I'm experiencing random crashes since X.org moved to 7.5. As of now, I could take the needed time to properly report them as I hate reporting just "keeps crashing on my system" bugs and make them RC (but, for me they are). I couldn't even take time to downgrade to 7.4 by using testing package while they were here. They mostly seem to be related to my dual screen setup at work (no crashes at home) and/or VirtualBox use (I use VB at work to host my work Windows machine). Of course, as expected, X.org 7.5 migrated to testing and now I'm screwed. I still have the crashes...and can't even easily downgrade..:-( Apart from beating myself, would there be a chance that someone still has these packages somewhere? I probably need 7.4+4 packages.

The first time we looked at this condo that we ended up buying, I looked at the cupboard under the stairs, and could visualise a small 19 inch rack in the lower part of it, with all of my computer gear in it. When we got structured cabling installed, I had the CAT-6 cabling terminated onto a basic 19 inch patch panel, with the intent of mounting this in a rack. Then it became a case of trying to find a rack. My favourite junk shop, Weird Stuff, didn't have anything that wasn't full height, so I started looking around online. There's certainly a lot of variance in price. I settled on a 26U Intellinet rack from New Tech, because based on the dimensions on the website, it would fit in the space I had in mind. When we assembled it, it became obvious that the dimensions quoted were the inside dimensions and not the outside dimensions, and it was about 4 centimetres too tall. In hindsight, I should have figured that out. 26 x 1.75 is 45.5. This is where I must give a shout out to New Tech. Sarah called them up, explained the situation, and they agreed to take the rack back, and sent us some shipping labels to ship it back. We managed to find a more simple rack, a 20U Middle Atlantic, which arrived today, and was significantly easier to put together. The only downside: no rear mounting holes. I don't think it'll be a huge problem, though. All I'm planning on mounting in the immediate future are the patch panel, some sort of cable management, a Catalyst switch, and a power strip. At some point in the future, once we stop spending money hand over fist, I'll look at getting a rack mountable server to replace the hodge podge of computers I'm currently running. What I wish we'd discovered about two days ago, was the Lack Rack. This would have been perfect (and so much cheaper).

We finished off moving everything from the old place last weekend. Our new home number can be determined by adding 59115 to the old one (or calling the old one in the next 60 days). The great unpacking is continuing. We're reluctant to unpack much in the kitchen because it's all going to have to come back out again in a couple of weeks. The cats have settled in well. Lily is still spending most of her time under the bed, but she comes out to explore from time to time. We had a bunch of fun with various bits of cable. I had some CAT-6 cabling installed throughout, and also some new RG-6 cabling was installed at the same time. The contractor ran a new line into our place and also a line to the patch panel he installed for the phone. When Comcast came out to activate our cable TV service, the poor guy had all sorts of problems, until I realised there was an air gap between the line coming into the condo and the cables outside. Then the next problem was my contractor had run the new line out to the wrong location. The bunch of cables where he ran the line to was some old out of commission stuff. The Comcast guy was really great. We located where the current cable came in (via the attic) and he spliced that into the new internal cabling, and the net outcome was close to what was supposed to have happened in the first place. The funniest thing was the Comcast cable guy left in such a rush because he ran over time on the job, that he left the cable cabinet wide open, including the box with all of the building's cable connections. If I new how to crimp RG-6, and was feeling lucky, I could remove the filter that's restricting the channels we receive. The next fun was getting the phone line sorted out. Again, my contractor had run a new CAT-6 line to the patch panel for the purposes of giving me a few jacks in the one place, so I could plug in my DSL modem, and a line into my Asterisk box. The only problem was he wasn't sure which pair was the phone line (and the phone wasn't active at the time) so there was an air gap between the end of the CAT-6 and the phone line. AT&T came out and sorted that out. I suspect the AT&T guy also left in a hurry, because he left a tone generator jumpered into the phone line. Once the phone was working we had to wait a couple of days for the DSL service to get moved to the new number. It turns out that I got new static IP addresses with the move, so it wasn't just a case of plugging everything back in and it working, unfortunately. But that's all behind us now. We're fully here, online again, and settling in. I think this weekend we'll bang in some picture hooks and hang up our paintings, and try to unpack a few more boxes.

