Dropbox is a popular service to synchronize files between multiple computers. The service is entirely proprietary but the company is Linux friendly and provides Linux binaries ready to use. They even provide Ubuntu packages that wrap the dropbox client and provide integration with Nautilus. A bit of story Unfortunately for Debian users, those packages do not work on Debian due to a dependency that can t be satisfied (because Ubuntu introduced an epoch on the version of their nautilus package that Debian doesn t have). This was even reported in a Squeeze review in Linux Weekly News. At some point, Ivan Borzenkov introduced a dropbox package to Debian but it was not based on the above package, instead it packaged directly the proprietary binaries. This was a bad decision because the binaries bundle a set of LGPL libraries and nobody from Debian wanted to do the required work to provide the corresponding source code. So the package got dropped (see bug #610300). More recently several persons filed ITP (Intent To Package) bugs stating their willingness to re-introduce dropbox in Debian, but after many months lingering in the bug tracking system (see #544499, #613788), they have been turned back to RFP (Request For Package) because they changed their minds. What I did Being a dropbox user myself (despite the recent proof that data stored on dropbox is not 100% private), I offered sponsorship to the volunteers who wanted to package dropbox. But it turns out this was not enough to motivate someone to complete the task. In the mean time I was still using the old dropbox package that was removed (it used to be downloadable from snapshot.debian.org). While this was good enough for me, it s clearly not OK in the long term and way too difficult for the majority of users. So this week-end I spent some hours to create a proper package. It s loosely based on the package provided by Dropbox but I upgraded the packaging and changed the way it works. I patched the dropbox wrapper to provide a dropbox update command that downloads and updates the proprietary binaries. They are now stored in /var/lib/dropbox instead of having a copy in each user s home directory (~/.dropbox-dist/). This update command is run by the postinst so that installing the package immediately downloads the proprietary binaries. Get the Dropbox packages for Debian The package nautilus-dropbox has been uploaded to Debian unstable, it s currently in the NEW queue but will shortly reach the mirrors. Then you will be able to You can install the package with a simple apt-get install nautilus-dropbox (provided that you activated the non-free section since that s where the package is hosted, it can t be part of Debian since it requires the proprietary binaries to be useful). In the mean time you can get the packages from the links below:

• For Debian Squeeze: i386 amd64
• For Debian unstable/wheezy: i386 amd64
• For Debian unstable with GNOME3 from experimental: i386 amd64

Get the Dropbox packages for Ubuntu I have setup a PPA for nautilus-dropbox. Feel free to use it in place of the upstream packages.

$ sudo add-apt-repository ppa:hertzog/nautilus-dropbox
$ sudo apt-get update
$ sudo apt-get install nautilus-dropbox

It also includes packages for oneiric, but in theory once the package is accepted into Debian, it should appear in oneiric shortly after. Package maintenance I have done the initial packaging work but I don t really want to maintain it in the long term. I have more than enough to do with dpkg and my other packages. So if you are interested in maintaining this package, please get in touch with me. You should know a bit of python since there are Debian-specific patches of the upstream code. The package is maintained in a git repository. Feedback If you have encountered a problem with one of those packages, feel free to leave a comment. If you re an happy user of the above packages, click here to find out how you can thank me. I hope you enjoy the packages!

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

On June 8, 2011, will be World IPv6 Day. Major web companies like Google, Yahoo! and Facebook will serve their content through IPv6. While some of them already provide IPv6 content, you either need to use a special URL or get your connection from some selected providers, known to provide working IPv6 connection (the selection is done through DNS).

What s World IPv6 Day about? World IPv6 Day is mostly about major content providers and network operators. End users are not expected to take any special actions. The content will be provided for one day on IPv6 and IPv4 without any special trick to circumvent broken IPv6 setup. If you have an IPv4 only connectivity, all will continue to work (this day is not about switching Internet to IPv6 only). If you have a working IPv6 connectivity of any kind, you will use it to access content. However, if you have a broken IPv6 connectivity, you will find yourself isolated during one day until you solve the problem. You can check your connectivity and determine if you will run into trouble on World IPv6 Day. The main point of this day is to check if IPv6 native content can be provided to anybody without leaving out a large part of the users (the ones with a broken connectivity) which is reason why we don t already have IPv6 native content everywhere. Google explains it this way:

We continuously conduct detailed measurements on the quality of IPv6 connectivity, and our latest results show that making Google services generally available over IPv6 at this time would lead to connection problems and increased latency for a small number of users. User experience is very important to us, and we do not want to impact users on networks that do not yet fully support IPv6. We will continue to re-evaluate the situation as the IPv6 Internet evolves.

While content provider are invited to provide IPv6 content, network operator are expected to troubleshoot their customers problem (and they cannot withdraw themselves from this operation since the main change is done by content providers). If they can enable IPv6 connectivity, that would be great too but this is not mandatory. UPDATED: It s a shame but Microsoft s participation to World IPv6 day is a fix to prefer IPv4 over IPv6. Moreover, this fix will be reverted a few days later. Let s hope that network operators don t follow the lead by not setting up DNS hijacking to remove AAAA record.

What can I do? Even if you are neither a network operator nor a major player in web content publishing, it is expected that World IPv6 Day will also be the playground for most IPv6 related things.

As a user Check if your ISP provides native IPv6 connectivity. This can be an opt-in option. For example, in France, Free allows its users to enable native IPv6 connectivity if they choose the appropriate option. That s easy to do and you get IPv6 in a matter of minutes. If your ISP does not provide native IPv6 connectivity, you can try to bug them about this. This is unlikely to work. You can also get IPv6 through a tunnel broker like SixXS (create an account and apt-get install aiccu) or Hurricane Electric. Wikipedia has a list of IPv6 tunnel brokers. Don t rely on any other transition mechanism, they are unreliable.

As a service provider If you host yourself some services, like your web page or your own mail server, you can enable IPv6 connectivity for those as well. Get an IPv6 connectivity for your services, either native or through a tunnel broker. Then, enable IPv6 on each service you provide. Check that they are listening on IPv6 with ss -6l. All the tips below are meant for very simple setup. Ensure that net.ipv6.bindv6only is set to 0. Here is what I have done:

• Enable SSH access through IPv6. With the default configuration of OpenSSH, this is already done. Check your ListenAddress directive if you have one.
• Serve Web content through IPv6. I am using Nginx as a web server and you can mix IPv4 and IPv6 in the same virtual host. Setting up IPv6 is easy. Ensure that IPv6 support is compiled in (nginx -V) use listen [::]:80 instead of listen 80. Apache is equally easy to configure and should even work out of the box since the Listen 80 directive will listen to both IPv4 and IPv6 by default. lighttpd also supports IPv6. An easy way to get started is to use server.use-ipv6 = "enable".
• Enable IPv6 in Postfix, Dovecot, BIND. For all of those, this works out of the box. Nowadays, most daemons will work fine with IPv6 without any special configuration since they will listen to :: by default which means any IPv4 or IPv6 address as long as net.ipv6.bindv6only is set to 0. Unfortunately, there is still some daemons that do not support IPv6.
• Add appropriate AAAA record to your DNS because providing IPv6 services is of little use if your users (including you) cannot access them without typing the appropriate IP address. You may also want to provide a NS record reachable through IPv6.
• Update your monitoring to use IPv6 as well. If you advertise IPv6 service and only IPv4 is working, most IPv6 client will be unable to access them.
• Adapt your firewall. In Linux, Netfilter handles IPv4 and IPv6 separately. You need to ensure you setup a correct firewall for IPv6 as you have done with IPv4.

