kloeri

I'll be going to Saint Petersburg this tuesday and plan to stay there a month or so. I'll be busy working on a commercial project while over there but there should be plenty of time to meet some local residents as well and enjoy the city.

I'm not sure what my schedule is going to be like but if you're in the area please give me a shout and we'll see if we can meet up for a beer or whatever. The best way to contact me after monday is probably by email at bryan.ostergaard@gmail.com but I hope to be present on irc as well.

And to celebrate the occasion I'll be looking back over the past year, recounting some of our many successes and also given a glimpse into the future - at least the way I see Exherbo's future.

But first I'd like to thank all the developers and users contributing in various ways to Exherbo. According to https://www.ohloh.net/p/exherbo there've been 52 contributors so far but that's leaving out people contributing to Exherbo related repositories that Ohloh doesn't know about or contributing in ways not involving commits. My guess is that we have had 60+ committers during this first year which is very good indeed.

A big thank you to all of you - Exherbo wouldn't have been anywhere near as usable without your continued commitment.

State of Exherbo
At this point I consider Exherbo very usable and quite stable. There're still major changes happening from time to time but usually the upgrade path can be easily explained in a few lines on the exherbo-dev mailing list.

As for packages we have supported KDE, Gnome, XFCE and Awesome on the desktop for a long time now. On the server side we have most of the usual suspects as well including the apache and lighttpd webservers, samba, exim, postfix, sendmail and so on.

Many people are likely still missing a couple of packages but that's easily solved using importare, writing your own exheres package or requesting it in the #exherbo IRC channel.

Many people have also started to test Exherbo after we started publishing Exherbo images for virtual machines. Just recently it became possible to easily build your own Exherbo images from scratch which will hopefully lead to lots of new ideas for Exherbo and make it easier to mold Exherbo to specific needs.

A year of accomplishments
There's been too many interesting things happening around Exherbo this past year to name them all but here's a mostly chronological list of major events.  All these events have helped shape Exherbo one way or another.

June 7th 2007
Stephen Bennett sets up the exherbo-dev mailing list. Everything keeps happening on IRC.

August 5th 2007
The old goatoo repository is killed and everything is moved to the new arbor and exherbo repositories. We still live in the dark subversion age.

October 13th 2007
Importare is born, makes life much easier as we have very few packages at this point in time. As described on http://ciaranm.wordpress.com/2008/05/20/managing-unpackaged-packages-or-whats-this-importare-thing/ importare is a paludis client allowing proper package installs, uninstalls and upgrades without an exheres.  At a point where we still had very few packages beyond what's required for a base system install this had a big impact on Exherbo. Importare is as important today as it was a year ago as it allow to concentrate on widely used packages instead of spending time on more obscure packages.

July 5th 2007
Support for the Exheres format is added to Paludis. Officially it's described as a test EAPI used to play around with new ideas that might not be suited for Gentoo.

July 24th 2007
We solve the problem with colliding source tarball names by introducing arrows. This allows us to rename distribution files on mirrors and locally to include package versions for example.

December 7th 2007
We add a commits mailing list. This is a big help for reviewing commits and lots of bugs are caught this way.

Early January 2008
Our mascot Zebrapig is born.

January 31th 2008
We add src_prepare and src_configure phases to exheres-0. For many packages this helps us write much cleaner packages as it matches the stages of the build process much better than just having one big src_compile phase.

March 14th 2008
First draft of Exheres-for-smarties is committed. Exheres-for-smarties becomes our main technical document on the Exheres format and repository structure.

March 15th 2008
We add :* and := support to specify slot dependencies more precisely.

May 2008
We gain a new, much better default src_install implementation which was later followed up by revamping pretty much every default function as well as the various helper functions. We also switched from subversion to git and had the frst archived discussion of replacing categories - this is still a frequently discussed topic.

