{frame}
+
\input med-end-en.tex
\end{document}
Modified: trunk/community/talks/200808_debconf8/get-archive-pages
===================================================================
--- trunk/community/talks/200808_debconf8/get-archive-pages 2008-08-04 13:49:42 UTC (rev 2387)
+++ trunk/community/talks/200808_debconf8/get-archive-pages 2008-08-04 19:01:30 UTC (rev 2388)
@@ -7,13 +7,18 @@
use DBI;
my $BASEURL = "http://lists.debian.org/debian" ;
-my @PROJECTS = ('med', 'edu', 'jr', 'accessibility', 'desktop', 'enterprise', 'lex', 'nonprofit', 'custom') ;
+my @PROJECTS = ('med', 'edu', 'jr', 'accessibility', 'desktop', 'enterprise', 'lex', 'nonprofit', 'science', 'custom') ;
my @MONTHES = ('01', '02', '03', '04', '05', '06', '07', '08', '09', '10', '11', '12');
my @ROBOTS = ('Debian Installer', 'bugzilla-skolelinux', 'Archive Administrator', 'hostmaster',
'Debian-med-request', 'Debian testing watch', 'Debian Bug Tracking System',
- 'Skolelinux archive Installer', 'Debian Wiki');
-my @SPAMAUTHORS = ('Pls check this new site');
+ 'Skolelinux archive Installer', 'Debian Wiki', 'gentoo-\w+\+help');
+## TODO: just consider mails containing these strings as SPAM
+## This has to be implemented in the code below
+my @SPAMAUTHORS = ('Pls check this new site');
+my @SPAMSUBJECTS = ('File blocked - ScanMail for Lotus Notes',
+ '^u?n?subscribe\s+.?$');
+
# if != 0 then extract of mailing list archives is stored in files in dirs
# The prefered method is to use only the database
my $storefiles = 0;
More information about the debian-med-commit
mailing list