Here is a minimal firewall that you can adapt:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

ip6tables -A INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT ip6tables -A OUTPUT -m state --state ESTABLISHED,RELATED -j ACCEPT for chain in INPUT OUTPUT; do ip6tables -A $chain -m rt --rt-type 0 -j DROP ip6tables -A $chain -p icmpv6 --icmpv6-type neighbor-solicitation \ -m hl --hl-eq 255 -j ACCEPT ip6tables -A $chain -p icmpv6 --icmpv6-type neighbor-advertisement \ -m hl --hl-eq 255 -j ACCEPT ip6tables -A $chain -m state --state NEW \ -p icmpv6 --icmpv6-type echo-request -j ACCEPT done ip6tables -A INPUT -i lo -j ACCEPT ip6tables -A OUTPUT -o lo -j ACCEPT # Put your rules here! ip6tables -N REJECT_ALL ip6tables -A REJECT_ALL -p tcp -j REJECT --reject-with tcp-reset ip6tables -A REJECT_ALL -p udp -j REJECT --reject-with icmp6-adm-prohibited ip6tables -A REJECT_ALL -j DROP for chain in INPUT OUTPUT FORWARD; do ip6tables -A $chain -j LOG --log-prefix "$chain-reject: " ip6tables -A $chain -j REJECT_ALL done

Debian status There is still a long way for full IPv6 support of all applications in Debian but it is a release goal since Etch. This is not an easy task since adding IPv6 support on some software can be really difficult. DSA also enabled IPv6 support on many servers used by the Debian project including the main web server, the main mirror, the security mirror and the name servers. You should be able to use Debian with an IPv6 only access.

As I write this, it's the morning of June 5th, 1981. A few people scattered across the US are waking up, going in to work, sitting down at their terminal with a coffee, and reading Usenet. Usenet is only getting a trickle of posts each day -- it's still in that period where it's easy to read every message posted to it. Many things lie in Usenet's future. It's still running A-News, which doesn't even have a real From header yet. Later this year it will switch over to B-News, and volume will begin to increase. In 1987 there will be The great renaming. And of course in 1994, the first spam will be posted to Usenet. But that's all a long way off, here in 1981. Right now, they're talking about 500 mb disk drives that only cost $38000. And rms is inciting flames about nuclear proliferation. And Postel is publishing an RFC for the new Mail Transfer Protocol. Good morning, Usenet. Who knows what will come next in this fledgeling electronic communications medium! a ten year real-time historical exhibit This morning, I'm announcing a new site: Olduse.net It's Usenet, updated in real time as it was thirty years ago. Planned to be available for the next ten years, unless I run out of inodes (again). If you missed it the first time around, this is your chance to follow Usenet's flowering. made possible by

• Henry Spencer at the University of Toronto, Department of Zoology, who archived Usenet. Back when it was really uncool and really expensive. Our view onto Usenet is thus slightly centric to Canada and Zoology, but that's ok.
• David Wiseman, who hauled 141 magtapes in a pickup truck.
• Many who worked to rescue data off the tapes. Including from the deleted stuff at the ends.
• Rich Skrenta, who somehow got a copy of the archive out from under the Google borg. Although one of the tar files is truncated. Just saying.
• The creator of Telehack, who pointed me in the right direction, ending my multi-year quest to find the archive. And if you think this is neat, Telehack will blow you away.
• The developers of Haskell, which enabled me to whip up a B-News to C-News converter, a custom uucp, date parsers for every crazy date format ever used on Usenet, and suitible queue data structures in a rock solid, maintainable way, in 500 lines of code written over 12 hours. When I realized I also needed an A-News to B-News converter, I knew it was worth it to have done things right, because that took only 43 more lines, and worked 100% on the first run! My code repository for olduse.net is here.

---

PS: You can post to olduse.net, but it won't show up for at least 30 years. :)

During the past week I finally got around to give the Vendor Profiles branch some attention. I suspect the most exciting change is that I got support for non-overridable tags. Along with making --ftp-master-rejects an alias for the ftp-master profile. This means that you will get the same behavior from Lintian and from the ftp-master auto-reject mechanism. The only issue is that the overrides are silently ignored. We also made dropped the default profile symlink in favor of using dpkg-vendor to find the best profile. This causes a slight conflict with solving #626476. I have asked the dpkg maintainers if that part of the libdpkg-perl API is stable enough for us to use. Finally I fixed an issue where profiles would ignore all display settings (--display-info etc.) and just show all tags (except experimental tags). Beyond the vendor profile branch, there are also a number of interesting changes applied in the master branch. Together with some people from #debian-python (ScottK and jwilk as I recall), I prepared some python related goodies. Particularly we fixed a false-positive missing-python-build-dependency with python3 and marked dh_pycentral and dh_python as deprecated. On a related note, we have also removed the tag uses-dh-python-with-no-pycompat. Based on feedback from Eric Lavarde, we redefined the classpath-contains-relative-path tag. It now only emits the bad parts of the Classpath and relative entries are allowed in they exists or point to /usr/share/java/$file.jar and there is a strong libX-java dependency. While the change breaks overrides, according to lintian.d.o we do not have any yet. The next release will feature a check for non-empty dependency_libs fields in la files and a fix to a flaw in logic for architecture dependent overrides. We will be reducing our dependency of dpkg-dev to a Suggests (#626476). Currently the xz-utils is in the pseudo-essential set, but it may not stay that way. So in the future you may need to explicitly install xz-utils for Lintian to support .xz source packages. Last, but not least, if you have felt that the lintianrc file has been rather useless to you, then you are not alone. I have made a prototype patch to improve the situation and hopefully soon you get to specify default options in your lintianrc.

This is for my own reference more than anything else; periodically I find myself needing to boot DOS. Usually because I want to upgrade a BIOS and the upgrade tool only works with Windows or DOS. Yes, I'm aware of flashrom, but the scary messages about toasting laptops means I'm much happier going the DOS route there. So I need a writable medium with a DOS boot image. The easy way to do it is to find a DOS floppy image and write that to a USB stick with dd, but it's basically a one off and means you can't fit a whole lot on the image. I wanted to do a native USB boot. The following did the trick for me. Firstly, this is all on a Debian testing box, in particular with dosfstools (3.0.9-1), mbr (1.1.11-4) and syslinux (2:4.04+dfsg-2). I don't think I'm using anything particularly new from these tools, so I suspect Debian stable will work just fine. In the below my USB stick is on /dev/sdb, I've got a copy of FreeDOS in ~/FreeDOS/ (I extracted the files from a balder10.img floppy image that's easy enough to find, but all you need is kernel.sys and command.com), and I wanted things neatly in their own subdirectories so I could try and keep the key for repeated use and know which bits I wanted to keep and which were transient. I also added a copy of the Debian Installer for convenience; it would need a copy of the netinst ISO dropped into the root to actually be useful, as there's not enough to do a full netboot from the wider Internet using just the HD initrd.

    # Create a single bootable LBA VFAT partition spanning entire stick.
    echo 0,,C,*   sfdisk /dev/sdb
    # Format as VFAT and give it a disk label.
    mkfs.vfat -n 'DOS BOOT' /dev/sdb1
    # Install an MBR
    install-mbr /dev/sdb
    # Mount it and create some directories for our files.
    mount /dev/sdb1 /media/DOS\ BOOT/
    cd /media/DOS\ BOOT/
    mkdir boot
    mkdir boot/syslinux/
    mkdir FreeDOS
    # Copy the chain.c32 syslinux tool over.
    cp /usr/lib/syslinux/chain.c32 boot/syslinux/
    # Copy FreeDOS over and move the important bits to /
    cp ~/FreeDOS/* FreeDOS/
    mv FreeDOS/command.com .
    mv FreeDOS/autoexec.bat .
    mv FreeDOS/fdconfig.sys .
    # Add the Debian Installer
    mkdir boot/d-i/
    cd boot/d-i/
    wget http://cdn.debian.net/debian/dists/squeeze/main/installer-amd64/current/images/hd-media/initrd.gz
    wget http://cdn.debian.net/debian/dists/squeeze/main/installer-amd64/current/images/hd-media/vmlinuz
    # Install syslinux
    syslinux -d boot /dev/sdb1

I made a boot/syslinux.help file containing:

    Available boot options are: FreeDOS d-i

and a boot/syslinux.cfg with:

    prompt 1
    display syslinux.help
    default FreeDOS
    label FreeDOS
        COM32 /boot/syslinux/chain.c32
        APPEND freedos=/FreeDOS/kernel.sys
    label d-i
        KERNEL /boot/d-i/vmlinuz
        INITRD /boot/d-i/initrd.gz

Initial testing was done using kvm which was a lot easier than constantly rebooting my laptop. Hopefully this is of help to someone. The only neat bit that I didn't see elsewhere when I was looking was the use of chain.c32 to load kernel.sys rather than having to use a FreeDOS provided boot sector image.