May 18th 2008
Announcing Exherbo on my blog
It took only an hour or two from my announcement being published to it hitting Slashdot, Digg and The Register to name but a few. The next several weeks was spend answering tons of questions and trying to resolve the worst misunderstandings.

May 23th 2008
We got tired of answering the same questions over and over so Ciaran wrote a quick install guide on http://ciaranm.wordpress.com/2008/05/23/installing-and-configuring-exherbo/.  This is of historical interest only but it was important at the time as it allowed us to get back to development for the most part. It's also interesting as a fairly accurate description of the state of Exherbo back then.

June 4th 2008
FOSS Aalborg takes place and I open with a talk describing the main ideas behind Exherbo, some of the bigger issues we want to solve and why I chose to start a new Linux distribution instead of joining an existing distribution.  Much interest shown and it was quite encouraging for myself to present my ideas before a large crowd of technical people. The video of my talk is still available at http://mirrors.dotsrc.org/blivklogere/foss_aalborg/2008/a_Linux_distribution_is_born--Bryan_Oestergaard--english--FOSS_Aalborg.mp4.

June 12th 2008
We add UnavailableRepository to Paludis and get a much better grip on the expanding number of package repositories. The script we use to build the package indexes for all the repositories hits Gentoo hard and we had to fiddle a bit with the updates before everybody was happy :)

August 17th 2008
My first "Exherbo goals" mail. This has become a series of mails where I describe the state of all the different ideas and features we're working on.

August 27th 2008
KDE 4.1.0 has landed! This marks the beginning of Exherbos KDE support and one of the more important milestones for desktop systems.

September 17th 2008
Markus Rothe announces his first PPC64 stage tarball. Markus ported Exherbo to PPC64 in fairly short time and is one of our many frequent contributors.

October 2nd 2008
Exherbo-cn, one of the early user managed repositories starts. It shows the strength of our distributed repository model by providing packages for Chinese support (fonts, input methods and so on). Exherbo-cn continues to be very active and one of the stronger parts of the community surrounding Exherbo.

October 4th 2008
The second day of the danish open source conference Open Source Days take place and I give a talk on my favorite subject - how we're rethinking Linux distributions and what it means to both developers and users. Unfortunately there's no video available of this talk.

Besides the talk we also had a fairly successful booth with plenty of visitors throughout the day. All in all a very good experience that I hope to repeat this year.

October 6th 2008
We add Unwritten repository support to Paludis and move all package requests from Bugzilla to unwritten so we can query them using paludis just like other packages.

January 26th 2009
Just in time for FOSDEM Ciaran adds AccountsRepository support to Paludis.  Packages can now depend on users and groups just like they would depend on various libraries. We quickly proceed to kill enewuser and enewgroup usage.

February 7th 2009
I was invited to FOSDEM as a maintrack speaker and had a blast! I gave a talk on '10 cool things about Exherbo' where I presented some of the cool things we've done to improve the user and developer experience. The rest of the weekend I was constantly approached by people wanting to know more about Exherbo and it was definitely my best FOSDEM experience so far. Video from my talk is available at http://video.fosdem.org/2009/maintracks/.

February 11th 2009
I reorganised our website and changed the build infrastructure to make it easier to maintain. The new website makes it much easier to find needed information and just as importantly it makes it quite easy to contribute updates and new content.

February 14th 2009
First mention that I can find of Sydbox, our future sandbox implementation written by Ali Polatel.

February 12th 2009
We add parametrised exlibs. This is quickly used to specify supported autotools versions, perl module authors and a host of other things making many exlibs much cleaner.

February 15th 2009
We add src_test_slow() phase for those packages that takes a ridiculous time to run their testsuites, often measured in hours. Users can control this with a build_option.

March 2nd 2009
Jonathan Dahan grabbed the chance and wrote an install guide for Exherbo as well as a short FAQ. This is the first major piece of user contributed documentation to the website.

March 3rd 2009
First virtual machine images are published and becomes quite popular. The images are all built manually which convinces me to start writing a script to build them.

