Hello Planet Debian reader.

And now you re thinking, who is this crazy guy who does not even know how to write English (is my sentence wrong?). My name is Guillaume Mazoyer and I am the Google Summer of Code student working on Jigsaw with Tom Marble and Sylvestre Ledru. I blog from France and I plan to write my GSoC reports right here.

Maybe you don t know what is Jigsaw yet. If so, go read this wiki page, you should be able to understand the goal of my project. Here is the description:

Since the beginning of Java, the JVM is a big virtual machine will all the features incorporated in it. This JVM provides many features which are useful in only limited areas (Corba) or which are not used at all in some contexts (AWT/Swing are useless for a web server).

The goal of the jigsaw project is to provide a modular JVM.

This will provide many improvements in the performance area (startup, solving of the dependencies, etc) but also the size of the dependencies.

There are already some works done by upstream (Oracle) but it does not complain to the Debian standards. The target is Java 8.

The goal is this Google Summer of Code project is to package the current development version of Jigsaw into Debian.

Please note that this project is not only about packaging. This will probably implies some developments and contributions to upstream (upstream is aware of this effort). Updating javahelper to manage specifities of Jigsaw could be considered.

I m pretty excited to work on this project and the work has already started. Can t wait to go deeper and enjoy working on two things that I love: Debian and Java.

John W. Dean wrote in insightful series of three articles for Findlaw about Authoritarian Conservatives [1]. In summary there are Authoritarian Followers who follow their leader blindly and Authoritarian Leaders who do whatever it takes to gain and maintain power. The Authoritarian mindset lends itself towards right-wing politics. Mick Ebeling gave an inspiring TED talk about his work developing a system to produce art that is controlled by eye movements [2]. The development work was started to support the quadriplegic graffiti artist TEMPT1. Mick s most noteworthy point is that all the hardware design and software are free so anyone can implement it without asking an insurance company or hospital (this is one of the few occasions when a TED speaker has received a standing ovation during a talk). The Eyewriter.org site has the designs and source which is licensed under the GPL [3]. Morgan Spurlock (who is famous for Supersize Me ) gave an amusing TED talk titled The Greatest TED Talk Ever Sold [4]. He provides some interesting information about the brand sponsorship process and his new movie The Greatest Movie Ever Sold . Ralph Langner gave an interesting TED talk about reverse-engineering the Stuxnet worm and discovering that it was targetted at the Iranian nuclear program [5]. The fact that the Stuxnet environment could be turned to other uses such as disrupting power plants is a great concern, particularly as it has special code to prevent automatic safety systems from activating. Angela Belcher gave an interesting TED talk about using nature to grow batteries [6]. She is evolving and engineering viruses to manufacture parts of batteries and assemble them, the aim is to scale up the process to manufacture batteries for the Prius and other large devices at room temperature with no toxic materials. She is also working on biological methods of splitting water into hydrogen and oxygen which has the obvious potential for fuel-cell power and also solar PV cells. As an aside she mentions giving a copy of the Periodic Table to Barack Obama and he told her that he will look at it periodically . Bruce Schnier gave a good overview of the issues related to human perceptions of security in his TED talk about The Security Mirage [7]. There isn t much new in that for people who have been doing computer work but it s good to have an overview of lots of issues. TED has an interesting interview with Gerry Douglas about his work developing touch-screen computer systems for processing medical data in Malawi [8]. This is worth reading by everyone who is involved in software design, many of the things that he has done go against traditional design methods. Mike Matas gave an interesting demo at TED of the first proper digital book [9]. The book is by Al Gore and is run on the iPad/iPhone platform (hopefully they will have an Android version soon). His company is in the business of licensing software for creating digital books. The demonstration featured a mixture of pictures, video, audio, and maps with the pinch interface to move them around. Dr Sommers of Tufts University wrote an interesting post for Psychology Today titled Why it s Never About Race [10]. It seems that there are lots of patterns of people being treated differently on the basis of race but for every specific case no-one wants to believe that racial bias was involved. The Register has an amusing article about what might have happened if Kate had left Prince William at the altar [11]. Fiorenzo Omenetto gave an interesting TED talk about synthetic silk [12]. He is working on developing artificial fibers and solids based on the same proteins as silk which can be used for storing information (DVDs and holograms), medical implants (which can be re-absorbed into the body and which don t trigger an immune response), and cups among other things. Maybe my next tie will have a no pupae were harmed in the production notice. ;) The CDC has released a guide to preparing for a Zombie apocalypse [13], while it s unlikely that Zombies will attack, the same suggestions will help people prepare for the other medical emergencies that involve the CDC. Salon has an interesting article by Glenn Greenwald who interviewed Benjamin Ferencz about aggressive warfare [14]. Benjamin was a prosecutor for war crimes at Nuremberg after WW2 and compares the US actions since 9-11 with what was deemed to be illegal by the standards of WW2. Eli Pariser gave an interesting TED talk about Online Filter Bubbles [15]. He claims that services such as Facebook and Google should give more of a mixture of results rather than targetting for what people want. The problem with this idea is that presenting links that someone doesn t want to click doesn t do any good. It s not as if the filter bubble effect relies on modern media or can be easily solved. Terry Moore gave a TED talk about how to tie shoelaces [16]. Basically he advocates using a doubly-slipped Reef Knot instead of a doubly-slipped Granny Knot. Now I just need to figure out how to tie a doubly-slipped Reef Knot quickly and reliably. Terry uses this as a mathaphor for other ways in which one might habitually do something in a non-optimal way.