My last update about running activities is dated back to Paris Marathon (April 10th). Time for a new update, then. After a very busy end March-early April, with that crazy succession of 55km Ecotrail and 42km Paris marathon with only two weeks between both, I needed a kind of "break". Moreover, I had another goal set with the 35m trail race named "Trail des Cerfs", on May 15th, in my neighbourhood. So, end of April was a little bit quieter. Still, one week after Paris marathon, I completed a very hard trail run in Seine Valley: only 18km, but about 600-700m positive slope. That was indeed a good moutain-style run (yes, in Seine valley....) and a good way to check that I recovered quite well. Later on, I slowed down a little with even a Sunday (24th) where I *didn't run at all*..:-).. Just a few short uns during the week and a short Saturday run. Well, that was indeed Easter Sunday and Monday is a holiday in France, so I ran an half-marathon on 25th...:-)...on a loop where I already have some references, the fastest being 1h46. This time it was 1h52'...but, well, that was two weeks after a marathon... Still, a small pain at the back of my ankle decided me to stop running long distances until May 15th. So, I only did short trainings up to no more than 12km, even during my stay in Germany for SambaXP. As a consequence, I was very well for the 35km (600m positive slope) Trail des Cerfs, which I completed in 3h31', 168th out of 328 runners. More interestingly, I completed it in negative split with a much faster second half, ending up at about 12km/h. For seomthing that (given the slope and the nature of the paths) is comparable to a marathon, that's very promising. The rest of May was spent quietly with a standard training and progressively resuming longer distances. I originally planned to do my spring solo marathon on May 28th in Versailles Castle park, but I stopped at km28 and completed the distance two days after with a 14km run..:-)... Stopping was wiser as I was feeling that the last loop would have been a great pain. That was indeed the first time I give up on a marathon. I prefer this to happen for a non official race, after all! June will not see any "big goal". I might be running one race somewhere in my neighbourhood but the point will mostly be accumulating kilometers before summer. Summer (which includes DebConf that makes training a little bit more hectic) will not see any race, probably (unless I find one in banja Luka!) but more some specific preparation for Berlin marathon. Here, the "safe" target for me is, first, breaking my record. I'm probably not ready for 3h30 and even 3h35 might be difficult. It seems that I'm now reaching a barrier where progress becomes harder. Before Berlin, I'll run the Bois d'Arcy half-marathon, two weeks before. Last year, I managed to set a very good mark, there, with 1h37. Not sure that i'll try to break this only two weeks before Berlin. After Berlin? We'll see, but my last big goal of the year will be the 67km "Le Puy-Firminy" night race, in November (no web site, this is indeed a "small" local race). Yet another adventure beyond marathon distances, in the unknown...

Erich Schubert's blog post reminded me that I've been meaning to writeup a post detailing how I'm keeping parts of my $HOME in git repositories. My goal has been to keep my home directory in a version control system effectively. I have a number of constraints however. I want the system to be modular. I don't always need X related config files in my home directory. Sometimes I want just my zsh related files and my emacs related files. I have multiple machines I check email from, and on those want to keep my notmuch/offlineimap files in sync, but I don't need these on every machine I'm on, expecially since those configurations have more sensitive data. I played around with laysvn for a while, but it never really seemed comfortable. I more recently discovered that madduck had started a "vcs-home" website and mailing list, talking about doing what I'm trying to do. I'm now going with madduck's idea of using git with detached work trees, so that I can have multiple git repositories all using $HOME as their $GIT_WORK_TREE. I have a script inspired by his vcsh script that will create a subshell where the GIT_DIR, GIT_WORK_TREE variables are set for me. I can do my git operations related to just one git repository in that shell, while still operating directly on my config files in $HOME, and avoiding any kind of nasty symlinking or hardlinking. Since I am usually using my script to allow me to quickly "move in" to a new host, I named my script "movein". It can be found here. Here's how I'll typically use it:

    stew@guppy:~$ movein init
    git server hostname? git.vireo.org
    path to remote repositories? [~/git] 
    Local repository directory? [~/.movein] 
    Location of .mrconfig file? [~/.mrconfig] 
    stew@guppy:~$ 

This is just run once. It asks me questions about how to setup the 'movein' environment. Now I should have a .moveinrc storing the answers I gave above, I have a stub of a .mrconfig, and an empty .movein directory. Next thing to do is to add some of my repositories. The one I typically add on all machines is my "shell" repository. It has a .bashrc/.zshrc, an .alias that both source and other zsh goodies I'll generally wish to be around:

    stew@guppy:~$ ls .zshrc
    ls: cannot access .zshrc: No such file or directory
    stew@guppy:~$ movein add shell
    Initialized empty Git repository in /home/stew/.movein/shell.git/
    remote: Counting objects: 42, done.
    remote: Compressing objects: 100% (39/39), done.
    remote: Total 42 (delta 18), reused 0 (delta 0)
    Unpacking objects: 100% (42/42), done.
    From ssh://git.vireo.org//home/stew/git/shell
     * [new branch]      master     -> origin/master
    stew@guppy:~$ ls .zshrc
    .zshrc

So what happened here is that the ssh://git.vireo.org/~/git/shell.git repository was cloned with GIT_WORK_TREE=~ and GIT_DIR=.movein/shell.git. My .zshrc (along with a bunch of other files) has appeared. Next perhaps I'll add my emacs config files:

    stew@guppy:~$ movein add emacs       
    Initialized empty Git repository in /home/stew/.movein/emacs.git/
    remote: Counting objects: 77, done.
    remote: Compressing objects: 100% (63/63), done.
    remote: Total 77 (delta 10), reused 0 (delta 0)
    Unpacking objects: 100% (77/77), done.
    From ssh://git.vireo.org//home/stew/git/emacs
     * [new branch]      emacs21    -> origin/emacs21
     * [new branch]      master     -> origin/master
    stew@guppy:~$ ls .emacs
    .emacs
    stew@guppy:~$ 

My remote repositry has a master branch, but also has an emacs21 branch, which I can use when checking out on older machines which don't yet have newer versions of emacs. Let's say I have made changes to my .zshrc file, and I want to check them in. Since we are working with detached work trees, git can't immediately help us:

    stew@guppy:~$ git status
    fatal: Not a git repository (or any of the parent directories): .git

The movein script allows me to "login" to one of the repositories. It will create a subshell with GIT_WORK_TREE and GIT_DIR set. In that subshell, git operations operate as one might expect:

    stew@guppy:~ $ movein login shell
    stew@guppy:~ (shell:master>*) $ echo >> .zshrc
    stew@guppy:~ (shell:master>*) $ git add .zshrc                                       
    stew@guppy:~ (shell:master>*) $ git commit -m "adding a newline to the end of .zshrc"
    [master 81b7311] adding a newline to the end of .zshrc
     1 files changed, 1 insertions(+), 0 deletions(-)
    stew@guppy:~ (shell:master>*) $ git push
    Counting objects: 8, done.
    Delta compression using up to 2 threads.
    Compressing objects: 100% (6/6), done.
    Writing objects: 100% (6/6), 546 bytes, done.
    Total 6 (delta 4), reused 0 (delta 0)
    To ssh://git.vireo.org//home/stew/git/shell.git
       d24bf2d..81b7311  master -> master
    stew@guppy:~ (shell:master*) $ exit
    stew@guppy:~ $ 

