NUUG-styremedlem Hans-Petter Fjeld meldte nettopp at han har sendt inn NUUGs høringsuttalelse angående Difi sin standardkatalog v3.1. Jeg er veldig glad for at så mange bidro og sikret at vår stemme blir hørt i denne høringen. Anbefaler alle å lese våre to sider med innspill.
In an article today published by Computerworld Norway, the photographer Eirik Helland Urke reports that the video editor application included with HTC One X have some quite surprising terms of use. The article is mostly based on the twitter message from mister Urke, stating:
"Drøy brukeravtale: HTC kan bruke MINE redigerte videoer kommersielt. Selv kan jeg KUN bruke dem privat."
I quickly translated it to this English message:
"Arrogant user agreement: HTC can use MY edited videos commercially. Although I can ONLY use them privately."
I've been unable to find the text of the license term myself, but suspect it is a variation of the MPEG-LA terms I discovered with my Canon IXUS 130. The HTC One X specification specifies that the recording format of the phone is .amr for audio and .mp3 for video. AMR is Adaptive Multi-Rate audio codec with patents which according to the Wikipedia article require an license agreement with VoiceAge. MP4 is MPEG4 with H.264, which according to Wikipedia require a licence agreement with MPEG-LA.
I know why I prefer free and open standards also for video.
I januar i fjor startet vi i NUUG arbeid med å gjøre informasjon om hvem som har stemt hva på Stortinget enklere tilgjengelig. I løpet av få måneder fant vi sammen med organisasjonen Holder de ord som arbeidet mot et lignende mål.
Siden den gang har vi fått tak i maskinelt lesbart informasjon om hvem som stemte hva mellom 1990 og våren 2010, og tilgang til stortingets nye datatjeneste som har informasjon fra høsten 2011 til i dag. Det gjenstår litt arbeid med det første datasettet, men datasettet fra høsten 2011 er klart til bruk. Begge datasettene er tilgjengelig via git.
På Go Open i morgen lanserer NUUG sammen med Holder de ord en test-tjeneste som viser hva som er og blir behandlet på Stortinget og hvem som har stemt hva siden oktober i fjor. Du får herved mulighet til å ta en sniktitt.
Here in Norway, the Ministry of Government Administration, Reform and Church Affairs is behind a directory of standards that are recommended or mandatory for use by the government. When the directory was created, the people behind it made an effort to ensure that everyone would be able to implement the standards and compete on equal terms to supply software and solutions to the government. Free software and non-free software could compete on the same level.
But recently, some standards with RAND (Reasonable And Non-Discriminatory) terms have made their way into the directory. And while this might not sound too bad, the fact is that standard specifications with RAND terms often block free software from implementing them. The reasonable part of RAND mean that the cost per user/unit is low,and the non-discriminatory part mean that everyone willing to pay will get a license. Both sound great in theory. In practice, to get such license one need to be able to count users, and be able to pay a small amount of money per unit or user. By definition, users of free software do not need to register their use. So counting users or units is not possible for free software projects. And given that people will use the software without handing any money to the author, it is not really economically possible for a free software author to pay a small amount of money to license the rights to implement a standard when the income available is zero. The result in these situations is that free software are locked out from implementing standards with RAND terms.
Because of this, when I see someone claiming the terms of a standard is reasonable and non-discriminatory, all I can think of is how this really is non-reasonable and discriminatory. Because free software developers are working in a global market, it does not really help to know that software patents are not supposed to be enforceable in Norway. The patent regimes in other countries affect us even here. I really hope the people behind the standard directory will pay more attention to these issues in the future.
You can find more on the issues with RAND, FRAND and RAND-Z terms from Simon Phipps (RAND: Not So Reasonable?).
Update 2012-04-21: Just came across a blog post from Glyn Moody over at Computer World UK warning about the same issue, and urging people to speak out to the UK government. I can only urge Norwegian users to do the same for the hearing taking place at the moment (respond before 2012-04-27). It proposes to require video conferencing standards including specifications with RAND terms.
Da jeg googlet etter noe annet kom jeg tilfeldigvis over en hovedfagsoppgave ved Universitetet i Oslo som diskuterer verdien av GPLs fire friheter for brukerne av IT-systemer. Jeg ble fascinert over det som presenteres der. Her er sammendraget:
Motivasjonen til å skrive denne oppgaven er en personlig undring over hvorfor det primært, og ofte eksklusivt, fokuseres på det økonomiske aspektet ved utredninger om fri programvare er et godt valg for det offentlige. Fri og produsenteid programvare bygger på fundamentalt forskjellige ideologier som kan ha implikasjoner utover økonomiske kostnader. Kunnskapskulturen som er med på å definere fri programvare er basert på åpenhet, og er en verdi i seg selv.