• [1] http://writ.news.findlaw.com/dean/20070925.html
• [2] http://www.ted.com/talks/mick_ebeling_the_invention_that_unlocked_a_locked_in_artist.html
• [3] http://www.eyewriter.org/developer/
• [4] http://www.ted.com/talks/morgan_spurlock_the_greatest_ted_talk_ever_sold.html
• [5] http://www.ted.com/talks/ralph_langner_cracking_stuxnet_a_21st_century_cyberweapon.html
• [6] http://www.ted.com/talks/angela_belcher_using_nature_to_grow_batteries.html
• [7] http://www.ted.com/talks/bruce_schneier.html
• [8] http://blog.ted.com/2011/04/29/fellows-friday-with-gerry-douglas/
• [9] http://www.ted.com/talks/mike_matas.html
• [10] http://www.psychologytoday.com/collections/201104/happy-birth-certificate-mr-president/why-its-never-about-race
• [11] http://www.theregister.co.uk/2011/04/28/operation_pumpkin/
• [12] http://www.ted.com/talks/fiorenzo_omenetto_silk_the_ancient_material_of_the_future.html
• [13] http://emergency.cdc.gov/socialmedia/zombies_blog.asp
• [14] http://www.salon.com/news/osama_bin_laden/index.html?story=/opinion/greenwald/2011/05/13/nuremberg
• [15] http://www.ted.com/talks/lang/eng/eli_pariser_beware_online_filter_bubbles.html
• [16] http://www.ted.com/talks/terry_moore_how_to_tie_your_shoes.html

Hello everyone, this is (not actually) my first post to the Planet Debian. There was some older posts I filed under this category, sorry. My name is Aron Xu, and I ve just passed my NM process so I m now a Debian Developer who has his shiny new account to work on more Debian stuff. During [...]

I have to say busybox is so cool! Today I did an installation of Debian Squeeze using daily netinst ISO image in an i686 qemu-kvm virtual machine. I was using ext4 for /boot, and btrfs for /. The installation process was quite normal as expected. After the installation I rebooted the virtual machine. Grub2 loaded [...]

Recently I read Alexandre Rosenfeld s blog post about low performance of btrfs while placing VM s disk on it, so I had a try and got almost the same result. I tried to have someone look into the problem, and now get the answer, quoting Bastian Blank : This is a result of the filesystem design, [...]

Today I received a mail asking me to give him some advice on how to become a Debian Maintainer and Debian Developer, I am glad to hear more people getting interested on joining Debian, :-) The 1st thing to keep in mind is that, involving Debian s development is easy, but becoming something _official_ requires a [...]

Debian Maintainer Debian Debian Maintainer (DM) Debian Debian Developer (DD) Debian 1. Debian 2. Debian New Maintainer s Guide Debian Debian 3. WNPP (Work-needing and Prospective Packages) Bug 4. Debian Developer DD sponsor DD Debian sponsor Debian [...]

