Spend millions visiting space TWICE, take Amazon referal money from non profit... priceless!



From skitterman.wordpress.com/2011/03/02/business-oppor<wbr></wbr><wbr></wbr>tunity-in-ubuntu/#comment-364:
Thanks to a former friend for pointing me to these two people who could express this better than me.

A few days ago Conrad Sanderson released version 1.1.6 of his Armadillo templated C++ library for linear algebra. Accordingly, I made a new release 0.2.14 of RcppArmadillo, our Rcpp-based integration into R which included a few local fixes. However, while I updated SVN at the R-Forge project I didn't update the final tarball so Armadillo 1.1.6 wasn't in it. That is corrected in the 0.2.15 upload which is now on CRAN and its mirrors. The short NEWS file extract for both releases follows, also containing Conrad's entry for 1.1.6. A few minor tweaks were added but no user-visible changes:

0.2.15  2011-03-04
    o   Upgraded to Armadillo Version 1.1.6    Baby Carpet Shark 
         * fixed size matrices and vectors can use auxiliary (external) memory
         * .in_range() can use span() arguments
         * subfields can use span() arguments
0.2.14  2011-03-02
    o   Support Run-Time Type Information (RTTI) on matrices by setting the
        state variable vec_state in Row and Col instantiation, with thanks to
        Conrad Sanderson for the hint
    o   fastLm code simplified further by instantiating the Armadillo matrix
        and vector directly from the SEXP coming from R
    o   inst/doc/Makefile now respects $R_HOME environment variable

And courtesy of CRANberries, here are the diffs to.2.14

 ChangeLog                                                   14 -
 DESCRIPTION                                                  8 
 inst/NEWS                                                   10 
 inst/doc/RcppArmadillo-unitTests.pdf                     binary
 inst/doc/unitTests-results/RcppArmadillo-unitTests.html     10 
 inst/doc/unitTests-results/RcppArmadillo-unitTests.txt      39 +-
 inst/include/armadillo_bits/Col_meat.hpp                    51 +++
 inst/include/armadillo_bits/Col_proto.hpp                    3 
 inst/include/armadillo_bits/Cube_meat.hpp                   52 +++
 inst/include/armadillo_bits/Cube_proto.hpp                   7 
 inst/include/armadillo_bits/Mat_meat.hpp                    51 +++
 inst/include/armadillo_bits/Mat_proto.hpp                   12 
 inst/include/armadillo_bits/Row_meat.hpp                    51 +++
 inst/include/armadillo_bits/Row_proto.hpp                    3 
 inst/include/armadillo_bits/arma_version.hpp                 6 
 inst/include/armadillo_bits/config.hpp                       9 
 inst/include/armadillo_bits/field_meat.hpp                 220 ++++++++++++++--
 inst/include/armadillo_bits/field_proto.hpp                 20 +
 inst/unitTests/runit.RcppArmadillo.R                        16 +
 19 files changed, 521 insertions(+), 61 deletions(-)

and to the previous release 0.2.13, respectivel

 ChangeLog                                                   14 ++++++
 DESCRIPTION                                                  6 +-
 inst/NEWS                                                   11 +++++
 inst/doc/Makefile                                           15 +++++--
 inst/doc/RcppArmadillo-unitTests.pdf                     binary
 inst/doc/unitTests-results/RcppArmadillo-unitTests.html     18 ++++----
 inst/doc/unitTests-results/RcppArmadillo-unitTests.txt      34 ++++++++--------
 inst/include/RcppArmadillo/Col_meat.h                       24 ++++++-----
 inst/include/RcppArmadillo/Row_meat.h                       19 ++++----
 src/fastLm.cpp                                              15 ++-----
 10 files changed, 94 insertions(+), 62 deletions(-)

More information is on the RcppArmadillo page. Questions, comments etc should go to the rcpp-devel mailing list off the R-Forge page.

The test suite for recently added ODBC support in Gammu was consistently failing on Win64 platform (all others were just fine). It took me some time to figure out why. The reason is SQLLEN type, which is in some sort special, that generated code for it is broken. All started with pretty much standard retrieving value from ODBC result:

SQLLEN size;
SQLRETURN ret;
/* Figure out string length */
ret = SQLGetData(res->odbc, 1, SQL_C_CHAR, NULL, 0, &size);
if (!SQL_SUCCEEDED(ret))  
    return NULL;
 
/* Did not we get NULL? */
if (size == SQL_NULL_DATA)  
    return NULL;
 
/* Allocate buffer for string */
out = reallocate(out, size + 1);

The first problem was comparing of size against SQL_NULL_DATA. It simply always failed even if size did contain -1 (what is actually value of SQL_NULL_DATA). Yesterday, I've come out with typecasting magic that worked fine for this case. However what I've missed is that even reallocate is failing with this parameter. No matter what the size would be, it would return NULL. Typecasting again did help here. Hey, should not be compiler clever enough to typecast this for me? It turns out that in case of MinGW64 it is not. Anybody has suggestion why? At the end I ended up with pretty ugly duplicating of the variable just to avoid mixing it up in future changes (as would happen with typecasting everywhere). I'd really like to know why my assumptions are wrong in this case.

Filed under: English Gammu 0 comments Flattr this!

Time for a seventh Debian XSF News issue!