If I want to create a brand new repository from files in my home directory. I can:

    stew@guppy:~ $ touch methere
    stew@guppy:~ $ touch mealsothere
    stew@guppy:~ $ movein new oohlala methere mealsothere
    Initialized empty Git repository in /home/stew/git/oohlala.git/
    Initialized empty Git repository in /home/stew/.movein/oohlala.git/
    [master (root-commit) 7abe5ba] initial checkin
     0 files changed, 0 insertions(+), 0 deletions(-)
     create mode 100644 mealsothere
     create mode 100644 methere
    Counting objects: 3, done.
    Delta compression using up to 2 threads.
    Compressing objects: 100% (2/2), done.
    Writing objects: 100% (3/3), 224 bytes, done.
    Total 3 (delta 0), reused 0 (delta 0)
    To ssh://git.vireo.org//home/stew/git/oohlala.git
     * [new branch]      master -> master

Above, the command movein new oohlala methere mealsothere says "create a new repository containing two files: methere, mealsothere". A bare repository is created on the remote machine, a repository is created in the .movein directory, the files are committed, and the new commit is pushed to the remote repository. New on some other machine, I could run movein add oohlala to get these two new files. The movein script maintains a .mrconfig file, so that joeyh's mr tool can be used to manage the repositories in bulk. Commands like "mr update", "mr commit", "mr push" will act on all the known repositories. Here's an example:

    stew@guppy:~ $ cat .mrconfig
    [DEFAULT]
    include = cat /usr/share/mr/git-fake-bare
    [/home/stew/.movein/emacs.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/emacs.git' 'emacs.git' '../../'
    [/home/stew/.movein/shell.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/shell.git' 'shell.git' '../../'
    [/home/stew/.movein/oohlala.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/oohlala.git' 'oohlala.git' '../../'
    stew@guppy:~ $ mr update
    mr update: /home/stew//home/stew/.movein/emacs.git
    From ssh://git.vireo.org//home/stew/git/emacs
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: /home/stew//home/stew/.movein/oohlala.git
    From ssh://git.vireo.org//home/stew/git/oohlala
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: /home/stew//home/stew/.movein/shell.git
    From ssh://git.vireo.org//home/stew/git/shell
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: finished (3 ok)
    stew@guppy:~ $ mr update        

There are still issues I'd like to address. The big one in my mind is that there is no .gitignore. So when you "movein login somerepository" then run "git status", It tells you about hundreds of untracked files in your home directory. Ideally, I just want to know about the files which are already associated with the repository I'm logged into.

Erich Schubert's blog post reminded me that I've been meaning to writeup a post detailing how I'm keeping parts of my $HOME in git repositories. My goal has been to keep my home directory in a version control system effectively. I have a number of constraints however. I want the system to be modular. I don't always need X related config files in my home directory. Sometimes I want just my zsh related files and my emacs related files. I have multiple machines I check email from, and on those want to keep my notmuch/offlineimap files in sync, but I don't need these on every machine I'm on, expecially since those configurations have more sensitive data. I played around with laysvn for a while, but it never really seemed comfortable. I more recently discovered that madduck had started a "vcs-home" website and mailing list, talking about doing what I'm trying to do. I'm now going with madduck's idea of using git with detached work trees, so that I can have multiple git repositories all using $HOME as their $GIT_WORK_TREE. I have a script inspired by his vcsh script that will create a subshell where the GIT_DIR, GIT_WORK_TREE variables are set for me. I can do my git operations related to just one git repository in that shell, while still operating directly on my config files in $HOME, and avoiding any kind of nasty symlinking or hardlinking. Since I am usually using my script to allow me to quickly "move in" to a new host, I named my script "movein". It can be found here. Here's how I'll typically use it:

    stew@guppy:~$ movein init
    git server hostname? git.vireo.org
    path to remote repositories? [~/git] 
    Local repository directory? [~/.movein] 
    Location of .mrconfig file? [~/.mrconfig] 
    stew@guppy:~$ 

This is just run once. It asks me questions about how to setup the 'movein' environment. Now I should have a .moveinrc storing the answers I gave above, I have a stub of a .mrconfig, and an empty .movein directory. Next thing to do is to add some of my repositories. The one I typically add on all machines is my "shell" repository. It has a .bashrc/.zshrc, an .alias that both source and other zsh goodies I'll generally wish to be around:

    stew@guppy:~$ ls .zshrc
    ls: cannot access .zshrc: No such file or directory
    stew@guppy:~$ movein add shell
    Initialized empty Git repository in /home/stew/.movein/shell.git/
    remote: Counting objects: 42, done.
    remote: Compressing objects: 100% (39/39), done.
    remote: Total 42 (delta 18), reused 0 (delta 0)
    Unpacking objects: 100% (42/42), done.
    From ssh://git.vireo.org//home/stew/git/shell
     * [new branch]      master     -> origin/master
    stew@guppy:~$ ls .zshrc
    .zshrc

So what happened here is that the ssh://git.vireo.org/~/git/shell.git repository was cloned with GIT_WORK_TREE=~ and GIT_DIR=.movein/shell.git. My .zshrc (along with a bunch of other files) has appeared. Next perhaps I'll add my emacs config files:

    stew@guppy:~$ movein add emacs       
    Initialized empty Git repository in /home/stew/.movein/emacs.git/
    remote: Counting objects: 77, done.
    remote: Compressing objects: 100% (63/63), done.
    remote: Total 77 (delta 10), reused 0 (delta 0)
    Unpacking objects: 100% (77/77), done.
    From ssh://git.vireo.org//home/stew/git/emacs
     * [new branch]      emacs21    -> origin/emacs21
     * [new branch]      master     -> origin/master
    stew@guppy:~$ ls .emacs
    .emacs
    stew@guppy:~$ 

My remote repositry has a master branch, but also has an emacs21 branch, which I can use when checking out on older machines which don't yet have newer versions of emacs. Let's say I have made changes to my .zshrc file, and I want to check them in. Since we are working with detached work trees, git can't immediately help us:

    stew@guppy:~$ git status
    fatal: Not a git repository (or any of the parent directories): .git

The movein script allows me to "login" to one of the repositories. It will create a subshell with GIT_WORK_TREE and GIT_DIR set. In that subshell, git operations operate as one might expect:

    stew@guppy:~ $ movein login shell
    stew@guppy:~ (shell:master>*) $ echo >> .zshrc
    stew@guppy:~ (shell:master>*) $ git add .zshrc                                       
    stew@guppy:~ (shell:master>*) $ git commit -m "adding a newline to the end of .zshrc"
    [master 81b7311] adding a newline to the end of .zshrc
     1 files changed, 1 insertions(+), 0 deletions(-)
    stew@guppy:~ (shell:master>*) $ git push
    Counting objects: 8, done.
    Delta compression using up to 2 threads.
    Compressing objects: 100% (6/6), done.
    Writing objects: 100% (6/6), 546 bytes, done.
    Total 6 (delta 4), reused 0 (delta 0)
    To ssh://git.vireo.org//home/stew/git/shell.git
       d24bf2d..81b7311  master -> master
    stew@guppy:~ (shell:master*) $ exit
    stew@guppy:~ $ 

If I want to create a brand new repository from files in my home directory. I can:

    stew@guppy:~ $ touch methere
    stew@guppy:~ $ touch mealsothere
    stew@guppy:~ $ movein new oohlala methere mealsothere
    Initialized empty Git repository in /home/stew/git/oohlala.git/
    Initialized empty Git repository in /home/stew/.movein/oohlala.git/
    [master (root-commit) 7abe5ba] initial checkin
     0 files changed, 0 insertions(+), 0 deletions(-)
     create mode 100644 mealsothere
     create mode 100644 methere
    Counting objects: 3, done.
    Delta compression using up to 2 threads.
    Compressing objects: 100% (2/2), done.
    Writing objects: 100% (3/3), 224 bytes, done.
    Total 3 (delta 0), reused 0 (delta 0)
    To ssh://git.vireo.org//home/stew/git/oohlala.git
     * [new branch]      master -> master