Hello everyone. Per tradition, I will use my first post to Planet Debian as a bit of an introduction. My name is Nathan Handler. I am Ubuntu Developer and a member of freenode staff. Currently, I am participating as a student in the Google Summer of Code program with Debian. I am working on Backend Tools and Infrastructure for DEX (aka dextools ) with Matt Zimmerman as my mentor. To quote the project description:

DEX is a new program designed to help improve Debian and its derivatives by merging in changes made downstream and encouraging discussions between the various projects. As this is a new project, most of the infrastructure does not exist (or is rather hackish and incomplete). This project will create the necessary backend tools and infrastructure so that all Debian derivatives can easily make use of the DEX project.

While I have not been great with updating my blog in the past, I am going to try and produce regular posts about my Summer of Code work. If you have any questions about DEX or are interested in getting involved, please check out the Get Involved section on the website.

The Alioth admins have been hard at work over the past few days to migrate Alioth to a new configuration split over a pair of servers. As a consequence, personal git repositories aren t working anymore for the time being. So, don t panic, the git repository will be back soon, one way or another.

What do Assholes do? Rusty just wrote a post titled If you didn t run code written by assholes, your machine wouldn t boot [1] about some of the anti-social tendencies demonstrated by programmers, including some that are implied to be fairly important. His post seems to imply that there are really great programmers who are anti-social and that we should just put up with it because of their great code. One of the problems with his post is that he doesn t define asshole . Holocaust deniers and all other Nazi supporters seem to clearly by assholes by any definition. People who have silly ideas about medicine and tell others seem to be merely misguided (although my dictionary gives stupid or irritating or ridiculous in the definition of asshole so technically they can meet the criteria). In the comments Rusty states that nuttiness is relative . While that is correct it doesn t seem to have much bearing on whether someone is an asshole. For example I know some very nice people who are utterly convinced by creationism. Is Anyone Essential to Free Software? Most projects have one person or several people in leadership positions, usually there seems to be a combination of project management and lead-programmer positions. Such people are obviously quite important to a project. But we have seen examples of people dying, being employed by Microsoft, retiring, relicensing the source in a bad way, and just losing interest without the project ceasing. It seems clear that in most cases when a project which has a significant amount of use has significant bugs and no maintainer then someone will step in. The cases where no-one takes over the project are often resolved by someone creating a competing project. If no-one takes over an abandoned Free Software project then it s a strong indication that the project wasn t particularly important anyway. I have no doubt that if any bug or missing feature made Linux systems stop booting then someone would fix it quite quickly. In a more general sense it seems that every time someone takes a position in a project that is of wide interest they are displacing someone else who might have done the job. When you volunteer to do significant work for a project you may be displacing someone who is more skillful than you this isn t necessarily a bad thing as there are plenty of other projects to work on, some of which require more skill. Growing More Programmers It seems to me that a large part of becoming a great programmer is facing great challenges. People who could be described as optimistic or arrogant will tend to take on more challenging tasks and therefore learn more. I m sure that there are a lot of people out there who have the potential to be great programmers apart from not taking on the challenging work, this seems to be an unfortunate waste of talent. Given a large enough population if someone leaves a senior position there should be someone else who can obtain the skills needed to take over. One advantage of this for Free Software development is that even if the best person to take on the challenge isn t living in the most convenient continent that won t be an obstacle, while with proprietary software development projects the teams are small and it s common that no-one else is capable of stepping up to a lead position. Another advantage is that when the lead developer leaves there are generally many candidates to replace them, all of whom can start work and be judged on the work that they do. I think that the best claims that can be made regarding essential people are not in regard to technical skill alone but to a combination of technical and people skills. Getting a group of programmers to work together is really hard but it s something that needs to be done for any significant project. Also the larger projects tend not to stand alone, being able to get changes included in other projects requires some skill. Ben Collins-Sussman and Brian Fitzpatrick gave an insightful talk at Google IO 2008 titled How to Protect Your Open Source Project From Poisonous People [2]. The first half of their talk is mostly about people who are misguided or difficult rather than what most people would consider poisonous and the second half is more about people who are actively poisonous and need to be removed. They advocate a community based on Politeness, Respect, Trust, and Humility. They describe in detail how the methods they advocate result in the members of their community being more productive, it seems obvious that those principles will lead to better career growth for people within the community and more friendly people wanting to join. When is being an Asshole OK? I once worked for a company that apparently had a team consisting solely of assholes. Apparently one asshole got promoted to management and after some internal transfers they ended up with all the assholes in the company on one team. I guess that when someone has negative interactions with everyone they won t notice the difference if they are put in a team where everyone is difficult. For a corporate environment that lacks a no jerks hiring policy this is