March 19th 2009
We replace versionator by internal functions. This way we can take advantage of Paludis own version comparison primitives instead of trying to keep a bash script in sync.

April 14th 2009
First release of Sydbox by Ali Polatel. Sydbox is intended to replace Gentoo's sandbox implementation in Exherbo and should fix most if not all the shortcomings of the existing sandbox implementation. This is an important example of core code being contributed to Exherbo from a user and shows that there's really no distance between users and developers in Exherbo.

May 10th 2009
I published my script for automatic KVM image creation. Several bugfixes and general clean up of the script is offered over the next few days.

Future expectations
Exherbo differs greatly from most other distributions and we don't really follow the normal pattern for distribution development. We have no release schedule for example - in fact we don't have any plans of a release at all!

That is not to say that a proper release might not happen but we'd need convincing arguments why a release is necessary before spending lots of time on it. So what do we do when we're not building new releases?

Small improvements
Well, most of our time is spend on what I consider small improvements. Much of the above list describe such improvements. Looked at individually they're interesting but rarely earth shattering. Based on this my predictions of what's to come is also going to be mostly about small but important things with a few bigger things thrown in as well.

Stable Exheres format
One of the most obvious things are the ever evolving Exheres format. At some point we're going to define our first stable format exheres-1 and convert our repositories to that. Before that happens I'd like to see proper support for binary packages and multi-ABI though.

Binary packages already works but we need to fix various problems before using it more officially. For multi-ABI we have the design more or less pinned down but there's some pretty annoying implementation issues that we need to work out.

Build infrastructure
We already have some parts of the infrastructure needed to build various Exherbo blobs like KVM images for example but lots more is needed.

Right now we can build KVM images for x86 and amd64 in a fairly inflexible manner. We need to expand current scripts and write new scripts allowing us to automatically build binary packages, several different kinds of image files, flexible configuration of partitions and file systems. And while at it we need to expand all this to be able to build images for CDs and USB sticks as well.

The build infrastructure should also be able to easily build customised images and be used for more or less unrelated purposes such as tinderboxing.

New init system
This is one of the more mysterious Exherbo projects but also one of the things that I'm most excited about personally. I've talked about it in public on several occasions so many of the basic ideas are already known. That said it's changed direction quite a bit and should be even more interesting when it's finally published.

Easier management of our distributed repository model
Our model of many Topic Repositories and Developer Repositories works fairly well as is but there's no doubt it can be improved further. Currently we want to implement a "repository of repositories" so you can install new repositories using paludis just like you install packages. As we continue to grow and refine our model I'm sure we're going to focus even more on this area and I'm looking forward to seeing what exciting ideas we're going to come up with in the future.

Better documentation
This is one area that haven't got a lot of focus so far. Our documentation mostly consists of Exheres-for-smarties and of course the paludis documentation. Lots of other areas needs to be documented and I'm hoping some users will step up to help with this important task.

Growing user community
This one seems obvious at first but a large part of our users come to Exherbo because of the flexibility of the distribution and our strong focus on technical design of new features as well as the rapid development happening.  This also means that many of our users actively participate in the development which is something I'm hoping to strengthen further as we go along.

It keeps the community very much alive and we seem quite capable of keeping the focus and direction of Exherbo despite having twice as many users contributing in the past year as there are official Exherbo developers.

New profiles
Our current profiles aren't very flexible or useful. We have some vague idea of "mix-ins" allowing us to "mix" several different profiles like an amd64 profile + a KDE profile for example. The idea is fairly vague at this point but at some point we'll get much more flexible profiles allowing for easier maintenance and use.

The great unknown
And perhaps the most exciting part of the future is the part that we can't foresee at all. The development rate have only increased since announcing Exherbo and we often get ideas from unexpected sources. Some of these ideas don't fit in very well with Exherbo and are quickly discarded but many ideas are used in one way or another. Usually that requires some molding to make the idea fit the rest of Exherbo properly which in turn might lead to new ideas.