• On the documentation side, I tried to come up with a squeeze backports policy. There are some new upstream releases and xorg-server 1.10 coming up (read below), as well as X stack s migration to testing, to consider before thinking of backporting though.
• Where backports will likely help is for old intel cards (see our FAQ entry). According to Chris Wilson, current kernels/intel drivers are better for 855+, but 830/845 still need a miracle.
• Migrating the X stack to testing needs a bit of coordination, because of the always painful dependency between the nouveau driver and the Linux kernel version. As mentioned in my mail to debian-release@ that would be quite self-contained otherwise. Bottom line: we re waiting for the kernel.
• As announced on debian-x@, the plan is to wait for this migration to happen before uploading new upstream releases of libdrm, the intel driver, and mesa to unstable. Both libdrm and mesa were prepared for Ubuntu, and will be likely uploaded to experimental soon. The intel driver was prepared, but only built against xorg-server 1.10 and uploaded to experimental for now.
• On the xorg-server 1.10 side, the (very late) RandR patches were proven broken, so reverting RandR 1.4 was decided. With some minor tweaks (to the radeon driver), that went quite well, and xorg-server 1.10 got released. Unfortunately, there s still some missing fix for the build system, which explains why I didn t upload a final 1.10 to experimental yet; several patches were proposed already (I checked a bunch of them), but we re still waiting for a final patch.
• As a result of the RandR 1.4 reverts, the video ABI was broken again. Fortunately I gave it a quick try right before 1.10rc3 got released, and noticed the breakage, so the video ABI got bumped. I was therefore very happy with my decision to only upload a bunch of drivers to experimental, built against 1.10rc* (see DXN#6), since I only needed to upload the following video drivers again: ati, dummy, fbdev, intel, mach64, nouveau, r128, vesa. It might make sense to add cirrus to that list next time, for tests under kvm and the like.
• After I poked Peter Hutterer, he kindly and promptly released snapshots for the following input drivers: kbd, mouse, and synaptics, which I uploaded both to unstable (built against 1.9), and to experimental (built against 1.10).
• If you re having troubles with your synaptics device, make sure you have no configuration issues in xorg.conf; even if that used to work somehow, autoconfiguration might fail if there s a misconfigured device (for example, wrong Device option there). You might want to have a look at our how to configure input documentation.
• On the driver side, for 1.9/unstable, Tormod Volden prepared updates for the mga and savage video drivers, which I sponsored.
• Julien Viard de Galbert prepared a new xserver-xorg-video-openchrome snapshot, which I uploaded to unstable.
• He also prepared the addition of a debug package, which I also sponsored after setting the Dm-Upload-Allowed field to yes since he s doing a pretty good job already. It s currently sitting in NEW accordingly.
• A while ago, Thibaut Girka requested the addition of an udeb for the libxtst library, it s now sitting in NEW. Thibaut is working on a virtual on-screen keyboard, initially for use in the Graphical Installer for devices lacking a keyboard (but likely useful in other circumstances). Work is going on, stay tuned.
• Steve Langasek was kind enough to MultiArch-ify both libx11 and xorg; both were uploaded.
• Wayland is getting way more funnier than planned: it is now needed to get mesa built against wayland libraries (to get the wayland EGL platform), and the needed mesa bits are (last I asked) only in master for now.
• On the bugs side, I spent many hours pinging submitters and triaging bugs, mostly on the ati and intel video drivers, as well as on xorg-server again, along with the xorg meta package. (Clicking on a bug graph below leads to a cached full-year view.)

  x-x-video-ati	x-x-video-intel	xorg-server	xorg

• According to the UDD view of XSF bugs, and to Julien s TXBW1, we went down from 880 bugs in november to 540 bugs now.

Why is that Mercedes blocking so much of the road through the station car park?It s because it s the second car in a 1 car parking spot!Here s a front view.These were taken in the Coburg Station car park on Wednesday. The car park was about half empty, so the alternative to blocking part of the road and blocking in someone who was legally parked was to just park about 10 meters further away from the station.Some people just shouldn t be driving.

This little blog is created/compiled with Chronicle, coded by Steve Kemp, and one of the cool features is that you can specify a command run before/after your blog gets compiled from the input files. This presents us with one problem and a lot of opportunities. First to the problem: the singular "a command" (from the manual page) might not be enough. Maybe you want to run more than one command. This can be achieved by creating a directory for the commands you want to run before the compilation and one for the commands you want to run after the chronicle run. Then you create a small shell script one directory above and call that from chronicle, the script for the post-run might look like:

#!/bin/sh
set -e
WORKDIR=" dirname $ 0  /post"
find $ WORKDIR  -name "*.sh" -exec " " ";"

This script makes the assumption, that the scripts to run have a .sh suffix and are in a subdirectory named "post". But you can surely adapt this to your needs. Now to one of the oppurtunities this offers: you can now, for example, create static pages, which share the general visual appearance with your blog entries but don't show up in the tags or archive system. I'm using this to generate an imprint for this blog. The script to generate the imprint looks like:

#!/bin/sh
set -e
OUTPUT_PATH= grep output $ HOME /.chroniclerc   cut -d= -f2   tr -d [:blank:] 
sed -ne'/<\?xml version/,/blosxomFirstDayDiv/  /blosxomFirstDayDiv/b;p ' \
        $ OUTPUT_PATH /index.html > $ HOME /tmp/imprint.html
cat  dirname $ 0  /imprint.data >> $ HOME /tmp/imprint.html
echo -e "\n" >> $ HOME /tmp/imprint.html
sed -ne'/imprint_match/,$p' $ OUTPUT_PATH /index.html >> $ HOME /tmp/imprint.html
mv $ HOME /tmp/imprint.html $ OUTPUT_PATH /imprint.html

There are again some assumptions made here, which you'd need to adopt to your setup:

• I'm using a modified version of the "Copyrighteous" theme (one of the modifications was detailed yesterday).
• The script expects the output setting in your .chroniclerc.
• The content/body of the imprint is stored in a file imprint.data in the same directory as the script which generates the imprint.
• There is a tmp directory below $ HOME (this is there to cope with a chroot environment).

Now you have a very powerful system to modify/extend your chronicle-generated blog. I hope you enjoy it as much as I do! A little P.S.: you might want to use version 4.5, as 4.4 had a nasty encoding bug.

This little blog is created/compiled with Chronicle, coded by Steve Kemp, and one of the cool features is that you can specify a command run before/after your blog gets compiled from the input files. This presents us with one problem and a lot of opportunities. First to the problem: the singular "a command" (from the manual page) might not be enough. Maybe you want to run more than one command. This can be achieved by creating a directory for the commands you want to run before the compilation and one for the commands you want to run after the chronicle run. [UPDATE]Then just add run-parts /path/to/ before,post .d to your .chroniclerc for the pre,post -build configuration entries.[/UPDATE] (Thanks to Steve for pointing the obvious out to me, don't know why I didn't think of run-parts myself.) Now to one of the oppurtunities this offers: you can now, for example, create static pages, which share the general visual appearance with your blog entries but don't show up in the tags or archive system. I'm using this to generate an imprint for this blog. The script to generate the imprint looks like:

#!/bin/sh
set -e
OUTPUT_PATH= grep output $ HOME /.chroniclerc   cut -d= -f2   tr -d [:blank:] 
sed -nre'/<\?xml version/,/blosxomFirstDayDiv/  /blosxomFirstDayDiv/b;p ' \
        $ OUTPUT_PATH /index.html > $ HOME /tmp/imprint.html
cat  dirname $ 0  /imprint.data >> $ HOME /tmp/imprint.html
echo -e "\n" >> $ HOME /tmp/imprint.html
sed -ne'/imprint_match/,$p' $ OUTPUT_PATH /index.html >> $ HOME /tmp/imprint.html
mv $ HOME /tmp/imprint.html $ OUTPUT_PATH /imprint.html

There are again some assumptions made here, which you'd need to adopt to your setup:

• I'm using a modified version of the "Copyrighteous" theme (one of the modifications was detailed yesterday).
• The script expects the output setting in your .chroniclerc.
• The content/body of the imprint is stored in a file imprint.data in the same directory as the script which generates the imprint.
• There is a tmp directory below $ HOME (this is there to cope with a chroot environment).