probably a good way of improving productivity overall. I am not aware of any significant Free Software project that was comprised of mostly jerks although I have seen a few with dysfunctional environments that encourage the worst behavior from their members. The smaller Free Software projects have less need for people who can relate to other people. There are many useful Free Software projects which have only one developer, in most cases anyone can take the source code and use it without dealing with the author. But even for a single-developer project an asshole can cause some serious problems. One example I know of concerns a developer who had unclear licenses and started making legal threats in response to a request for a clear license. Another example is of a developer who released code that was designed to not work when one particular user compiled it and redistributed the binaries. Both of them caused some significant amounts of time to be wasted by people who were unfortunate enough to develop systems that interacted with the code in question, and even more time was wasted when some misguided people defended them in the inevitable flame-wars. Even for a project with only one developer it s still better for everyone if that developer isn t an asshole. One comment I ve seen related to this issue suggesting that some types of asshole behavior shouldn t be a problem an example that was cited is a colleague who cheats on a romantic partner. Jeremy Clyman (who is currently doing a Ph.D in Psychology) has written an interesting article about this for Psychology Today [3]. He reviews the movie The Dilemma which deals with someone catching their colleague s wife cheating. Jeremy analyses the psychological issues involved and how they can (among other things) impact the ability for such people to work. I once worked in an office where two married employees were very open about having an affair and we were all apparently expected to lie on their behalf if necessary, it really affected the quality of the working environment. Extreme Assholes There are lots of people involved in Free Software development who are difficult and many who are to some extent assholes. But some of them take being an asshole to the extreme, such as Holocaust deniers (an example which Rusty used). In the comments on his post the Westboro Baptist Church is also mentioned. It is possible to entirely disagree with someone on a contentious issue such as abortion but still be able to get along with them. But when someone supports a hate-based organisation such as the WBC or supports Nazis in any way then there will be many people who just can t tolerate them and no-one should be expected to tolerate such people. I have seen two instances where Free Software developers advocated pro-Nazi positions (one had an archive of neo-Nazi propaganda and the other claimed that Nazis were not responsible for the Holocaust). Neither of the pro-Nazi programmers was evicted for defending Nazis, but both of them ended up leaving the community in adverse ways after causing other damage in the mean time. I don t think it takes any great ability to predict the future to determine that someone who defends Nazis will eventually end up doing something that requires expulsion and drive away users and developers in the mean-time. There is no possibility that someone can support the Nazi or WBC ideology only when not associated with your project, it will affect all aspects of their life. When a Holocaust denier is allowed to be a member of a community it also sends out a message that members of the groups which were persecuted by Nazis aren t particularly welcome in the community. Helping Minor Assholes There are a lot of people who don t have malevolent aims but who unintentionally cause some difficulty (it seems that the truly malicious are a tiny minority). I don t think that excusing the bad things that they do on the basis of writing good code helps them in the long term. Many of the suggestions that Ben Collins-Sussman and Brian Fitzpatrick make seem likely to help people who don t want to be assholes and direct them towards positive involvement in the community. One trend that seems apparent is the non-linear response to certain types of bad behavior. There is often little difference in severity between something that gets almost no attention and something that results in a large and extremely hostile reaction. If someone persists in acting like an asshole for long enough it seems to be inevitable that they will eventually exceed some threshold for what is tolerated and get a very significant negative response. It would be good if things didn t need to get to that stage. I think that the most unfortunate aspect of Rusty s blog post is that most people will probably interpret it as encouragement to write better code as a way of getting a free pass for being an asshole. I know that this isn t what Rusty intended, but most people on the Internet don t know Rusty as well as I do. Conclusion Ben Collins-Sussman and Brian Fitzpatrick seem to have some of the best ideas for how to deal with these issues when you control a project, but most of us aren t in that position. Everyone can advocate better behavior. Extreme assholes need to be removed quickly and without a great debate about their contributions, freedom of speech, or other issues. Since considering this issue I ve been wondering about when one should avoid the lesser assholes and asshole-positive environments. People tend to adapt to their environment, so if you associate with assholes a lot then there s a good chance you will start to become like them. Is

• [1] http://rusty.ozlabs.org/?p=196
• [2] http://www.youtube.com/watch?v=-F-3E8pyjFo
• [3] http://www.psychologytoday.com/blog/reel-therapy/201105/what-causes-the-dilemma

The recent Alioth migration (thanks, Alioth folks!) led to slight variations in the URL schemes for various VCS. I guess some migration path will be considered but in the meanwhile that seemed to be a good example for the nice git-config insteadOf trick. Here s the bottom of ~/.gitconfig:

[url "git://anonscm.debian.org"]
    insteadOf = "git://git.debian.org"
[url "ssh://git.debian.org"]
    pushInsteadOf = "git://git.debian.org"

And tada! debcheckout picks the appropriate service (anonscm for anonymous access), and one is still able to push (through ssh) without having to change the URL in each and every .git/config file (which would also lead to higher CPU usage when updating afterwards).