This process of constantly exploring new ideas helps keep Exherbo at the forefront and definitely keeps it a fun project to work on.

Thank you all for being part of this project - Exherbo might be my baby but you're all helping it grow up and shaping it into something very exciting.

I've uploaded new KVM images based on the 20090504 stages a day or two ago. The images are available at http://dev.exherbo.org/images/ as usual.

But more interestingly I've now made the script used to build the images available so you can build new images yourself whenever you like. The script is available in the scripts/ directory of the exherbo repository.

All you need to build your own image is a few prerequisites and this script. The script requires kvm (for kvm-image) and parted (used to manipulate the partition table) and sfdisk (used to get some partition table information) installed.

"paludis --install kvm parted util-linux" will ensure you have all the needed prerequisites. After installing those all you need is to specify a few options to the script and everything should be automatic from there on.

The script describes the available options and their defaults when passed -h or --help.

# ./create-kvm-image --help
Usage: create-kvm-image [OPTIONS]
Options:
    --arch=amd64|x86                Target architecture for image file
    --kernelversion=            Kernel version to be used in image
    --stageversion=             Date of tarball, for example 20090504 or current
    --kvmtmpdir=/path/to/image      Where to build the image file. Defaults to /tmp/kvm-tmp/
    --kvmtmpkernel=/path/to/kernel  Where to build the kernel. Defaults to /rootfs
    --kvmimagename=/path/to/image   Image filename (including path). Defaults to /exherbo-x86_64.img
    --kvmimagesize=          Size of image file in gigabytes. Defaults to 6G
    --jobs=                  Number of make jobs when building the kerne. Defaults to 4

If you're satisfied with the defaults all you need to specify are kernel version, stage tarball version and architecture. Which gives you a command like ./create-kvm-image --kernelversion=2.6.29.2 --stageversion=20090504 --arch=amd64

And a few minutes later (takes about 5 minutes on my quad-core Core2 box) you'll have a brand new KVM image called exherbo-x86_64.img in /tmp/kvm-tmp/. Please note that we don't support cross compiling yet so you'll have to specify the same target architecture as your host architecture for now.

And as always, I welcome git format-patches to add support for other image types (virtualbox, vmware, ..) and other features.

To avoid name clashes and silly names we're adding a new set of rules for naming repositories. The rules affect profiles/repo_name and not the actual sync url which can differ if neccessary. I'd recommend using the same name however.

The new rules for repo_name is as follows:
- Official topic repositories uses the topic as name.
- Personal repositories uses use the owners (nick)name
- Personal topic repositories use owners (nick)name-topic

So the official KDE repository is named 'kde' and you can find all it's packages using for example 'paludis --list-packages --repository kde'. Ingmars personal repository is named 'ingmar' and if he had a personal topic repository for office type packages it would be named 'ingmar-office'.

I hope the new rules will make the status of repositories easier to understand.

Just a quick note that I left all gentoo IRC channels for good as a few gentoo developers are always either putting words in my mouth or attacking me in silly ways when I try to participate in technical discussions. I have absolutely no inclination of getting dragged down to that level so I've simply left all the channels now.

If anybody needs my help I'm sure you know where to find me but please consider carefully if it has any relation to gentoo before contacting me and DON'T contact me if the answer is yes. Thanks for your consideration.

In one of my recent blog posts a user commented that it would be great if we added our documentation to a git repository as that would make it much easier to contribute.  In fact all of our website has always been in git but it hasn't always been public and we haven't been very good at describing the setup so I'll try to describe our current setup now.

Our setup
Our website is spread over two different repositories.

The first repository is called www.git and contains what be considered the "main" website - that is our frontpage, the list of developers and links to various resources and the available documentation. The www repository isn't particularly interesting and doesn't get a lot of updates.

The second repository is docs.git which all the real documentation available through the website such as Exheres-for-smarties, our FAQ and installation guide. This repository gets frequent commits as we update Exheres-for-smarties to reflect changes in the exheres-0 EAPI and correct or clarify existing documentation. This is the repository most patches should be made against.