Oppgavens tema er programvarelisensen GPL og frihet. GPL-lisensiert programvare gir visse friheter i forhold til produsenteid programvare. Mitt spørsmål er om, og eventuelt i hvilken utstrekning, disse frihetene blir benyttet av ulike brukere og hvordan de manifesterer seg for disse brukerne. Sentrale spørsmål i oppgaven er:
- Hvordan fordeles handlekraft gjennom lisensieringen av programvaren?
- Hvilke konsekvenser har programvarelisensen for de ulike brukere?
Fri programvare gir blant annet brukeren mulighet til å studere og modifisere kildekoden. Denne formen for frihet erverves gjennom kunnskap og krever at brukeren også er en ekspert. Hva skjer med frihetene til GPL når sluttbrukeren er en annen? Dette diskuteres i dialog med informantene.
Jeg har i denne oppgaven samlet inn intervjudata fra IKT-ansvarlige ved grunnskolene i Nittedal kommune, driftsansvarlig og IKT-veilederen for skolene i kommunen, samt IKT-koordinator for utdanning i Akershus fylkeskommune og bokmåloversettere av OpenOffice.org. Den empiriske delen av oppgaven er delt inn i to seksjoner; den første omhandler operativsystemet Skolelinux, den andre kontorprogrampakken OpenOffice.org.
Som vi vil se gir GPL lokal frihet og kontroll gjennom omfordeling av makt fra produsent til bruker. Brukerens makt analyseres gjennom begrepene brukermedvirkning og handlingsfrihet. Det blir også lagt vekt på strukturelle forhold rundt bruken av teknologi, og spesielt de økonomiske begrepene nettverkseksternaliteter, innlåsing og stiavhengighet. Dette er begreper av spesiell nytte når objektet som omsettes eller distribueres er et kommunikasjonsprodukt, fordi verdien til et slikt gode for en potensiell bruker avhenger av antall eksisterende brukere av godet. I tilknytning til denne problematikken inneholder oppgaven også en diskusjon rundt åpne standarder og formater.
Oppgaven konkluderer med at de «fire frihetene» som GPL-lisensen er laget for å beskytte er av avgjørende betydning for bruken av OpenOffice.org og Skolelinux, i Akershus fylkeskommune såvel som i skolene i Nittedal. Distribusjonen av handlekraft er ikke helt symmetrisk. Det er først og fremst de profesjonelle utviklerne i Skolelinux som direkte kan nyttiggjøre seg friheten til å endre kode, mens en sluttbruker som Nittedal kommune nyttiggjør seg den økonomiske friheten til å kunne distribuere programmene. Det er imidlertid også slik at ingen aktør klarer seg uten alle disse «frihetene».
Jeg fant også en masteroppgave fra 2006, men der ligger ikke komplett oppgave tilgjengelig. På tide å holde et øye med Skolelinux-søket til DUO...
Behind Debian Edu and Skolelinux there are a lot of people doing the hard work of setting together all the pieces. This time I present to you Andreas Mundt, who have been part of the technical development team several years. He was also a key contributor in getting GOsa and Kerberos set up in the recently released Debian Edu Squeeze version.
Who are you, and how do you spend your days?
My name is Andreas Mundt, I grew up in south Germany. After studying Physics I spent several years at university doing research in Quantum Optics. After that I worked some years in an optics company. Finally I decided to turn over a new leaf in my life and started teaching 10 to 19 years old kids at school. I teach math, physics, information technology and science/technology.
How did you get in contact with the Skolelinux/Debian Edu project?
Already before I switched to teaching, I followed the Debian Edu project because of my interest in education and Debian. Within the qualification/training period for the teaching, I started contributing.
What do you see as the advantages of Skolelinux/Debian Edu?
The advantages of Debian Edu are the well known name, the out-of-the-box philosophy and of course the great free software of the Debian Project!
What do you see as the disadvantages of Skolelinux/Debian Edu?
As every coin has two sides, the out-of-the-box philosophy has its downside, too. In my opinion, it is hard to modify and tweak the setup, if you need or want that. Further more, it is not easily possible to upgrade the system to a new release. It takes much too long after a Debian release to prepare the -Edu release, perhaps because the number of developers working on the core of the code is rather small and often busy elsewhere.
The Debian LAN project might fill the use case of a more flexible system.
Which free software do you use daily?
I am only using non-free software if I am forced to and run Debian on all my machines. For documents I prefer LaTeX and PGF/TikZ, then mutt and iceweasel for email respectively web browsing. At school I have Arduino and Fritzing in use for a micro controller project.
Which strategy do you believe is the right one to use to get schools to use free software?