Now you have a very powerful system to modify/extend your chronicle-generated blog. I hope you enjoy it as much as I do! A little P.S.: you might want to use version 4.5, as 4.4 had a nasty encoding bug.

Christian is a figure of Debian, not only because of the tremendous coordination work that he does within the translation project, but also because he s very involved at the social level. He s probably in the top 5 of the persons who attended most often the Debian conference. Christian is a friend (thanks for hosting me so many times when I come to Paris for Debian related events) and I m glad that he accepted to be interviewed. He likes to speak and that shows in the length of his answers but you ll be traveling the world while reading him. My questions are in bold, the rest is by Christian. Who are you? I am a French citizen (which is easy to guess unless you correct my usual mistakes in what follows). I m immensely proud of being married for nearly 26 years with Elizabeth (who deserves a statue from Debian for being so patient with my passion and my dedication to the project). I m also the proud father of 3 wonderful kids , aged 19 to 23. I work as team manager in the Networks and Computers Division of Onera the French Aerospace lab , a public research institute about Aeronautics, Space and Defense. My team provides computer management services for research divisions of Onera, with a specific focus put on individual computing. I entered the world of free software as one of the very first users of Linux in France. Back in the early 1990 s, I happened (though the BBS users communities) to be a friend of several early adopters of Linux and/or BSD386/FreeBSD/NetBSD in France. More specifically, I discovered Linux thanks with my friend Ren Cougnenc (all my free software talks are dedicated to Ren , who passed away in 1996). You re not a programmer, not even a packager. How did you come to Debian? I m definitely not a programmer and I never studied computing (I graduated in Materials Science and worked in that area for a few years after my PhD). However, my daily work always involved computing (I redesigned the creep testing laboratory and its acquisition system all by myself during my thesis research work). An my hobbies often involved playing with home computers, always trying to learn about something new. So, first learning about a new operating system then trying to figure out how to become involved in its development was quite a logical choice. Debian is my distro of choice since it exists. I used Slackware on work machines for a while, but my home server, kheops, first ran Debian 1.1 when I stopped running a BBS on an MS-DOS machine to host a news server. That was back in October 1996. I then happened to be a user, and more specifically a user of genealogy software, also participating very actively in Usenet from this home computer and server, that was running this Debian thing. So, progressively, I joined mailing lists and, being a passionate person, I tried to figure out how I could bring my own little contribution to all this. This is why I became a packager (yes, I am one!) by taking over the geneweb package, which I was using to publish my genealogy research. I applied as DD in January 2001, then got my account in July 2001. My first upload to the Debian archive occurred on August 22nd 2001: that was of course geneweb, which I still maintain. Quite quickly, I became involved in the work on French localization. I have always been a strong supporter of localized software (I even translated a few BBS software back in the early 90 s) as one of the way to bring the power and richness of free software to more users. Localization work lead me to work on the early version of Debian Installer, during those 2003-2005 years where the development of D-I was an incredibly motivating and challenging task, lead by Joey Hess and his inspiring ideas. From user to contributor to leader, I suddenly discovered, around 2004, that I became the coordinator of D-I i18n (internationalization) without even noticing You re the main translation coordinator in Debian. What plans and goals have you set for Debian Wheezy? As always: paint the world in red. Indeed, this is my goal for years. I would like our favorite distro to be able to be used by anyone in the world, whether she speaks English, Northern Sami, Wolof, Uyghur or Secwepemcts n. As a matter of symbol, I use the installer for this. My stance is that one should be able to even install Debian in one s own language. So, for about 7 years, I use D-I as a way to attract new localization contributors. This progress is represented on this page where the world is gradually painted in red as long as the installer supports more languages release after release. The map above tries to illustrate this by painting in red countries when the most spoken language in the country is supported in Debian Installer. However, that map does not give enough reward to many great efforts made to support very different kind of languages. Not only various national languages, but also very different ones: all regional languages of Spain, many of the most spoken languages in India, minority languages such as Uyghur for which an effort is starting, Northern Sami because it is taught in a few schools in Norway, etc., etc. Still, the map gives a good idea of what I would like to see better supported: languages from Africa, several languages in Central Asia. And, as a very very personal goal, I m eagerly waiting for support of Tibetan in Debian Installer, the same way we support its sister language, Dzongkha from Bhutan. For this to happen, we have to make contribution to localization as easy as possible. The very distributed nature of Debian development makes this a challenge, as material to translate (D-I components, debconf screens, native packages, packages descriptions, website, documentation) is very widely spread. A goal, for years, is to set a centralized place where translators could work easily without even knowing about SVN/GIT/BZR or having to report bugs to send their work. The point, however, would be to have this without making compromises on translation quality. So, with peer review, use of thesaurus and translation memory and all such techniques. Tools for this exist: we, for instance, worked with the developers of Pootle to help making it able to cope with the huge amount of material in Debian (think about packages descriptions translations). However, as of now, the glue between such tools and the raw material (that often lies in packages) didn t come. So, currently, translation work in Debian requires a great knowledge of how things are organized, where is the material, how it can be possible to make contribution reach packages, etc. And, as I m technically unable to fulfill the goal of building the infrastructure, I m fulfilling that role of spreading out the knowledge. This is how I can define my coordinator role. Ubuntu uses a web-based tool to make it easy to contribute translations directly in Launchpad. At some point you asked Canonical to make it free software. Launchpad has been freed in the mean time. Have you (re)considered using it? Why not? After all, it more or less fills in the needs I just described. I still don t really figure out how we could have all Debian material gathered in Rosetta/Launchpad .and also how Debian packagers could easily get localized material back from the framework without changing their development processes. I have always tried to stay neutral wrt Ubuntu. As many people now in Debian, I feel like we have reached a good way to achieve our mutual development. When it comes at localization work, the early days where the everything in Rosetta and translates who wants stanza did a lot of harm to several upstream localization projects is, I think, way over. Many people who currently contribute to D-I localization were indeed sent to me by Ubuntu contributors .and by localizing D-I, apt, debconf, package descriptions, etc., they re doing translation work for Ubuntu as well as for Debian. Let s say I m a Debian user and I want to help translate Debian in my language. I can spend 1 hour per week on this activity. What should I do to start? Several language teams use