This morning I awoke to a FedEx box landing on my head courtesy of my father. I jumped up and took a look at it I quickly opened it to find myself some goodies and a nice welcome letter Ooooh shiny Finally the holy grail ;) Just thought I d make everyone jealous of my shiny Google things :P
Now back to work, ~Sam

Guys, we have a problem. The name of that problem is NVidia and their Optimus technology. The idea of that tech is quite neat take a laptop, put two video cards in it, use the powerful card when you need 3D power, use the weak card when you need to conserve battery. The problem is that any laptop with this technology is currently an expensive paperweight on Linux (or rather it was so until a couple weeks ago, see below). And NVidia has no plans for fixing that. Let me explain the technicalities. In all modern laptops (such as the new and sexy Dell XPS 15z and many, many others) NVidia Optimus is implemented as follows there is an NVidia card and an Intel card, all display outputs are connected to the Intel card, when the user wants more 3D power, the NVidia card is powered on and it renders an image into a framebuffer in the video memory of the Intel card which is the responsible for actually outputing that framebuffer to a display. Currently in Linux if you are using NVidia drivers the Intel card is left uninitialised, your session starts up fine on the NVidia card, it outputs the resulting picture onto the video memory of the Intel card, but the Intel card is not configured in any way and so it does not do anything further with this picture and the display stays black. The alternative is to use Intel video card driver and then the NVidia card stays there in your laptop as dead weight. It does not even power down (out of the box) so you have all the power consumption and none of the functionality. There is some support from NVidia to using that card for CUDA computing purposes in such situation, but nothing else. After years of neglect and statements from NVidia that Optimus support in Linux is impossible and thus not even planned, one individual stepped up and made it work. Yes, the solution is hacky, yes it is a hassle to set up and make it work, but this is the first glimmer of hope for actually working Optimus on Linux. Over the last couple years the share of laptops with Optimus has greatly increased almost all laptops that have NVidia video cards have Optimus now. This Bumblebee project is currently our only real hope on making Linux work on these laptops. This project needs our support in a multitude of ways:
1) help test the actual project on different hardware;
2) look at the code and help improve on it, both the core code and system integration, simplify it so that everyone can install and use this;
3) figure out how we can integrate that support into distributions, so that it becomes possible to have working Linux out-of-the-box on Optimus hardware;
4) pressure NVidia more to provide official Optimus support on Linux (at least in always-on-NVidia mode, without breaking Bumblebee along the way). Martin has already showed the superiority of open source by implementing something that NVidia considered to be too hard to do, let s show the power of our community to push such solutions to production quality.

In February Borders went bankrupt [1], since then they have been in the slow process of closing down. Now Borders is trying to clear the last of their stock and offering 80% discounts off the marked price. I bought a book by Stephen Baxter and one by Peter F. Hamilton and those appeared to be the last two books worth buying (IMHO) on the almost empty sci-fi shelves, the books were a little tattered but at 80% discount I m not complaining. It s been almost four years since I last bought books, and I still haven t read all the free sci-fi stories and watched all the free sci-fi movies from the net which interest me [2]. So I m not planning to buy many more books unless I see something better than a 50% discount. Paul Wayper writes about the difficulty of buying ebooks [3]. It s ironic that some people have claimed that ebooks were part of the cause of Borders financial troubles given that they really aren t working well, not even for the most dedicated buyers. In related news Kobo (the company that runs the Borders ebook store) has assured customers that they won t lose the books that they own [4]. There are very few situations in which a company needs to assure customers that they won t lose property that they have paid for and received due to a corporate bankruptcy. As further evidence that Digital Restrictions Management (DRM) is a bad thing, Apple have shut down the iFlow Reader [5] so that they can monopolise ebook sales on the iPhone and iPad. This is a good reason to avoid restricted platforms (such as anything from Apple) and encrypted content. It seems to me that Cory Doctorow s scheme for giving copies of his books to libraries is a more effective way of donating in return for a free ebook [6] (which is rather similar to the buy one get one scheme that they used to run for OLPC). Hopefully Charles Stross will end up doing something similar to make Paul Wayper happy.