You can browse through both repositories using gitweb or clone the repositories using git clone git://git.exherbo.org/docs.git and git clone git://git.exherbo.org/www.git. Looking through the contents you should notice two things:

1. We use Markdown for all our documentation. Markdown is a very easy format to describe paragraphs, headers, lists and so on in a format very close to normal text files.
2. HTML generation is driven by Makefiles. These makefiles are run from a git hook and automatically generates the HTML code used on our website every time one of us pushes new content to the repositories. We only have one webserver at the moment so there's no syncing between nodes or any other complexities involved in this process.

The Makefiles is also worth looking at if you don't have much experience with make as they use a couple tricks to generate the needed rules based on wildcards.

Contributing documentation
If you want to help fix current documentation (typos, badly worded text and so on) you should follow the same instructions as if you were contributing a package. There's a few differences of course as you can ignore all the exheres specific stuff but the git configuration should follow that guide and we're going to want a git format-patch that we can apply directly to our repositories.

For larger contributions you'll likely want to talk to me first (kloeri on irc.freenode.net or kloeri@exherbo.org) to make sure that we want to add your document(s) to the website and that we agree on the contents of it before spending a lot of time on writing. We're quite open to contributions but we also have fairly specific ideas on the direction we want for Exherbo now and the near future so it's important that we coordinate larger updates to the documentation and/or website.

That said, we could use an update for Contributing to cover documentation as well - please let me know if you're interested in helping with this and I'll try to coordinate it.

Update:
I've just fixed the www.git Makefile to use maruku like docs.git always have. So you should check formatting using the dev-ruby/maruku package.

We've had several requests for disk images for kvm, vmware and virtualbox over the past few months. This request was repeated at FOSDEM so I finally took some time out to make it happen and I hope that you'll enjoy it.

The images allow you to easily test Exherbo without touching your current system. In case you're using some other virtualization product you'll probably be able to use the kvm images as those are raw disk images.

Keep in mind though that we don't provide any official support if you try Exherbo so you're expected to resolve problems yourself.

In one of my recent blog posts I asked for volunteers to help with our documentation needs. Jonathan Dahan kindly offered his help and wrote a short FAQ, a document on some of our features as well as a basic installation guide.

You can find all the new documentation here.

I'm still looking for help with documentation, be it new documentation or updates to existing documentation so please contact me if you have any ideas in this regard.

The video of my FOSDEM talk (10 cool things about Exherbo) is online now along with all the other great FOSDEM talks. I've already had quite a bit of positive feedback about my talk but don't miss the other videos either - there was quite a lot of good talks this year and I'll have to catch up on a few videos myself.

Just before FOSDEM the Exherbo website got a much needed update. The old website was mostly untouched from our announcement almost a year ago and many things have changed since then.

So what's new?
- We've reorganised the content to make it easier to find downloads, our developer documentation and various other resources.
- It's friendlier - we've added tons of packages during the last year and we're hardly ever breaking Exherbo completely anymore so many of the old warnings don't apply anymore. We're still not promising any support though so you need to have a good grasp of how linux works if you want to try Exherbo.
- Much nicer buildsystem for the website. The old website was a conglomeration of make, sed, cat and files in at least 4 different formats spread across two git repositories making it a bit unwieldy. The new website is built using just make and markdown for all the content. We still have content spread across two repositories but we've removed some of the weird interdependencies between them.

What's in the future?
I expect that we'll start adding more documentation in the coming months and that there might be a bit more reorganisation of the content as the website grows. If you have a fair grasp of how Exherbo works and good english writing skills this is an area where we could use some help.

Please contact me (kloeri on #exherbo / irc.freenode.net or kloeri@exherbo.org) if you think anything is wrong with the new website or you want to contribute some material.

Just a quick note that I was interviewed prior to FOSDEM. The interview is available at FOSDEM interview.