Debian mailing lists to coordinate their work. If you re lucky enough to be a speaker of one of these languages, try joining debian-l10n-<yourlanguage> and follow what s happening there. Don t try to immediately jump in some translation work. First, participate to peer reviews: comment on others translations. Learn about the team s processes, jargon and habits. Then, progressively, start working on a few translations: you may want to start with translations of debconf templates: they are short, often easy to do. That s perfect if you have few time. If no language team exists for your language, try joining debian-i18n and ask about existing effort for your language. I may be able to point you to individuals working on Debian translations (very often along with other free software translation efforts). If I am not, then you have just been named coordinator for your language I may even ask you if you want to work on translating the Debian Installer. What s the biggest problem of Debian? We have no problems, we only have solutions We are maybe facing a growth problem for a few years. Despite the increased welcoming aspects of our processes (Debian Maintainers), Debian is having hard times in growing. The overall number of active contributors is probably stagnating for quite a while. I m still amazed, however, to see how we can cope with that and still be able to release over the years. So, after all, this is maybe not a problem Many people would point communication problems here. I don t. I think that communication inside the Debian project is working fairly well now. Our famous flame wars do of course still happen from time to time, but what large free software project doesn t have flame wars? In many areas, we indeed improved communication very significantly. I want to take as an example the way the release of squeeze has been managed. I think that the release team did, even more this time, a very significant and visible effort to communicate with the entire project. And the release of squeeze has been a great success in that matter. So, there s nearly nothing that frustrates me in Debian. Even when a random developer breaks my beloved 100% completeness of French translations, I m not frustrated for more than 2 minutes. You re known in the Debian community as the organizer of the Cheese & Wine Party during DebConf. Can you tell us what this is about? This is an interesting story about how things build themselves in Debian. It all started in July 2005, before DebConf 5 in Helsinki. Denis Barbier, Nicolas Fran ois and myself agreed to bring at Debconf a few pieces of French cheese as well as 1 or 2 bottles of French wine and share them with some friends. Thus, we settled an informal meeting in the French room where we invited some fellows: from memory, Benjamin Mako Hill, Hannah Wallach, Matt Zimmermann and Moray Allan. All of us fond of smelly cheese, great wine plus some extra p t home-made by Denis in Toulouse. It finally happened that, by word of mouth, a few dozens of other people slowly joined in that French room and turned the whole thing into an improvized party that more or less lasted for the entire night. The tradition was later firmly settled in 2006, first in Debconf 6 in Mexico where I challenged the French DDs to bring as many great cheese as possible, then during the Debian i18n meeting in Extremadura (Sept 2006) where we reached the highest amount of cheese per participant ever. I think that the Creofonte building in Casar de C ceres hasn t fully recovered from it and is still smelling cheese 5 years after. This party later became a real tradition for DebConf, growing over and over each year. I see it as a wonderful way to illustrate the diversity we have in Debian, as well as the mutual enrichment we always felt during DebConfs. My only regret about it is that it became so big over the years that organizing it is always a challenge and I more and more feel pressure to make it successful. However, over the years, I always found incredible help by DebConf participants (including my own son, last year a moment of sharing which we will both remember for years, i think). And, really, in 2010, standing up on a chair, shouting (because the microphone wasn t working) to thank everybody, was the most emotional moment I had at Debconf 10. Is there someone in Debian that you admire for their contributions? So many people. So, just like it happens in many awards ceremonies, I will be very verbose to thank people, sorry in advance for this. The name that comes first is Joey Hess. Joey is someone who has a unique way to perceive what improvements are good for Debian and a very precise and meticulous way to design these improvements. Think about debconf. It is designed for so long now and still reaching its very specific goal. So well designed that it is the entire basis for Joey s other achievement: designing D-I. Moreover, I not only admire Joey for his technical work, but also for his interaction with others. He is not he loudest person around, he doesn t have to .just giving his point in discussion and, guess what? Most of the time, he s right. Someone I would like to name here, also, is Colin Watson. Colin is also someone I worked with for years (the D-I effect, again ) and, here again, the very clever way he works on technical improvements as well as his very friendly way to interact with others just make it. And, how about you, Rapha l? I m really admirative of the way you work on promoting technical work on Debian. Your natural ability to explain things (as good in English as it is in French) and your motivation to share your knowledge are a great benefit for the project. Not to mention the technical achievements you made with Guillem on dpkg of course! Another person I d like to name here is Steve Langasek. We both maintain samba packages for years and collaboration with him has always been a pleasure. Just like Colin, Steve is IMHO a model to follow when it comes at people who work for Canonical while continuing their involvment in Debian. And, indeed, Steve is so patient with my mistakes and stupid questions in samba packaging that he deserves a statue. We re now reaching the end of the year where Stefano Zacchiroli was the Debian Project Leader. And, no offense intended to people who were DPL before him (all of them being people I consider to be friends of mine), I think he did the best term ever. Zack is wonderful in sharing his enthusiasm about Debian and has a unique way to do it. Up to the very end of his term, he has always been working on various aspects of the project and my only hope is that he ll run again (however, I would very well understand that he wants to go back to his hacking activities!). Hat off, Zack!I again have several other people to name in this Bubulle hall of Fame : Don Armstrong, for his constant work on improving Debian BTS, Margarita Manterola as one of the best successes of Debian Women (and the most geeky honeymoon ever), Denis Barbier and Nicolas Fran ois because i18n need really skilled people, Cyril Brulebois and Julien Cristau who kept X.org packaging alive in lenny and squeeze, Otavio Salvador who never gave up on D-I even when we were so few to care about it. I would like to make a special mention for Frans Pop. His loss in 2010 has been a shock for many of us, and particularly me. Frans and I had a similar history in Debian, both mostly working on so-called non technical duties. Frans has been the best release manager for D-I (no offense intended, at all, to Joey or Otavio .I know that both of them share this feeling with me). His very high involvment in his work and the very meticulous way he was doing it lead to great achievements in the installer. The Installation Guide work was also a model and indeed a great example of non technical work that requires as many skills as more classical technical work. So, and even though he was sometimes so picky and, I have to admit, annoying, that explains why I m still feeling sad and, in some way, guilty about Frans loss. One of my goals for wheezy is indeed to complete some things Frans left unachieved. I just found one in bug #564441: I will make this work reach the archive, benefit our users and I know that Frans would have liked that.