• [1] http://www.smartcompany.com.au/retail/20110217-borders-australia-and-angus-robertson-chains-collapse.html
• [2] http://etbe.coker.com.au/2007/09/23/dmca-and-sci-fi/
• [3] http://www.mabula.net/tbfw/blosxom.cgi/2011/05/23#2011-05-23-dear-charles-stross
• [4] http://www.readwriteweb.com/archives/kobo_reassures_its_e-book_customers_after_borders.php
• [5] https://www.iflowreader.com/Closing.aspx
• [6] http://craphound.com/walh/contribute/give-a-copy

Well today wasn t as productive as I had hoped. After my Skype call with vorlon I had layed out certain things I wanted to get done today and only got half of them done. This was mainly due to the fact my computer died midday and I had to reinstall Ubuntu. Luckily I didn t lose any data (thank god for advanced partitioning). Guess tomorrow will be twice as much work but I don t mind. Well until tomorrow, ~Sam

Or how to get rid of Evolution without removing your whole system There is a complain I have heard several times from people after they installed Debian: I use Icedove/Thunderbird so I do not need Evolution: I tried to remove it but this would remove my whole system, that sucks! Explanation The reason behind this situation is that GNOME is installed by a meta-package, which is a dummy package that depends on each component of the official GNOME desktop, including Evolution. So it got installed as a dependency of that dummy package. This means that you, or rather, the Debian installer, never asked explicitly to install either Evolution or any other specific software: it came automatically. Now, you want to remove Evolution: very well, but this implies removing the meta-package, as you are removing one of its dependencies. But if you are removing this meta-package, there is no reason to keep all this software you did not asked for, so, as they got installed automatically, they are now removed automatically as well The solution So, do you still want to remove Evolution but keep the remaining of your GNOME desktop? Of course you will not keep GNOME , because GNOME includes Evolution, so strictly speaking GNOME without Evolution is not GNOME any more. Well, there is a solution: marking all the dependencies of the meta-package as manually installed, that is, faking that you installed them explicitly Aptitude is able to do that with its powerful search functions:

# aptitude unmarkauto '?reverse-depends(gnome)   ?reverse-recommends(gnome)'
# aptitude unmarkauto '?reverse-depends(gnome-core)   ?reverse-recommends(gnome-core)'
# aptitude unmarkauto '?reverse-depends(gnome-desktop-environment)'\
'   ?reverse-recommends(gnome-desktop-environment)'

What did we do? Well, exactly what is written: unmarking as automatically installed the packages that reverse-depend or reverse-recommend gnome or gnome-desktop-environment. In other words, those that are dependencies or recommendations of gnome, gnome-core or gnome-desktop-environment. Why these three packages? Because the GNOME environment that Debian provides by default is composed of several meta-packages: I shall let you refer to their respective descriptions for further information. Now, you can safely remove Evolution without pulling out your whole desktop system:

# aptitude remove evolution

The Debian kernel team regularly backports driver updates to the Linux kernel in stable releases to add support for new hardware, and I've prepared several updates intended for point release 6.0.2. Since the kernel team does not have a large collection of hardware on which to test driver changes, we would appreciate test reports from users. It is important to test not just that new devices are supported properly, but that there are no regressions in support for older devices. Changes I have updated these drivers to the versions found in Linux 2.6.38, modulo driver API changes:

• bna: New driver for Brocade Ethernet adapters. (#627699)
• e1000e: Add support for i82567V-4, i82579. Bug fixes for i82577, i82578, i82583. (#627700)
• hpsa: New driver for HP Smart Array controllers. It also supports some devices that are currently handled by cciss, but I have disabled that for 'squeeze'. (#627701)
• igb, igbvf: Add support for i82576-ET2, 82580, DH89xxCC, i350. (#627702)
• pm8001: New driver for PMC-Sierra SAS/SATA HBAs. (#627703)
• r8169: Add full support for RTL8168DP. Bug fixes for RTL8102E. (#627704)
• tg3, broadcom: Add support for BCM5717, BCM5719, BCM57765, and BCM50610M and BCM5241 PHYs. Bug fixes for BCM5755. (#627705)