FOSDEM is only 3 weeks away now and as usual it's promising to be a great experience. It's by far the biggest Open Source conference in Europe. I've been to FOSDEM for the past 3 years and it's been a great experience every year. This year the staff have (as usual) outdone themselves with a torrent of talks, stands, developer rooms and non-technical events such as the traditional beer event and a new offer of guided tours for spouses/partners.

But this year is even more exciting for me as I've been invited to give a talk about Exherbo and some of the interesting things we're doing in our small part of the world. The talk is called 10 cool things about Exherbo and I hope to see all of you there.

Tomorrow is Software Freedom Day - a yearly event where people all over the world get together to celebrate free software, enjoy talks related to free software and just as importantly get to meet lots of people.

If you happen to be in Copenhagen tomorrow you can meet myself and several other people from SSLUG at Copenhagen Business school. SSLUG's SFD program includes talks on Free Software, Linux, Open Office and GIMP. Everybody else can look up their local Software Freedom Day events - there's more than 500 teams registered all over the world so there's probably going to be an event nearby.

And regarding Planet Larry.. Steve Dibb just announced that he's setting up a feed for retired Gentoo developers which is very good news in my opinion. Lots of retired developers blog and they often have interesting comments on things related to Gentoo or tips that other people can benefit from. And this way people can know whether the blog posts they're reading comes from a normal user or a retired developer. I would probably have prefered marking retired developers another way instead of having multiple feeds but I can see why some people wants to know who's who and I'd much rather have a seperate feed than nothing at all. Oops, I was a bit too quick - Exdevs are now going in the main feed instead and will be marked using colour or some other way instead of a seperate feed.

And since I've been having this discussion with Steve on and off for quite some time: Thank you Steve :)

That's right! Exherbo will be at the Open Source Days conference october 3-4 and ticket sale opened a few days ago.

Open Source Days expects 2000+ visitors and have lots of talks ranging from open source project management to more technical networking talks as well as talks large scale open source deployment. I'll be talking about the Exherbo project of course - how it all began, the current ideas defining Exherbo and some of the technical features support those ideas. Besides my talk Exherbo will also have a booth where you can meet several of our developers.

I hope to meet lots of people at OSD and get to discuss different aspects of Exherbo - it's an exciting project with lots of interesting features and above all room for innovative new features and improvements.

All open source developers get for their hard work most of the time is credit - other people knowing they did some hard work and sometimes even other people acknowledging and thanking them for it.

Anybody who's been doing open source development for a few years knows this which is why I'm extremely surprised to see Diego "flameeyes" Pettenò trying to hide other developers work.

In one of his many blog posts Diego talks about ABI dependencies and how that's important for cross compiling among other things. As it turns out I agree this is important and more to the point, David Leverton (a Paludis and Exherbo developer) pointed out quite a bit of prior work being done on this exact problem in Exherbo.

David did so in the following comment:


This comment is nothing but helpful and friendly even if Diego might not want to talk to spb as they have some history - the comment still points directly at prior work being done in this exact area that Gentoo could easily benefit from and hiding it (by not allowing the comment) is dishonest at best.

I hope Diego is going to be a lot more cooperative in the future and actually follow the open source spirit instead of trying to mimic behaviour that's primarily seen from closed source companies.

Just a quick note that I wrote up a long list of things Exherbo still needs on http://lists.exherbo.org/pipermail/exherbo-dev/2008-August/000254.html.

Reading that list and any follow-up discussion is probably a good way to follow Exherbo development for the foreseeable future and might give some insights into what goes on behind the doors for those of you with no real experience in distribution development.

Some of you might remember me doing a presentation on Exherbo and primarily our planned initsystem a few months ago at the FOSS Aalborg conference in Denmark.

Most of the videos are now available including the one of my presentation. The presentation isn't highly technical but I talk a lot about why I don't think other initsystems solve the problems very well as well as what I expect from an initsystem. The video should be fairly interesting to all the people who wasn't at the FOSS Aalborg conference.