Above, the command movein new oohlala methere mealsothere says "create a new repository containing two files: methere, mealsothere". A bare repository is created on the remote machine, a repository is created in the .movein directory, the files are committed, and the new commit is pushed to the remote repository. New on some other machine, I could run movein add oohlala to get these two new files. The movein script maintains a .mrconfig file, so that joeyh's mr tool can be used to manage the repositories in bulk. Commands like "mr update", "mr commit", "mr push" will act on all the known repositories. Here's an example:

    stew@guppy:~ $ cat .mrconfig
    [DEFAULT]
    include = cat /usr/share/mr/git-fake-bare
    [/home/stew/.movein/emacs.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/emacs.git' 'emacs.git' '../../'
    [/home/stew/.movein/shell.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/shell.git' 'shell.git' '../../'
    [/home/stew/.movein/oohlala.git]
    checkout = git_fake_bare_checkout 'ssh://git.vireo.org//home/stew/git/oohlala.git' 'oohlala.git' '../../'
    stew@guppy:~ $ mr update
    mr update: /home/stew//home/stew/.movein/emacs.git
    From ssh://git.vireo.org//home/stew/git/emacs
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: /home/stew//home/stew/.movein/oohlala.git
    From ssh://git.vireo.org//home/stew/git/oohlala
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: /home/stew//home/stew/.movein/shell.git
    From ssh://git.vireo.org//home/stew/git/shell
     * branch            master     -> FETCH_HEAD
    Already up-to-date.
    mr update: finished (3 ok)
    stew@guppy:~ $ mr update        

There are still issues I'd like to address. The big one in my mind is that there is no .gitignore. So when you "movein login somerepository" then run "git status", It tells you about hundreds of untracked files in your home directory. Ideally, I just want to know about the files which are already associated with the repository I'm logged into.

In February I started the process of moving my phone and my wife s phone to Virgin from Three [1]. The reason is that Three didn t offer any good phones on affordable contracts, the cheapest that was suitable was a HTC Desire HD which would have cost me $55 per month, while I could justify spending that for my own phone (which is used for responding to SMS from Nagios to fix client servers) I didn t have the budget to spend that much on my wife s phone too and I really want us to have the same type of phone for ease of support. So I chose Virgin who offers the Sony Ericsson Xperia X10 for as little as $29 per month I chose a $39 per month deal that included 1500MB of data transfer and also had three months free which makes it effectively $34.12 per month. When using previous phones that weren t particularly smart I had also carried a Netbook and a 3G modem with me most of the time. Now that I have a phone that is a ssh client I don t need that so I tried to cancel the contract today. Three allows you to do almost everything over the Internet except cancel a contract their web site doesn t even give a phone number to call for that purpose. This must keep their support people busy, Vodaphone (which has just merged with Three) has recently had a horrible security breach because their sales booths used public Internet access for all customer data [2]. Also there is currently a law suit against Vodafone for poor network performance and misleading claims about service areas [3]. My experience with Three performance has been reasonably good apart from the fact that they advertised 3G service in Bendigo and provided none. As Three are apparently desperate to retain customers they offered me free service for 6 months if I don t close the account now. So I have a SIM that supports 1G of 3G data transfer per month for no charge until December (worth $90). What can I do with it? I don t own a 3G modem as I gave that to my parents (who are quite happy with pre-paid 3G net access via Three) and the phones that I have which can be used for tethering are a little slow (usable for ssh and basic web access but not for Youtube etc). Is there a way of selling such a SIM? Note that my name is still on the contract and any excess data or roaming fees will be billed to me so I can t just put it on ebay. I guess that one thing I can do is to use the SIM for receiving phone calls. For example if a friend was visiting from another country and wanted to receive calls without paying roaming fees I could lend them a phone. Any ideas?

• [1] http://etbe.coker.com.au/2011/02/18/xperia-x10i-first-experience/
• [2] http://www.smh.com.au/technology/security/mobile-security-outrage-private-details-accessible-on-net-20110108-19j9j.html
• [3] http://www.theaustralian.com.au/australian-it/it-business/users-join-legal-suit-on-vodafone-network-performance/story-e6frganx-1226044675451

Hello, this is the official project report for the Debian GSoC admins and everyone. It s been few days since the official coding phase started. Most of the time I ve been figuring out the details regarding integration of non-x86 images to the Eucalyptus IaaS and working on the ARM image, as you ll learn later. I m also in the process of starting to design the set of tools we ll be delivering as part of the project. The project s goal is to be a useful resource for Debian developers and porters. I ve joined some porters list in order to learn what are the things they are struggling with and how the project can help to address them. I ll be starting a wiki page with such topics. Bonding period:
Bonding period has been more fruitful than I expected. As mentioned in my previous post, Steffen and I got in touch with the Eucalyptus team, since they show their early interest on the project. We are coordinating cooperation between our teams. Besides my mentor and I defined to start working on the ARM image. Coordination went OK. We ll be having meetings every two weeks. Project Status:
Resuming Dominique s work I began working on the ARM image. Aur lien s work was a great resource. Currently I have a working and updated Debian testing ARM image using the versatile kernel that can be used under qemu-arm. I m currently making tests and figuring out the next phase which is cloudify it. Over this weekend I expect to finish on that side and then began on the design and implementation of the tools that will be the project s result.

Until now I haven t faced much issues, but the Eucalyptus part since the docs are targeting x86 and amd_64 archs. However I think I can run qemu-arm on top. I ll talk to them about the idea.

Future plans: week 3: - Finish tests with the ARM image. - Design the base tool-set, thinking integration with dpkg-buildpackage and qemubuilder. week 4-8: - Write the necessary code. mid-term evaluation. Other bits: Latest days have been quite busy for me, since I m still have to attend class, my semester ends in five weeks. This week was quite overwhelming with media , my University published an article (Spanish) about me and the program. I m the third student of the CS school participating in GSoC, in the past we had Gentoo Project and KDE students. Later, a local Cable TV program interview me on the same matter, it went fine. I managed to talk about Debian and what we do (we get free Advertising yay!) :) also about Google s Summer of Code program and invited students to participate both in FLOSS projects and the program. Still don t know if they have put it on air, I think it will be next week.

DECLARATIVE DIVERSIONS ====================== Introduction A diversion is when it is possible to have dpkg not overwrite a file when it reinstalls the package it belongs to, and to have it put the file from the package somewhere else instead. Declarative diversions involves a new control file with a declarative syntax which dpkg will parse and process directly as part of the package unpack and removal phases, eliminating the problems resulting from non-atomic handling of diversions. Topics There are a number of topics involved in implementing this kind of project * What syntax do we use for the new control file? * What dpkg does with the control file * How do we handle diversions to a non-existant directory? * How do we order unpacking a new package which adds a diversion? * How do we order unpacking a new package that removes a diversion? * How do we order removing a package which had a diversion? * How do we handle errors? * What happens to dpkg-divert? - Details - Control File Syntax - It will conform to RFC2822 style with the following format: * Divert-From: * Divert-To: * Blank lines and lines beginning with # will be comments Divert-To will be optional and if it is ommitted then files being diverted will have their filename changed to file.distrib The above style has it s advantages, one of the main ones being that there is no need to escape whitespace within filenames. - Details - Control File Handling - Within control.tar.gz the file should be named diversions This file is then copied to /var/lib/info/$package.diversions This is ensuring we have a copy of the control file just in case it is needed. Any input on this subject would be appreciated. - Details - Handling Diversions to non-existant directories - Diverting files to directories that don t exist can cause a number of problems. If the package does not own the directory it may be left orphaned on removal of the package The package is responsible for ensuring the availability of the target directory in the unpack phase. - Details - Ordering Requirements - =>Unpacking a new package that adds a diversion *1. Unpack folders *2. Add diversions *3. Unpack files ==Notes== Performing diversions this way avoids trouble with creating folders and accidentally overwrtiting the wrong version of files. If we don t use this method the problems that can occur include creating folders before the package itself would create them which may cause permission errors. Another major problem would be extracting a new version of the file and overwriting the old one. This might cause the wrong version of the file to be diverted and the package to break. =>Unpacking a new package that removes a diversion *1. Unpack folders *2. Add diversions *3. Unpack files ==Notes== =>Removing a package which had a diversion *1. Remove files *2. Remove diversions *3. Remove folders ==Notes== This ensures that all files, diversions and folders are removed correctly Details - Error handling Errors in diversions will have to handled with a great deal of care due to the fact that if they are not the package could be broken. This means that a great deal of checks must be done to ensure that all the files can be diverted properly before any actual diverting takes place. If they can t the package installation/update must be stopped and rolled back to avoid the package being installed incorrectly or broken. - Details - dpkg-divert - When we impliment the new diversion method we should keep the current dpkg-divert. This allows maintainers to catch up with the new method without breaking their packages. It also allows maintainers to perform some operations that aren t support by the new method. Example Usage #1 The file to be diverted is /usr/share/foo It needs to be moved to /usr/share/bar The syntax of the control file would be: #Start File Divert-From: /usr/share/foo Divert-To: /usr/share/bar #End File Example Usage #2 In this example the maintainer doesn t want to move the file to any specific folder The syntax of the control file would be: #Start File Divert-From: /usr/share/foo #End File This would divert the file to /usr/share/foo.distrib - Footnotes - RFC2822 Guide: * http://www.faqs.org/rfcs/rfc2822.html First Email Thread on Declarative Diversions (First Message in Thread): * http://lists.debian.org/debian-dpkg/2011/05/msg00102.html Declarative Diversions Wiki: * http://wiki.debian.org/SummerOfCode2011/DeclarativeDiversions My Blog for this Project: * http://blog.sam-dunne.com