---

Thank you to Christian for the time spent answering my questions. I hope you enjoyed reading his answers as I did. Subscribe to my newsletter to get my monthly summary of the Debian/Ubuntu news and to not miss further interviews. You can also follow along on Identi.ca, Twitter and Facebook.

7 comments Liked this article? Click here. My blog is Flattr-enabled.

It looks like many peoples decided it is right time to bloat my system with stuff I don't use. The first think I did not like when squeeze went out was libpcsclite1 pulling pcsd and hal into my system. This is being solved on two fronts - porting pcsd to udev (in #612971) and disabling libpcslite1 support in wpasupplicant. Both of them being good ideas. What actually drove me to writing this post is today's update, where aptitude suddenly wanted to install whole latex and thus using about 760MiB of extra disk space. I know disk space is supposed to be cheap these days, but it's not that cheap with SSD in notebook. The culprit turned out to be doxygen, which up to now happily lived without single bit of latex, but now needs it (though temporarily according to changelog). I know it is tempting to enable every possible feature in your package, but in same cases it is really not something what most of people would appreciate.

Filed under: Debian English 4 comments Flattr this!

The next version of jffnms is nearing completion and is now at Release Candidate 1. Version 0.9.0 has a major amount of work in cleaning up and securing the code. The majority of the work has been in the complete re-write of the engines that do the polling, autodiscovery and consolidation. The parent/child communication has changed as has the way the processes are forked. On the front-end, the requirement to register globals has finally been removed, with the code explicitly specifying and sanitising the variables it requires. This will make it easier to debug problems and make the application webservers more secure. Finally there is better support for High Capacity interface counters and some support for IPv6, meaning you can see how slow ipv6.google.com is from your place. JFFNMS 0.9.0rc1 is available from SourceForge at https://sourceforge.net/projects/jffnms/files/jffnms%20RC/

I was interested to hear about a company here in the UK called ALLOW Ltd., offering marketing database management under a we ll get you off lists, then pay you to go back on at your pleasure basis. That sounds a fair deal to me, so I decided to sign up for it.

Our technology is built using some of the best and most secure tools in the industry. We have partnered with infrastructure providers who handle some of the most sensitive data in the UK (such as medical and financial records). Both the digital and physical security measures we have implemented are amongst the strongest available anywhere. This includes full encryption of all data at all times, full implementation of secure socket layers, security certificates and physical restriction of access to the data, our servers and our offices. Our systems have been fully penetration tested (that means we ve asked people to try and break in).

(There are other suitable assertions in various places they even have a set of principles about safeguarding data.) Unfortunately, this promise is rather undermined in several ways after noticing the first couple, I did a little digging to see what else was exploitable. Here s the final part of the joining process, where you choose a username and password combination: The text I ve cropped too eagerly says Choosing a secure password is an essential part of protecting your personal information , or thereabouts. I duly chose a complex password that fitted the requirements, and to my surprise it was rejected. I tried another, and it was rejected; then a third and a fourth. By trial and error I worked out what was going on: 1. The password must contain only the listed special characters, not just include one of them. That s a bit of a problem, because even assuming a basic ASCII set, 15 characters are unavailable to users; 80 are left, so that s about a 15% fall in the available combinations*. Not a good start. More concerning is the presence of a security question field. It s used for resetting the password in the event losing it, but this technique for recovery has long been ridiculed the shared secret is often common knowledge amongst friends, and sometimes (as in this case) the available questions are fairly easy for an attacker to find the answer to in public records or solicit from the victim without arousing much suspicion: 2. The security questions available include First pet s name and worse, First school name . It s pretty pointless enforcing stringent password requirements, and then bypassing them with something so susceptible to a dictionary attack. I was pleased to find that failing to log in to an account more than a few times results in a temporary lockout, which should deter casual brute force attacks. But I wanted to know how that security question would be used, so I forgot my password and followed the links to reset it. Here s the form: Actually the first form, not shown here, initially just asks for a username, giving an error message if it isn t registered, and here s another problem: 3. The password reset process confirms the existence, or non-existence, of a given username half the credentials required to log in to any visitor. I d be prepared to take a bet that most users will choose What was the name of your first school? as a security question. The first pet you have is often at such a young age you can t remember it clearly; the name of the street you grew up on might change a couple of times if you moved house. But first school I attended? I ll never forget that, so it makes most sense to use as a backup password . It s also the best one for an attacker to try and find out from public sources. But that aside, as you can see the password is not generated at random and communicated to the real account holder out-of-band, in the manner of many other sites. Instead: 4. A new password is immediately set to a value already known by the attacker. Once inside, an attacker can also change the security question or answer, or both, so you can t even regain your account by telephoning the company unless you can convince them you re genuine, in which case the security question was a total waste of time anyway. I awarded some marks for notifying the user by email that the password has been changed, but immediately docked them again because bingo! You re now a victim of identity theft! Let s assume you ve been locked out, the security question has been changed and you want your account back. ALLOW don t let you telephone them; you either have to dig around and find an address to send an email, which we all know can be intercepted, or (and you re encouraged to) contact them through a form on the site. You ll probably include some personal details, because you want to convince them of your real identity; indeed, two of the options on the form are I ve got a question and Something doesn t work . I sent my findings through this very form, under the latter heading, and to my surprise: 5. Despite promises of full encryption of all data at all times, full implementation of secure socket layers , the contact form is transmitted to ALLOW in the clear, with no protection whatsoever. So now anyone listening in your connection knows all about you too: your ISP, any of the peers along the route, the deep packet inspection advertisers if your ISP is less than reputable, and the neighbour who connects to your wireless and slips you a fiver every month for the privilege. Nice work, privacy specialists. (For the record:

• I have sent these findings to ALLOW and await a response have a response.
• Yes, this scenario is unlikely, but it doesn t fill me with confidence about the rest of their business activities.)

* please feel free to correct my maths. It was never my strongest subject.

---

Privacy specialists should hire security specialists is a post from: jwiltshire.org.uk Flattr

As mentioned earlier this week, I changed the server for my hoster from server4you to Hetzner. Because the contract for my old server is still running until the 14th of April, I thought I can make some good use of it by running an Tor exit node on it. Tor, The onion router is a world-wide network of server to anonymize network connections, and used for example by freedom-loving people in repressive countries, or by Facebook users in repressively proxied university networks Naturally, such a system is also used by people with lesser motives, and after nearly one week of running the server, some complaints were reaching server4you a honeynet supposedly detected botnet activity from my IP address, another company hired by Paramount Pictures claimed that someone downloaded Iron Man 2 over BitTorrent via my server. These were forwarded to my by server4you, demanding an explanation. I explained that I am running a Tor node, that such things are unavailable, and that neither I nor server4you is going to be held responsible. Nevertheless, they insist that I should stop whatever I am doing (I am not sure if they really understood what tor is), and when I said that I won t, they decided to unilaterally and immediately end my contract by tomorrow. Thus a powerful exit node that might have made life better for Tor users is doing to disappear again. They will not reimburse the advance fee that I had paid in advance and made a non-specific threat about further claims they might have. So server4you was, at least in my case, not only an unreliable and inflexible provider, they also prefer to cut off their customer s server just because of two complaints by third parties. I conclude that this is not a hoster that you should trust important services to. And I am not a unique case: There are reports of similar server shut-downs by server4you (in German).