Hope you like the presentation(s) and thanks to the FOSS Aalborg organisers for getting the videos online and for having a great conference.

This is shiny but pretty much useless if you ignore the geek factor.. Not one to ignore the geek factor I've made a quick movie of arbor commits :)

There's movies of the Apache, Postgres, Eclipse and Python projects available on the authors homepage.

The last couple of days I've seen a bunch of links to 'Never Heard of Exherbo? Never Will, Either.' pop up all over the internet and people have asked me about it on freenode as well.

The short answer is that yes, I've seen it and I've even made a couple of comments on that website. (Their comment system is horrible by the way and removes any line breaks making longer comments a big mess). But since I keep getting comments about it I'd like to clear up a few misconceptions from that article and some of the comments.

• Exherbo is not intended as a user distribution
• Exherbo is not currently intended to be used by users but we're only concentrating about the current status. Saying that Exherbo is not intended as a user distribution completely misses the fact that we only have a very small part of the distribution at this point and clearly can't start to support users in this stage of development.
• We play only with our friends
• That's not really true either. We state fairly clearly that we don't want users at this point in time as it would be a waste of everybodys time but quite a few developers from different distributions have shown interest in what we're doing and are actively following our development to see if there's anything they might benefit from. A good example of this is Gentoo bug 33544 where Gentoo apparently wants to copy some of our code for a future EAPI. Making our code available to other developers is a basic part of what open source is all about and we're doing so even if we're not yet ready for users.
  
• Usability, error checking and formatting comes from user requests
• A huge part of the user requests I've seen during the years as an open source developer haven't been for any of those things - quite the opposite in fact. Many users keep requesting that we (open source developers that is) skip tests, ignore errors and so forth so silly things like that doesn't get in the way of installing applications. Never mind that the exact same users later complain that the applications doesn't work (big surprise there..). We definitely don't want any of that advice even if users believes it's good advice. We're well aware of how to build reliable software that's easy to use and make a big point of that already.

It's really not that we don't want users or developers - it's just that the time for that hasn't come yet and I wish people writing articles and commentaries about Exherbo would understand that or at least ask us about our intentions before spreading these misconceptions any further. Hopefully I've made it a little more clear now.

I got back from the FOSS Aalborg conference yesterday evening after spending a couple of days with several friends from the Exherbo project and enjoying the conference immensely. The talks were interesting and the speakers were quite good as well. And for my own talk about Genesis I got lots of positive feedback throughout the day and people seemed to like my ideas quite a bit. Slides for my presentation is available now and the video will be available on the FOSS Aalborg website as soon as the video editing and conversion is finished. The slides are probably not very interesting without the video however.

I found Martin von Haller Grønbæks talk about the legal aspects of reverse engineering quite interesting as well as that's a fairly important issue to software developers in general. He was talking about danish law that's derived from the EU directives so it should (mostly) be the same in the rest of the european union as I understand it. His points can be summed up like: "reverse engineering is legal if it's used to improve interoperability and it's needed due to lack of documentation" (my words, don't blame Martin for oversimplifying things this much). He kept mentioning interoperability frequently throughout his talk which made me think of companies using reverse engineering and whether they did it to support interoperability or not.

This of course lead me to ask how much you can stretch the interoperability paradigm as antivirus companies are among the few types of companies using reverse engineering in their day to day operations. This question (while very serious as I like to tinker with reversing malicious code myself from time to time) made everybody laugh and Martins very clear answer was that it's not legal to use reverse engineering for anti-virus purposes as that's not an interoperability issue (more like the opposite I guess). This is an interesting issue for many people and/or companies and it might be worth contacting a lawyer if you think you can possibly be hit by this as Martins answer is definitely not to be taken as legal advice - I'm sure there can be all kinds of details making it less black and white than the very general situation (completely void of any details) that I sketched in my question.