The beginning of this Google Summer of Code has been busy and pretty cool. Not a lot of code was produced but I did learn a lot of things.

During this GSoC I will work on the next generation JDK. This means that I will (and I already have) compile the JDK. That s pretty awesome but it requires a lot of power. That s why Tom Marble created me an account on his 8 CPUs machine to help me.
Since I m probably going to sign some packages, I created a new GPG key with a 4096 RSA strength.
My old key was:

pub 1024D/F144A319 2008-10-18

And the new key is:

pub 4096R/EE2BBBC7 2011-05-03

I cross signed my keys using this great tutorial.

I read some documentation about Jigsaw and watched Tom s talk at DebConf 10.
I have spent a little more than a week to understand the current OpenJDK packaging to be able to rebuild it. I have learnt from where do the sources come from. The current packaging is a little hard to understand. There are sources that come from IcedTea, OpenJDK, and more.

After being able to rebuild the OpenJDK 6 package, I have downloaded the source of the OpenJDK 8 and Jigsaw. I used the Mercurial forest to get the sources.

For OpenJDK 8:
hg clone http://hg.openjdk.java.net/jdk8/jdk8 jdk8
cd jdk8 && sh get_sources.sh

For Jigsaw:
hg clone http://hg.openjdk.java.net/jigsaw/jigsaw jigsaw
cd jigsaw && sh get_sources.sh

I have managed to build the sources of OpenJDK 8 and Jigsaw using the same steps:
cd jigsaw
. jigbuild
. jdk/make/jdk_generic_profile.sh
export ALLOW_DOWNLOADS=true
make sanity
make all

Jigsaw specific:
make modules-sanity
make modules

Each build took several minutes. But they were successful. Tom helped me a lot to build OpenJDK 8 and Jigsaw by giving me a solution. For example, after getting the Mercurial forest, there were still missing source files and the build failed. Tom told me that I needed to export a environement variable: ALLOW_DOWNLOADS=true.
After building Jigsaw, I was able to see that the version string is not legal in Debian and that is a real problem.
I have written some simple Hello World examples to compile and run with Jigsaw. To modularize a program, it needs to contain a module-info.java files. This file includes the modules needed to run the program and the main class to run. For a classic hello world, the module-info.java files looks like:

module test.hello @ 1
requires jdk.base @ 7-ea;
class test.hello.Main;


Tom gave me some packages to build: jtharness and jtreg. I succesfully built jtharness using CDBS and its Ant helper but I ll have to repackage it to use javahelper instead of CDBS. I still didn t try to build jtreg because it requires jtharness.

My next plan is to get jtharness and jtreg packaged with javahelper soon and understand what modules does what in Jigsaw. Understaning modules will help me to write some more complex examples. I hope to start working on Jigsaw a little more soon (when I my internship will end [next week]).

So I just read that Google will only support modern browsers starting 2 months from now.

As of August 1st, we will discontinue support for the following browsers and their predecessors: Firefox 3.5, Internet Explorer 7, and Safari 3. In these older browsers you may have trouble using certain features in Gmail, Google Calendar, Google Talk, Google Docs and Google Sites, and eventually these apps may stop working entirely.

Given the importance of Google, the impact is huge. This company has acquired the power to basically dictate what browser you can provide to your users - otherwise they won t be able to access what many of them now consider vital functionality. Such a decision denotes a grave misunderstanding of the workstation ecosystem from the Google people. It means they consider their only target to be nerdy users with home computers they can (and want to) upgrade and break every 3 months with the latest version of Windows or Fedora. What about corporate computers? What about non-techy people who buy a computer and stick with the OS that was sold with it for 4 years? I m afraid they are still the vast majority of web users. You can t decide to deploy a new version of IE of Firefox on a large number of computers for next month. Sometimes, this is not even possible (hello Windows 2000/XP users). For Debian squeeze, this means no more Firefox for you. Epiphany and Konqueror might still work, but Google loves sending JavaScript that make old versions of Webkit struggle. And anyway, this is just the beginning. In a few months they will tell us to upgrade again to Firefox 4.2 and IE 12. One week after their release, yeah! Let s quote a comment which should help understanding the reasoning behind such decisions.

Andy, while I understand staying on LTS, I think it's a little bit silly to use a mission critical machine for web browsing in that way. Also, there is no reason your browser has to be tied in lockstep to your OS. Two simple solutions:
1. Don't use the built-in browser for your main web browsing. Install Chrome, for example. or,
2. Since LTS is designed for servers and other "can't have any chance of downtime" machines, quit using that machine as your web browsing box and use a personal laptop for such things, which you can keep up to date with the current OS release instead of waiting 2 years.

Belief #1: the browser can (and should) be independent from the OS . It s interesting to note that the same people who say this are the ones who also jerk off at the idea of desktops and phones with tight web integration. This integration comes at a cost: this restricts your ability to change everything in the browser from one day to another. Belief #2: long-term OSes are for mission-critical servers only . Yeah sure, that s why Windows has a lifecycle of 3 years. Desktops are no different at all from servers on this matter. You don t upgrade your desktop every 6 months when you do serious work with it; the cost and the risk are just too high. And anyway, this comes again from the same people who want to upgrade every single component of said mission-critical servers every 2 months to install the latest version of their preferred web framework. Belief #3: people can upgrade their browsers or OSes . No really they can t. Many people wouldn t know how to do this, even with a step-by-step documentation. And in enterprise deployments, they are restricted from doing so. Thanks for spreading the clich that web developers are clueless, spotty nerds, incapable of understanding the needs of production environments. Apparently Google is not exempt from this disease.

IIT Bombay, like most likely most Indian educational instituates, has some rules and regulations about the interaction between boys and girls that, at least to the uninitiated eye of a foreigner who was brought up in a liberal society, seem arbitrary and archaic. At IIT Bombay, for example, there may be no visits in hostels by students of the other gender (yes, there are separate boys and girls hostels) between 10pm and 7am, and during the day any such visit has to be registered at the reception security guard. And when there was need for more room for girls and they reassigned one wing of a boys hostel to the adjacent girls hostel, they closed the connection by bricking it and split the lawn by a fence. With barbed wire on top. I took this as a theme for a satirical inspection of these policies which comes to the conclusion that IIT Bombay is a very progressive institute that puts these rules in place as affirmative action towards same-gender relationships. After some discussion with the editors the text got actually published in the institutes bi-monthly newspaper Raintree , unfortunately only now, two weeks after I left India. You can read the text on the bottom of page 8 in the 6MB PDF file. I cannot predict how it will be received, but at least Utkarsh Raut, who illustrated the text, clearly got the message right. Thanks for the nice and very fitting drawing!