One of the major problems is the vendor lock-in from top to bottom: Especially in combination with ignorant government employees and politicians, this works out great for the "market-leader". The school administration here in Baden-Wuerttemberg is occupied by that vendor. Documents have to be prepared in non-free, proprietary formats. Even free browsers do not work for the school administration. Publishers of school books provide software only for proprietary platforms.
To change this, political work is very important. Parts of the political spectrum have become aware of the problem in the last years. However it takes quite some time and courageous politicians to 'free' the system. There is currently some discussion about "Open Data" and "Free/Open Standards". I am not sure if all the involved parties have a clue about the potential of these ideas, and probably only a fraction takes them seriously. However it might slowly make free software and the philosophy behind it more known and popular.
Jeg har tenkt meg på konferansen Go Open 2012 i Oslo 23. april. Medlemsforeningen NUUG deler ut prisen for fremme av fri programvare i Norge der i år. Kommer du?
It take all kind of contributions to create a Linux distribution like Debian Edu / Skolelinux, and this time I lend the ear to Justin B. Rye, who is listed as a big contributor to the Debian Edu Squeeze release manual.
Who are you, and how do you spend your days?
I'm a 44-year-old linguistics graduate living in Edinburgh who has occasionally been employed as a sysadmin.
How did you get in contact with the Skolelinux/Debian Edu project?
I'm neither a developer nor a Skolelinux/Debian Edu user! The only reason my name's in the credits for the documentation is that I hang around on debian-l10n-english waiting for people to mention things they'd like a native English speaker to proofread... So I did a sweep through the wiki for typos and Norglish and inconsistent spellings of "localisation".
What do you see as the advantages of Skolelinux/Debian Edu?
What do you see as the disadvantages of Skolelinux/Debian Edu?
These questions are too hard for me - I don't use it! In fact I had hardly any contact with I.T. until long after I'd got out of the education system.
I can tell you the advantages of Debian for me though: it soaks up as much of my free time as I want and no more, and lets me do everything I want a computer for without ever forcing me to spend money on the latest hardware.
Which free software do you use daily?
I've been using Debian since Rex; popularity-contest says the software that I use most is xinit, xterm, and xulrunner (in other words, I use a distinctly retro sort of desktop).
Which strategy do you believe is the right one to use to get schools to use free software?
Well, I don't know. I suppose I'd be inclined to try reasoning with the people who make the decisions, but obviously if that worked you would hardly need a strategy.
Recently I have spent time with Skolelinux Drift AS on speeding up a Debian Edu / Skolelinux Lenny installation using LTSP diskless workstations, and in the process I discovered something very surprising. The reason the KDE menu was responding slow when using it for the first time, was mostly due to the way KDE find application icons. I discovered that showing the Multimedia menu would cause more than 20 000 IP packages to be passed between the LTSP client and the NFS server. Most of these were NFS LOOKUP calls, resulting in a NFS3ERR_NOENT response. Because the ping times between the client and the server were in the range 2-20 ms, the menus would be very slow. Looking at the strace of kicker in Lenny (or plasma-desktop i Squeeze - same problem there), I see that the source of these NFS calls are access(2) system calls for non-existing files. KDE can do hundreds of access(2) calls to find one icon file. In my example, just finding the mplayer icon required around 230 access(2) calls.
The KDE code seem to search for icons using a list of icon directories, and the list of possible directories is large. In (almost) each directory, it look for files ending in .png, .svgz, .svg and .xpm. The result is a very slow KDE menu when /usr/ is NFS mounted. Showing a single sub menu may result in thousands of NFS requests. I am not the first one to discover this. I found a KDE bug report from 2009 about this problem, and it is still unsolved.
My solution to speed up the KDE menu was to create a package kde-icon-cache that upon installation will look at all .desktop files used to generate the KDE menu, find their icons, search the icon paths for the file that KDE will end up finding at run time, and copying the icon file to /var/lib/kde-icon-cache/. Finally, I add symlinks to these icon files in one of the first directories where KDE will look for them. This cut down the number of file accesses required to find one icon from several hundred to less than 5, and make the KDE menu almost instantaneous. I'm not quite sure where to make the package publicly available, so for now it is only available on request.
The bug report mention that this do not only affect the KDE menu and icon handling, but also the login process. Not quite sure how to speed up that part without replacing NFS with for example NBD, and that is not really an option at the moment.
If you got feedback on this issue, please let us know on debian-edu (at) lists.debian.org.
About two weeks ago, I was interviewed via email about Debian Edu and Skolelinux by Bruce Byfield in Linux Weekly News. The result was made public for non-subscribers today. I am pleased to see liked our Linux solution for schools. Check out his article Debian Edu/Skolelinux: A distribution for education if you want to learn more.
Created by Chronicle v4.4