I have also cherry-picked some small changes:

• bnx2i: Add support for BCM5771E
• wl1251: Add support for PG11 chips
• bnx2x: Add support for BCM84823
• ar9170usb: Add several additional USB device IDs

There are more drivers that I think should be added or updated (see #624794) but they will probably have to wait for release 6.0.3. Packages The source package and binary packages for i386 and amd64 are available on people.debian.org. They can be verified by the checksums in the signed changes file. The current packages are version 2.6.32-35~test1, but there may be further test versions before an official stable update. How to test For network drivers, I suggest the following regression tests:

1. If the driver tries to load firmware (only required for some chips), does this work once the firmware file(s) are installed?
2. Can you receive and transmit VLAN-tagged frames after creating a VLAN interface?
3. Does the interface work after suspend and resume?
4. Does the interface work after removing the cable for 10 seconds and reinserting it?
5. Does multicast configuration work? (IPv6 autoconfiguration or mDNS will cover this.)
6. Can the interface send and receive TCP/IP across a LAN at the same speed, before and after these changes? (Use e.g. netperf to test this, but don't forget to remove the netperf package after use.)
7. Are any warnings or errors logged by the kernel during the preceding tests?

For storage drivers, unfortunately I don't have a good idea of what tests would be suitable. In any case, please don't test on disks storing valuable data! Please send test reports to the bug reports linked above, stating the driver name, the PCI ID for the device you tested (from lspci -n) and any other device identification that the kernel log (for example, r8169 logs the 'XID' of the device).

Jazz on the table

<iframe frameborder="0" marginheight="0" marginwidth="0" scrolling="no" src="http://rcm.amazon.com/e/cm?lt1=_blank&amp;bc1=000000&amp;IS2=1&amp;bg1=FFFFFF&amp;fc1=000000&amp;lc1=0000FF&amp;t=dedasys-20&amp;o=1&amp;p=8&amp;l=as4&amp;m=amazon&amp;f=ifr&amp;ref=ss_til&amp;asins=B004IYISQW" style="width: 120px; height: 240px; float: left; margin-right: 20px; margin-bottom: 8px;"></iframe> Many small business owners start a business with the idea of greater freedom in mind, yet end up chaining themselves to something that takes even more dedication than a regular job. This book discusses the dos and don'ts that business owners need to be conscious of in order to create a business that is independent of its owner(s) in order to be able to sell it. It does so in an easy to read format that's a bit of a gimmick - a fictional story illustrating the authors points. Like most business books, the core idea isn't that hard to explain. the book even includes a handy summary of the main points, which I'll paraphrase here:

1. Focus on one thing you can do really well; even a service business can be 'productized' by creating a standard service that is the focus of the business.
2. If your business revolves around one or two key clients, the risk inherent in that approach will lower the value or scare of potential buyers completely.
3. Put a process in place, from sales through production.
4. Don't "be" your company. If the company is all about you, what's a potential buyer really getting if you sell it and leave?
5. By focusing on products, you can charge up front rather than having poor cash flow. If you pay people to work on a project for 3 months, and don't get paid for another month, you are, by the end, out 4 months salary while waiting for the payment.
6. Say no to projects outside the scope of your business. Only by focusing narrowly can you really excel at what you do.
7. Spend time doing some research and calculations to estimate your potential market size; buyers will want to know this.
8. If you have a business that has sales people, hire at least two so that they'll compete with one another.
9. You want people who are good at selling products, not services; the latter will want to tweak your offering for each and every client, rather than selling it as-is and trying to find how it can meet the customer's needs that way, which is the best strategy for a product.
10. If you were previously running a more 'generic' services company, and you switch to a more productized approach, be prepared to take a hit the year when you switch.
11. Potential acquirers will want to see at least a couple of years of steady growth with the new model after making the switch.
12. If you grow, you'll need a management team that can work without you. Put an incentive system into place to reward their loyalty and results.
13. When looking for an adviser to sell your company, aim for one where you will not be the largest or smallest client.
14. If your adviser really only has one company in mind to sell yours to, they may be trying to sell you off cheap as a favor.
15. Consider how much you could grow with the resources of a buyer. Think big, and show them what kind of growth could happen with the right backing.
16. Think and speak like a product business with 'customers' rather than 'clients'.
17. Stock options are more complex than simpler options like bonuses that are paid out over a period of time, in terms of ways to give people an incentive to stay.