The Event I just attended a lecture by Dr Ziggy Switkowski (see the above picture) on the topic of nuclear power, for a long time Ziggy has been the most prominent advocate of nuclear power in Australia, so naturally the talk was pro-nuclear. The lecture was free for alumni of Swinburne University and introduced by Swinburne s Chancellor who described Ziggy as a friend. Before the lecture started I handed out flyers for Beyond Zero Emissions [1] with a plan for zero carbon use in Australia by 2020. That was more difficult than handing out Greens polling cards on election day. The people who attended were there for a purpose and they could see the free food and wine on offer inside the building so getting them to stop and take a brochure wasn t easy. There were also people from Friends of the Earth [2] and some other anti-nuclear organisations giving out brochures. There were also some anti-nuclear protesters dressed as Uranium miners and the specter of nuclear contamination, they made for an interesting picture (which is amazingly good for a phone camera 20 minutes after sunset), but got in the way when I was handing out brochures. There were a few university security people at the door trying to keep the opposition away who were a bit annoying. There was also one policeman in attendance who was quite friendly, he asked whether there was a competition between the different groups as there is when handing out how-to-vote cards on election day. 15 minutes before the lecture was due to start I quit handing out the brochures and went inside for free food and drink (food always tastes better when the other team pays the bill). The Lecture I felt that Ziggy did a poor job of advocating nuclear power. There were a number of gaping flaws in his argument, but I have to commend him for his honesty in his appraisal of how hard it would be to develop a nuclear industry in Australia. Most nuclear advocates are a lot more optimistic about the process of building nuclear power plants and training all the people to run them. I will write a post tomorrow rebutting some of Ziggy s arguments. After the Lecture Immediately after the lecture the vice-chancellor stated that 500 people registered which made it the best attended such lecture that they have held. I don t know how many of the 500 attended, but I did notice that about 1/4 of the seats in the lecture hall were empty. It s rather sad that they can t get more people to such lectures, is TED.com stealing the audience? When I got out of the building I started handing out brochures to people as they left. I noticed that they were a lot friendlier than they were on the way in. I believe that people who left the lecture had greater interest in renewable energy because Ziggy just failed to make a good case for nuclear power. I expect that anyone who hadn t already made up their mind would immediately want to research the alternatives after watching Ziggy s lecture.

• [1] http://beyondzeroemissions.org/
• [2] http://www.foe.org.au/

Well jffnms 0.9.0 was a, well, its a good example of what can go wrong without adequate testing. Having it written in PHP makes it difficult to test, because you can have entire globs of code that are completely wrong but are not activated (because the if statement proves false for example) and you won t get an error. It also had some database migration problems. This is the most difficult and annoying part of releasing versions of JFFNMS. There are so many things to check, like:

• What has changed between this version of the database?
• Does the development database have all the new fields?
• Will importing an old database and using the diffs give me the same result as importing the new database
• Does the structure field match the object which matches the database table

I ve been looking at sqlalchemy which is part of turbogears. It s a pretty impressive setup and lets you get on with writing your application and not mucking around with the low-level stuff. It s a bit of a steep learning curve learning python AND sqlalchemy AND turbogears but I ve got some rudimentary code running ok and its certainly worth it (python erroring on un-assigned varables but not forcing you to define them is a great compromise). The best thing is that you can develop on a sqlite database but deploy using mysql or postgresql with a minor change. Python and turbogears both emphasise automatic testing. Ideally the testing should cover all the code you write. The authors even suggest you write the test first then implement the feature. After chasing down several bugs, some of which I introduced fixing other bugs, automatic testing would make my life a lot easier and perhaps I wouldn t dread the release cycle so often.

Related articles

• I m beginning to hate PHP? (abdulapopoola.wordpress.com)
• pTest: PHP Unit Tester in 9 Lines Of Code (blogs.sitepoint.com)
• PHP to the Rescue! (blogs.sitepoint.com)

Network effects are such a pernicious thing. They give us the internet, but they also give us an internet full of monopolistic worse-is-better players that use network effects to stifle improvements and competition. Sometimes the best you can manage is to be a parasite. facebook So I'm back on facebook ... sort of. My strategy for facebook is to be as near to a featureless datapoint as I can. I won't post anything. I entered as bad quality information as I could get away with. I turned all privacy controls off, on the theory that this way I don't have to worry about facebook eroding privacy since everything about me on facebook is aggressively public. By never posting anything to facebook, I am trying to fight the network effect; trying to avoid increasing the value of facebook due to me being on it. I've adblocked the text entry field on facebook just to avoid the temptation. I want to erase the Like buttons too. Technologically, I think a good approach would be a browser plugin that does these things while extracting all available content and optionally sending it to a RECAP-type system. I only friend people who I need to keep up with and who have little or no other internet presence. This is because I am aware that each person I friend still propels facebook's network effect a little bit; even though they'll never see any useful information from me, they will know that they are able to talk to me via facebook. I only log into facebook once a week or so (this is easy when you're not interacting with it anyway), and read whatever sample of farmville and other posts facebook selects as "important" (worse than usual probably since I never give it feedback). So people I friend can't rely on me reading everything they do, which reduces the value a little more. I will probably delete this account occasionally, and then re-add it later. Just often enough to be really annoying. twitter Similarly, I no longer gate my identi.ca account to twitter, and have deleted as many tweets as I could stand to (one click and one keypress per tweet; no mass-delete; ugh!), and protected my account to avoid accidential new tweets leaking out. Since twitter is not quite as evil as facebook I feel slightly less bad about exposing more of my social graph there. Only the folks who are too lame to get on an OpenMicroBlogging site like identi.ca. phone Of course there was a network effect with the telephone as well, and while it got deregulated it's still pretty bad, and also a technology that I mostly dislike using. Oddly my phone behavior is backwards to the other two networks; I only make outgoing phone calls, never answering the phone (unless real life intervenes very severely). I listen to some random fraction of my voice mail and read whatever ones manage to get machine-transcribed.

When I installed my new laptop, I decided to switch from sleepd to GNOME Power Manager (GPM). The rationale behind this decision was to get screen dimming for free while keeping other features because GPM uses UPower which uses pm-suspend and pm-powersave behind the scene. This means that it is easy to execute scripts on suspend or when the AC is unplugged by putting them in /etc/pm. I don t use a full GNOME desktop. My desktop is running FVWM which is just a window manager. However, I already use some parts of the GNOME environment:

• GDM as a display manager because it is the only one that is able to handle ConsoleKit correctly. This means that GDM is able to register your session to ConsoleKit and make it active. Without that, you won t be considered as the user sitting in front of the screen and DBus won t allow you to interact with most of your hardware unless you add yourself to a lot of groups. ConsoleKit enables the at_console property used by most policies.
• Network Manager to handle both wired and wireless connections. Connecting to some random AP using only command line tools is rather cumbersome (assume that the AP is using WPA). Connecting to a broadband network with a 3G dongle is even more cumbersome.
• GNOME bluetooth applet because interacting with BlueZ bluetooth stack is now done with DBus only and the command line tools have been removed. Therefore, you need to write your own tools or use this small applet that do most of the work for you.

What s working It happens that most of the features just do not work. However, suspend when closing the lid works fine. The icon in the systray works fine too (but there is a small bug). The screen dimming feature on idle also works (but there are annoying bugs with it). The visual feedback when pressing brightness keys works too.

What s not working It seems that I am not the only one to have problem with GPM: there are currently 115 bugs opened in the Debian BTS. If you want to get some debug output from gnome-power-manager, run it with --verbose flag. --debug flag does nothing (and supersedes --verbose).