I used to maintain a membership with IEEE. I no longer do. The latest nonsense to come from them is a change for the worse in their copyright assignment policy, which i thought was already problematic. Other engineering societies (IETF, USENIX, etc) continue to do socially relevant, useful work without attempting to control copyright of work contributed to them. This just makes IEEE look like a power- and money-hungry organization, rather than a force for positive advancement of technology. For shame, IEEE. I will not consider reinstating my membership unless the organization changes their copyright assignment and publication policy to better reflect the spirit of scientific inquiry and technological advancement they should stand for.Tags: freedom, ieee, policy

As Squeeze was just released, some might be on the lookout for a replacement for LiquidWeather++, a SuperKaramba script. Others might just search for the first time for a way to have a weather forecast displayed on their desktop. With my yaWP maintainer hat on, I'd like to recommend yaWP for the job. Why yaWP? Short answer: because it's the best (SCNR). Longer answer: because yaWP is easy to use, yet highly customizable. yaWP can track the forecasts of several cities for you, you can put yaWP on your desktop or in your control bar. You can limit the amount of days displayed (in the control bar mode). It's localized for many languages (if your language isn't among the already existing ones, please consider translating yaWP). yaWP can work with multiple services, is themeable and can display a satellite image for your area. Now, some of you might still (don't ask me why) prefer a different Plasmoid for displaying the weather forecasts. But even those can benefit from yaWP and use one of its three data engines (AccuWeather, Google Weather Service and Weather Underground (Wunderground)), thanks to KDE's Weather Ions. Ok, and how does it look? Please remember, that yaWP can be themed, so the picture below just shows one option of many.
Included from screenshots.debian.org (License: GPL2 or later + OpenSSL exception)

Another month and thus another issue of Flattr FOSS. Time flies but Flattr s usage seems to remain relatively strong. I was fearing that people get bored after some time but that does not seem be to the case, my Flattr income does not really drop (it doesn t take off either ). I get a bit less for dpkg than at the start but a bit more for my blog. Enough babbling, let s go straight to the 5 projects to flattr this month:

1. Linux Mint (Flattr link) is a Linux distribution based on Debian and Ubuntu that targets home desktop users. I have never tried it (I m too much of an hardcore Debianer) but I ve read several good reviews saying that it works well out of the box with good multimedia support. If you try it out, be sure to pick Linux Mint Debian Edition (LMDE).
2. Network Block Device (NBD) user-space support tools (Flattr link) is the set of tools to setup a block device whose content is really on a remote server. It requires support of NBD in the kernel of course. The user space tools are maintained by Debian developer Wouter Verhelst.
3. Tahoe LAFS (Flattr link) is a filesystem that implements a cloud storage. The data are distributed across multiple servers in such a way that it can continue to work even if one of the servers fails.
4. Gajim (Flattr link) is a full featured and easy to use Jabber client. If you don t need support of multiple instant messaging protocols, this Jabber client might be a good fit for you.
5. Aur lien Gateau (Flattr link) is a KDE developer who takes one day per week off his work to work on KDE. He seeks support from the community to compensate for the loss of revenue. Much like me, he has a support page with a complete history of the amount of donations received. I can only sympathize with people who try to live true to their passion good luck Aur lien!

This article is part of the Flattr FOSS project.

5 comments Liked this article? Click here. My blog is Flattr-enabled.

I picked up a couple of cheap Linux devices at the weekend. First of all, a $99 Android tablet from CVS, made by Craig. It's a generic RK2818 device and of course it's lacking any kind of GPL offer in the documentation. As far as I know the only company that's released any Rockchip source so far has been Archos, and even then they haven't released the tools you need to actually build an image - they seem to be floating around the internet anyway. But it's straightforward to get it to run the Android market, and it runs Shortyz quite well, so fit for purpose from my point of view. I am, obviously, attempting to contact Craig to find out how they're going to satisfy their obligations but haven't got past their bizarre text-to-speech based support menu system that dumps you to answerphone after 5 minutes of being on hold. Next attempt will involve pressing more buttons.

The other one was a Sharper Image Literati e-reader, $49 from Macy's (on clearance, obviously). This one's interesting by virtue of not being an Android device. Instead it's got a fairly recognisable Busybox-based Linux environment that's even got udev and dbus running. It brings up a framebuffer and just dumps a QTE-based reader (from Kobo) onto it. Other than being woefully underpowered and slow, it actually seems very competent. There seem to be several versions of the hardware - the one I got has an ARM SoC from SiRF on it. SiRF make GPS chipsets, and it turns out that their Atlas 5 platform is actually intended for Linux-based GPS units. The embedded world always seems to find a way. What surprised me more is that it's probably the most polished looking Linux I've bought for under $300. No bizarre kernel spew. echo mem >/sys/power/state works. Standard backlight interface.

Oh, and no source. Obviously. But an interesting device regardless.

Australia s Department of Finance has mandated that the MS-Office document format should be the standard document format for all agencies [1]. Paul Wayper notes the fact that MS doesn t plan to support it s own standards and suggests ODF, also known as ISO/IEC standard 26300:2006. [2].Psychology Today has an interesting blog post by Marnia Robinson about scientific reasons for avoiding porn addiction [3]. Her post links to a number of other posts on the topic on Psychology Today.Psychology Today has an interesting post by Noam Shpancer titled What Doesn t Kill You Makes You Weaker [4]. It seems to me that the belief to the contrary is due to people rationalising bad things that happened to them, it s not all bad as it made me stronger .Psychology Today has an interesting article by Noam Shpancer about the soldier as hero myth [5]. He argues that soldiers are really just employees, the risk to American soldiers in war zones really isn t that great and that they are just pawns to the military-industrial complex.Anders Ynnerman gave an interesting TED talk about visualising medical data which includes many startling 3D pictures of humans and animals that are being scanned [6]. He describes how this is useful in forensic analysis by the police and in better understanding biological processes.Margarita Manterola gave an interesting Debconf talk titled Making Debian Rule, again , she has some interesting ideas for improving Debian [7]. She is most focussed on social problems and the questions and comments had some interesting ideas. The Center for Non-Violent Communication [8] was mentioned in the questions, apparently their success includes teaching better communication skills to violent criminals, so it seems that they can help with some big problems.Benjamin Mako Hill wrote an interesting article about the ratio of writers to readers on Wikipedia [9]. While he makes some good points I m wondering about the measure of an editor , there have been few months where I have made 5 edits, so I guess I m just a reader. I probably should take more care in logging in when I make changes though, I ve made more than a few small changes from random systems (client sites, Internet cafes, etc) without bothering to login. This would mess up the stats a bit.The Reid Report has an interesting article about Bush being limited in his international travel due to the fact that he has admitted being a war criminal [10].Understanding Prejudice is an interesting article that summarises a lot of psychological research [11]. It gives many ideas for anyone who wants to reduce prejudice.Robin Harris wrote an interesting ZD Net article about vibrations and the impact on disk performance [12]. Apparently it s very common to be able to get more performance from drive arrays if you dampen vibration, previously I believed that unless you had extreme vibration to cause catastrophic problems then nothing would go wrong. But it seems that even smaller amounts of vibration can cause a 30% performance loss or worse! There is work in progress on vibration-damping racks.In a story published in The Australian about the NSW ambulance service being shut down due to a virus, Professor Caelli cites SE Linux as an example of a technology to make computers more virus resistant [13]. Professor Caelli also states that there should be a legal requirement to have backup systems for computers that have such grave consequences if they crash.Sinclair Community College is offering a new class Fundamentals of Linux Security , it includes SE Linux [14].Slate has an interesting article about Pickpockets [15]. Apparently they are becoming extinct in the US as law enforcement has prevented the Fagins from training younger people, and it s only in Eastern Europe that the skills base remains.