All in all, it was an enjoyable read, with solid points. Even if you have no intention of selling your business, thinking of the business itself as a sort of "product" that is not dependent on you to work correctly is a sensible way to go about creating and growing a business. He's certainly not, nor claims to be the first one to discuss this idea, and indeed references the well known E-Myth Revisited which focuses less on selling a business, and more on how to go about extricating you and your skills from the business.

long time no post about my RC bug fixing activities. & indeed, I haven't been very active during the last weeks. here's a quick update; it includes two NMUs done for the perl 5.12 transition, the rest is just "normal" maintainance work (interesting to see how much stuff can break behind our backs ...).

• #578940 libhttp-proxy-perl: "libhttp-proxy-perl: FTBFS with perl 5.12.0-1: POD coverage"
  add patch as a workaround (pkg-perl)
• #603251 libsys-syslog-perl: "libsys-syslog-perl: obsoleted by perl in squeeze"
  upload new upstream release (pkg-perl)
• #612780 libnet-ssleay-perl: "libnet-ssleay-perl: FTBFS against openssl 1.0.0c-2 from experimental"
  upload package with patch by chrisb (pkg-perl)
• #615571 libclone-fast-perl: "libclone-fast-perl: FTBFS with perl 5.12.3: test failures"
  upload with patch from niko tyni (pkg-perl)
• #616566 libconfig-model-itself-perl: "libconfig-model-itself-perl: Failing test t/itself.t"
  upload package prepared by dod (pkg-perl)
• #621019 libwww-perl: "typo in depends makes package uninstallable"
  fix typo in dependency (pkg-perl)
• #621135 libwww-perl: "libwww-perl: Breaks partial upgrades from squeeze"
  upload package prepared by periapt (pkg-perl)
• #621468 libmodule-cpants-analyse-perl: "libmodule-cpants-analyse-perl: FTBFS: Can't locate Test/YAML/Meta/Version.pm in @INC"
  grab patch from upstream bugtracker (pkg-perl)
• #621469 src:libmojomojo-perl: "libmojomojo-perl: FTBFS - failing tests"
  upload new upstream release with fixed dependency (pkg-perl)
• #622919 libjifty-dbi-perl: "libjifty-dbi-perl: security SQL injection"
  upload new upstream release (pkg-perl)
• #623510 libjgoodies-looks-java-doc,libjgoodies-common-java: "libjgoodies-common-java and libjgoodies-looks-java-doc: error when trying to install together"
  fix Document files in .doc-base file
• #624767 src:libbuffy-bindings: "libbuffy-bindings: FTBFS"
  update build dependencies, upload to DELAYED/2
• #624811 src:mapserver: "mapserver: FTBFS: ld: cannot find -lepsilon"
  add missing build dependency, upload to DELAYED/2
• #625840 libnet-freedb-perl: "libnet-freedb-perl: FTBFS - test suite needs internet access"
  disable tests that need internet access (pkg-perl)
• #625908 libmath-random-perl: "libmath-random-perl: Not whitelisted for autobuildding in non-free"
  add XS-Autobuild field (pkg-perl)
• #627170 src:libdata-amf-perl: "libdata-amf-perl: Can't locate YAML/Base.pm in @INC"
  add missing build dependency (pkg-perl)
• #627177 src:libtest-www-mechanize-perl: "libtest-www-mechanize-perl: Failing tests"
  upload new upstream release (pkg-perl)
• #627229 src:libaudio-ecasound-perl: "libaudio-ecasound-perl: Ecasound.xs:5:23: fatal error: ecasoundc.h: No such file or directory"
  upload package with fixed build dependencies prepared by ghedo (pkg-perl)
• #627231 src:libdbix-class-candy-perl: "libdbix-class-candy-perl: Failing tests t/irc-schema.t"
  adjust versioned build dependencies (pkg-perl)
• #627238 src:libcatalyst-modules-perl: "libcatalyst-modules-perl: failing tests"
  update the faulty module (pkg-perl)
• #627270 src:libstring-errf-perl: "libstring-errf-perl: failing test in t/basic.t"
  upload package prepared by ghedo (pkg-perl)

I still hope to find some more time in the future to look at other RC bugs; after all we want to have a not-too-long freeze for wheezy, don't we?