Nothing happens when pressing the power button I have configured GPM to ask me what to do when I press the power button. But when I press the button, nothing happens. Looking at the source, here is the snippet of code that is executed:

1 2 3 4 5 6 7

if (policy == GPM_ACTION_POLICY_INTERACTIVE) GpmSession *session; egg_debug ("logout, reason: %s", reason); session = gpm_session_new (); gpm_session_logout (session); g_object_unref (session);

In fact, this is not GPM that will ask you the question, but GNOME session manager because GPM triggers logout. If you don t run a full GNOME session, nothing will happen. My workaround is just to set the action to Suspend .

Does not sleep when idle GPM is configured to go to sleep after 2 hours on AC or 30 minutes on battery. This does not work for me because idle detection is delegated to external programs. There are four variables used by GPM to determine if we are idle:

• session_idle which is the idle hint from ConsoleKit,
• idle_inhibited which is set when another program asks to disable idle detection (a video player which does not want the screen to be dimmed),
• suspend_inhibited which means a program do not want we going to sleep (a music player for example)
• x_idle which tells if X is idle or not (someone is moving the mouse or typing on the keyboard).

To allow GPM to suspend when we are idle, x_idle and session_idle must be true while suspend_inhibited and idle_inhibited must be false. For screen dimming purpose, you just need to have x_idle set to true and idle_inhibited set to false. Currently, only GNOME Screensaver seems to tell ConsoleKit if the session is idle or not. XScreenSaver does not contain anything about this. I am using xautolock with xlock and I am pretty sure they won t tell GPM if I am idle. You can use dbus-send to circumvent this problem:

1 2 3 4 5 6 7 8 9 10 11

$ dbus-send --system --dest=org.freedesktop.ConsoleKit \ > --type=method_call --print-reply --reply-timeout=2000 \ > /org/freedesktop/ConsoleKit/Manager \ > org.freedesktop.ConsoleKit.Manager.GetCurrentSession method return sender=:1.6 -> dest=:1.106 reply_serial=2 object path "/org/freedesktop/ConsoleKit/Session1" $ dbus-send --system --dest=org.freedesktop.ConsoleKit \ > --type=method_call --print-reply --reply-timeout=2000 \ > /org/freedesktop/ConsoleKit/Session1 \ > org.freedesktop.ConsoleKit.Session.SetIdleHint boolean:true method return sender=:1.6 -> dest=:1.108 reply_serial=2

While I understand what is the rationale behind those choices (no code duplicate and allowing external programs to hinder power management), it lacks some fallback in case you don t use GNOME Screensaver (which is the key component here). I first though this was bug #594706 but this is not the case. I have opened (and closed) bug #628755. UPDATED: Jump to the conclusion for an updated solution for this problem.

Brightness is not restored to the previous value after idle When GPM undims the screen, it does not restore the previous value. It seems to be bug #617310 which is marked as fixed in upstream. I did not investigate it.

Wrong number of batteries reported I only have one battery but after one suspend/resume cycle, GPM thinks I have two. Maybe this is a bug of UPower. Maybe this is bug #606434. I did not investigate it. Maybe there is an happy interaction with bug #456280.

Inability to set configure actions when the UPS runs out of battery This is bug #390602. In addition to laptop batteries, UPower also supports UPS. In this case, GPM displays a special icon to tell if the UPS is fully charged. However, there is no possibility to execute special actions when the UPS is unplugged, like hibernate after 10 minutes or something like this. This seems to be a problem with the interface because looking with gconftool-2, we can spot variables like low_ups or critical_ups which allow to set actions when the UPS is almost out of battery. Moreover, the UPS icon only displays after one suspend/resume cycle but I am too lazy to report and debug this issue. Since, the icon disappear if I restart GPM, I am confident that the bug is in GPM.

Conclusion The conclusion is that GPM is not really that buggy but it really needs to be run in a full-blown GNOME desktop. The main problem is not really GPM. New technologies like UPower and ConsoleKit get pushed by freedesktop.org initiative, they get implemented by GNOME (and, I suppose, KDE) because they have tons of people following the appropriate mailing lists and any alternative just gets crushed by all those unpublicized changes. UPDATED: I got several helpful comments about this problem. There are two possible solutions:

1. Use xfce4-power-manager. It handles brightness control in a better way and don t need an external program to tell when the current session is idle. It still needs a session manager to ask what to do when using the power button. The drawback to this solution is that you start to pull another desktop environment into your environment. You had some parts of GNOME and now you have some parts of XFCE. This solution was proposed by Sylvain Collilieux.
2. Use gnome-session, gnome-screensaver and keep gnome-power-manager. As pointed by Julien Valroff, gnome-screensaver needs gnome-session to detect idleness. However, gnome-session starts a whole GNOME desktop. Again, there are two solutions. The first one is to replace gnome-session by something lighter implementing the bare minimum to make gnome-screensaver happy. Another suggestion is to configure gnome-session to start your favorite environment.

I start mine from ~/.xsession and here is what I have added at the top of it:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

# If not done, restart ourself with gnome-session if [ -z "$SESSION_MANAGER" ]; then [ -d ~/.config/autostart ] mkdir ~/.config/autostart rm -f ~/.config/autostart/* cat <<EOF > ~/.config/autostart/xsession.desktop [Desktop Entry] Name=Xsession Comment=Start .xsession Exec=$(readlink -f $0) Type=Application X-GNOME-Autostart-enabled=true EOF exec gnome-session --autostart=$HOME/.config/autostart fi # Otherwise, keep starting stuff # [...] exec fvwm-crystal

I know it has been a while, but after reading a blog post by Anand Kumria over at planet.debian.org I decided to have a quick look at one of the problems he described.

Basically, Anand wants to force the local resolver to be used for each and every network connection, may that connection be established manually or via NetworkManager. He wrote that fixing this configuration for every new connection manually is tedious, and I fully agree on that. So here is a solution to do this all automatically, using resolvconf:

After installing the resolvconf package every time /etc/resolv.conf is to be updated resolvconf takes care of that. Using the files in /etc/resolvconf this process can be controlled and the resulting file modified to fit one own's needs.

So at first we would like the local resolver to be used for every connection. This works by simply adding the "nameserver 127.0.0.1" directive to the /etc/resolvconf/resolv.conf.d/head file. Simple as that. Every time /etc/resolv.conf gets generated the contents of the head file are actually used as /etc/resolv.conf's header.

Using this method the local resolver is used for every connection. But Anand wanted to use only the local resolver and discard any resolvers possibly obtained via DHCP for example. Guess what, this is also possible using resolvconf.

Adding TRUNCATE_NAMESERVER_LIST_AFTER_127="yes" to /etc/default/resolvconf does exactly that. Now every nameserver directive after the 127.0.0.1 one is ignored and will not make it into /etc/resolv.conf. You can of course add more nameservers to the head file above the 127.0.0.1 directive.

Problem fixed I guess.
Don't forget to re-connect to the network or manually force re-creation of /etc/resolv.conf so the changes you made get populated. I really hope this is of use to some of you facing similar problems.

Two weeks ago, I was lucky to celebrate my 33th birthday with my closest friends in l'Alqueria. When asked to wish something before blowing the candles on Victor's delicious apple cake, I thought I have basically everything I'd want, but it'd be cool if some real changes happened in this world. Not long after, big demonstrations asking for Real Democracy Now happened throughout the Spanish state, and today, that Sunday seems to be an eternity away. Huge assemblies, thousands of strangers working together, more demonstrations, an election campaign eclipsed by #15m, hundreds of well thought, plausible claims published, the movement crossing the Spanish borders and leaking into France and Greece, the feeling that this is the good one, the basis for a fresh start that can make our lives better, our society a fair one and the possibility to stand in front of the fuckers who have made our lives a lot harder, to tell them it's not going to work like that anymore. All of this in 15 days. Unfortunately, revolution came when I'm in a crucial month to finish my studies and swamped with other little things, so I've been unable to be in the Valencian camp site for more than 3 days. Hopefully when I'm done on the 22nd people will still be taking the square, because Mako and Mika will be visiting then. Yay!