• [1] http://www.itnews.com.au/News/245276,australia-mandates-microsofts-open-office-xml.aspx
• [2] http://www.mabula.net/tbfw/blosxom.cgi/2011/01/19#2011-01-19-minister-leigh-letter
• [3] http://www.psychologytoday.com/blog/cupids-poisoned-arrow/201007/not-all-warnings-about-porn-are-moralistic-or-unscientific
• [4] http://www.psychologytoday.com/blog/insight-therapy/201008/what-doesnt-kill-you-makes-you-weaker
• [5] http://www.psychologytoday.com/blog/insight-therapy/201101/soldier-hero
• [6] http://www.ted.com/talks/anders_ynnerman_visualizing_the_medical_data_explosion.html
• [7] http://penta.debconf.org/dc10_schedule/events/627.en.html
• [8] http://www.cnvc.org/
• [9] http://mako.cc/copyrighteous/20110206-00
• [10] http://blog.reidreport.com/2011/02/bush-cancels-geneva-and-other-consequences-of-torture/
• [11] http://www.understandingprejudice.org/apa/english/
• [12] http://www.zdnet.com/blog/storage/bad-bad-bad-vibrations/896
• [13] http://www.theaustralian.com.au/australian-it/world-focus-on-ambulance-system-failure/story-e6frgakx-1226005951185
• [14] http://www.sinclairclarion.com/home/campus/2011/02/07/new-linux-class-starting-in-spring-quarter/
• [15] http://www.slate.com/id/2286010/pagenum/all/

I was going to make this a comment against Ritesh Raj Sarraf's post about patents and pharmaceuticals but the comment got way too long.

The basic problem is that patents on pharmaceuticals cannot be directly compared to patents on software and the explanation gets a bit involved. It's worse than apples and oranges, it's more like comparing an algorithm to a piece of DNA.

Patented pharmaceuticals must declare the precise chemical formula of the molecule (and a whole lot else besides) before the molecule can be brought to market but the company holding the patent is legally prevented from protecting themselves against slightly different molecules which are blatantly based on the protected molecule. All modification results in a new patentable product, legally separated from the original.

Therefore, the first company to patent a molecule from a new class is instantly joined in the field by other companies who are (obviously) watching the research news channels very carefully. (In their shoes, you would do the same or you go bust.) These companies simply add a flourine ion or change a hydrogen to a hydroxl group or any other number of minor, apparently inconsequential, changes. Changes which, in software, would be deemed a derivative work and would be roughly equivalent to changing a single local variable from a 16bit integer to a 32bit integer in a single function in a single file in the entire codebase. i.e. a trivial bug fix or even a typo.

A molecule with a couple of hundred atoms can have a change affecting one atom and be a completely new patentable item without any risk of the other company being sued for patent infringement, plagiarism, copyright violation (there is no copyright) or any other barrier. It's common to find that molecule foo is joined on the market by a desoxyfoo and a flourofoo etc. with a chlorofoo just around the corner, within months of the class-leader being launched. It can't be compared to software - effectively, every new release gets a new patent which covers the entirety of everything contained in that release, even if 99.999999999% of it is precisely the same as is covered under other several other patents held by extremely litigious competitors.

Modification is simply not prohibited - copying without making any changes whatsoever to the active molecule is prohibited. Changes to the inactive ingredients will be prosecuted - changes to the active ingredient will not. Comparing this with software is likely to drive you insane.

For example, the mere process of packaging an upstream release for a distribution would be patent infringement in pharmaceutical terms - UNLESS you deliberately patch the original source to make an utterly trivial and apparently pointless change. At that point, the upstream completely disown your entire release and you would have to fix every bug yourself - without the original upstream even telling you whether they fixed the same bug themselves, because they aren't ABLE to modify the upstream code without starting a whole new project and they won't tell you why or how they changed things. There is no such thing as an alpha-release of a molecule. Every release is final and no subsequent modification is possible - unless it only affects the inactive ingredients, e.g. by adding a slow release coating. These changes can affect the way that the molecule acts on the test subject / patient but must not change the active ingredient itself in any identifiable way. So the upstream can make a new wrapper (new packaging) but nobody else is allowed to do so until the patent expires. Once the patent expires, every one and their dog is allowed to make identical copies of the active ingredient in their own inactive packaging and these are called generics. Generics still need testing to ensure that the active ingredient truly is the same and that the generic behaves in the same way as the original but this process is cheaper than making a change to the active molecule.

A changed active molecule is an entirely new substance, it gets a new patent and it needs to go through largely the same amount of testing as the patented original but the new company still gets a head start on what the molecule is likely to do and what side-effects (by now noticeable with the class-leader product which is in large scale trials or even on the market) would be harmful to the future usefulness of their new, derivative, molecule. The new company simply jumps onto the bandwagon shouting "I want a piece of that action!" and there is absolutely nothing the original company can do about it, except try to show that their original molecule is better than the new competitors and to try and build a substantial market share before everyone switches to the new derivative. (Prescribers are human and tend therefore to be lazy, so unless there's a reason to change, the class leader will still have substantial share even after several derivatives exist.)

It's an arms race - each company tries to create as many of these potential "me-too" variants as possible in-house, put as many as can be (financially) supported through trials, keep pursuing those which don't have prohibitive side-effects / problems in tests and hope that, at the end, one of their versions will actually turn out to have better effectiveness or fewer side-effects. If not, they just have to market one as "as-effective-and-no-more-harmful-but-CHEAPER-than" the class leader. The majority of "me-too" molecules come under this category - even if the company concerned would try to point to infinitesimal differences in trial data to pretend otherwise. All this with the knowledge that the very hour that the patent expires, a dozen generic manufacturers will flood the market with copies which do not need to recover the massive costs of the original research and are therefore massively cheaper.

How much cheaper? patented molecules may cost 30 for 28 doses on Monday and one generic may cost 3p for 500, another may cost 40p for 10,000 on Tuesday, the day after the patent expiry. Generic houses only need to cover the actual costs of manufacture, not the costs of the original patent etc. and by the time the patent expires, the efforts of the patent holder to generate market share mean that economies of scale become evident. The absolute cheapest isn't necessarily the one who gets the contract either, these things all have expiry dates and all 10,000 have to be sold / supplied from a single location within that date to actually deliver the full savings. So then you get satellite generics who buy in quantities of a few million and pack down into 56 or less so that the total purchased quantity can sit on many shelves instead of just one. It's the same tablet in the end, with the same markings too. All that differs is the marking on the foil and the colour of the cardboard. (Oh and the same foil and colour cardboard can contain tablets with different markings from different manufacturers from one batch to the next, all within the branding of the same repackaging company.)

The original patent-holding company also has an interest in generating their own "me-too" derivatives, sometimes marketed under the name of a company which turns out to be solely owned by the parent of the original company. It can get horribly recursive.
Sometimes, the derivative from the sibling company hits the market before the "true" competitor from a different company - simply because all companies try to have many molecules in development at any one time (because most changed molecules fail in testing) and sometimes they get lucky and come up with two v.similar molecules and have to do something to justify the cost of developing both. That can turn out to be a poisoned chalice rather than two golden geese.

Sometimes, the company fails to get enough variants through testing and the company then gets bought out by a rival - who then gains their research and continues experimenting with other derivative molecules based on the ones already tested and dismissed. The limiting factor is the money.

At this point, I must now declare my cards. I used to be a pharmacist and although I am still entitled to call myself the holder of a Batchelor of Sciences degree in Pharmaceutical Sciences I am most definitely no longer a pharmacist. I therefore defer to my still-registered colleagues in case I've messed up any of the detail here. The above cannot be declared as the opinion of a pharmacist. I am an ex-pharmacist who voluntarily chose to resign his registration to take up full time work as a software developer rather than pay the exorbitant costs of maintaining such registration and all the "proof-of-competency" requirements which such registration requires.

Finally, before you ask, I am very happy to be an ex-pharmacist - my erstwhile colleagues have my ongoing sympathy but no matter how much some of them may beg, they all know that I will not be rejoining them on that side of the dispensary bench if I've got any choice in the matter.

Software is just better. It's all just recycled electrons.

Mosf of us would be well versed with the Patent system in general. We have patents in every sector - Technology, Agriculture, Pharmacy etc. Most readers in our profession must already be well versed of the Pros/Cons of the Patent system in Software/Technology. Patents in software are more like ammunition. The more you have, the stronger you are. We don't seem much cat fight in Software/Technology because not one organization own all the patents for a product. An IBM ThinkPad might be using a cool track pad feature which might be a patent of Dell. Since the finished product comprises of many patents from different owners in the same industry, it is better to play good with each other. You scratch my back, I scratch yours. But what about the Pharmaceutical Industry? There, one single patent could comprise the product. I was reading this document from Bruce Lehman which touches upon Patents and Pharmaceutical Industries, but couldn't find an answer to my question. read more

We have customers, which require a report about what we've done during maintenance windows. Usually this includes a report about upgrades, newly installed packages etc. and obviously everything we've done apart from that.
Till now we've prepared them manually. For a greater bunch of systems this is a big PITA, because to be somehow useful you have to collect the data for all systems and after that, prepare a report where you have:

• The upgrades done on all systems (e.g. a libc or kernel update)
  
  seperated from
  
• the updates specific to a certain system or system class

Its also error-prone because humans make mistakes.

Perl to the rescue!
At least the part about generating a report about installed/upgraded/removed packages could be automated, because dpkg writes a well-formed logfile in /var/log/dpkg.log. But I noticed that there appearently is no library specialised at parsing that file. Its not a big deal, because the format of that file is really simple, but a proper library would be nice anyway.
And so I wrote such a library.
It basically takes a logfile, reads it line by line and stores each line parameterized into a generic object.

Its features include:

• Parsing the logfile (obviously)
• Storing each line in a logfile as a DPKG::Log::Entry object that holds informations about the entry type (e.g. action or status update), the package associated (if any), timestamp, verbatim line, line number etc.

• Limiting the number of lines parsed to a time range specified by DateTime objects

Based on that, I wrote another library DPKG::Log::Analyse, which takes a log, parses it with DPKG::Log and then extracts the more relevant information such as installed packages, upgraded packages, removed packages etc.

This, in turn, features:
- Info about newly installed packages
- Info about removed packages
- Info about upgraded packages
- Info about packages which got installed and removed again
- Info about packages which stayed halfinstalled or halfconfigured at the end of the logfile (or defined reporting period)


These libraries are already uploaded to CPAN and packaged for Debian.
They passed the NEW queue very quickly and are therefore available for Sid:

http://packages.debian.org/sid/libdpkg-log-perl

As an example use (and for my own use case, as stated above), I wrote dpkg-report, which uses the module and a Template::Toolkit based template to generate a report about what happened in the given logfile.
It currently misses some documentation, but it works somehow like this:

Report for a single host over the full log:
dpkg-report

Report for a single host for the last two days:
dpkg-report --last-two-days

Report for multiple hosts (and logfiles):
The script expects that each log file has the name <systemname>.dpkg.log so that it can guess the hostname from the system and can grab all such log files from a directory if a directory is specified as log-file arguments:</systemname>

dpkg-report --log-file /path/to/logs

This will generate a report about all systems without any merging.

Report for multiple hosts with merging:
dpkg-report --log-file /path/to/logs --merge

This will do the following:

• Report packages which where installed/removed/etc. on all systems seperate from the upgrades only done on the specific systems
• If systems start with a common name and end on a number (e.g. mail1, mail2, mail3 etc.), report packages which where installed on all systems with that common name seperate from the upgrades only done on the specific systems
• For a specific system list only the changes on that particular system and nothing else.

A (fictive) report could look some what like this:

dpkg-Report for all:
------------------------
Newly Installed:
abc
Removed:
foo
Upgraded:
bar (0.99-12 - 1.00-1)

dpkg-Report for test*:
------------------------
Newly Installed:
blub

dpkg-Report for test1:
------------------------
Upgraded:
baz (1.0.0-1 -> 1.0.0-2)

dpkg-Report for test2:
------------------------
Upgraded:
zab (0.9.7 -> 0.9.8)

Currently this report generator is only included in the source package or in the git(hub) repository of the library. I wonder if it makes sense to let the source package build another binary package for it.
But its only a 238 lines perl script with a dependency on the perl library so I'm unsure if it warrants a new binary package. What do others think?