Registration is now open for DebConf10. The dates of the conference are August 1-7, 2010, with arrivals at our group lodging permitted as of 3 PM on July 31 and departures required by 11 AM on August 8. The conference is preceded by DebCamp from July 25-31 including the arrival day. To receive announcements regarding DebConf and DebCamp, please subscribe to the debconf-announce mailing list. In order to avoid spamming Planet Debian with a long post detailing registration procedure, travel sponsorship, DebCamp info, and cost of attendance, please instead see the DebConf10 website s registration page to answer all your questions. As always, feel free to contact the DebConf team via email or IRC for answers not provided by the website. DebConf10 is going to be great. We re looking forward to seeing you in New York in August! the DebConf team

Romain uploaded our first release of RProtoBuf to CRAN yesterday. RProtoBuf provides bindings for GNU R to the Google Protobuf implementation. Google Protobuf is (and I quote) a way of encoding structured data in an efficient yet extensible format that is used for almost all internal RPC protocols and file formats at Google. RProtoBuf had a funny start. I had blogged about the 12 hour passage from proof of concept to R-Forge project following the ORD session hackfest in October. What happened next was as good. Romain emailed within hours of the blog post and reminded me of a similar project that is part of Saptarshi Guha's RHIPE R/Hadoop implementation. So the three of us--Romain, Saptarshi and I---started emailing and before long it becomes clear that Romain is both rather intrigued by this (whereas Saptarshi has slightly different needs for the inner workings of his Hadoop bindings) and was able to devote some time to it. So the code kept growing and growing at a fairly rapid clip. Til that stopped as we switched to working feverishly on Rcpp to both support the needs of this project, and to implement ideas we had while working on this. That now lead to the point where Rcpp is maturing in terms of features, so we will probably have time come back to more work on RProtoBuf to take advantage of the nice templated autoconversions we now have in Rcpp. Oddly enough, the initial blog post seemed to anticipate changes in Rcpp. Anyway -- RProtoBuf is finally here and it already does a fair amount of magic based of code reflection using the proto files. The Google documentation has a simple example of a 'person' entry in an 'addressbook' which, when translated to R, goes like this:

R> library( RProtoBuf )                      ## load the package
R> readProtoFiles( "addressbook.proto" )     ## acquire protobuf information
R> bob <- new( tutorial.Person,              ## create new object
+   email = "bob@example.com",
+   name = "Bob",
+   id = 123 )
R> writeLines( bob$toString() )              ## serialize to stdout
name: "Bob"
id: 123
email: "bob@example.com"
R> bob$email                                 ## access and/or override
[1] "bob@example.com"
R> bob$id <- 5
R> bob$id
[1] 5
R> serialize( bob, "person.pb" )             ## serialize to compact binary format

There is more information at the RProtoBuf page, and we already have a draft package vignette, a 'quick' overview vignette and a unit test summary vignette. More changes should be forthcoming as Romain and I find time to code them up. Feedback is as always welcome.

So I finally got off my lazy butt and started the paperwork to get a formal recognition of studies equivalence for a undergraduate studies (Licenciatura en ingenier a de software) via CENEVAL's Acuerdo 286 (licenciatura). Part of the paperwork involves filling the form I photographed and attached to this node. This is the utmost example of ber-redundant paperwork... Where it requires my personal data, some of the fields are:

CURP

Clave nica de Registro Poblacional, Unique Populational Registration Key. This is a string composed by:

• WOIG: First letter and first vowel after the first letter of my first family name, first letter of my second family name, first letter of my given name
• 760427: birth date, yymmdd
• H: Sex (H = male, M = female, I guess)
• DF: State or federative entity I was born in
• LSN: First consonant after the first letter of the first and second family name and of the given name
• 03: Deambiguation digits

So far, so good.

Age

Well, cannot it be deduced from the CURP? It has my full birth date!

Sex

Well, cannot it be deduced from the CURP? It is a specific field in it

Birth date

Man, you already have it literally on top of the field!

Nationality

This tramit is only for Mexicans, so... what's the point?

Birth place

The CURP states already the state or federative entity I was born... But I'll accept this one, as not all states have only one city as DF does

Quite a nice catalog of redundancies. As a cherry on top of the cake, the phone number states I should write my long distance phone code (LADA historically, Larga Distancia Autom tica, Automatic Long Distance) only if I am in any of the states. I can only ask myself why... Anyway... Lets continue filling paperwork. Grah. Hopefully I will be able to get my papers... somewhen in the next half century.

I'm starting to notice a pattern here:

• Google executives are on trial because some sorry excuses for human beings picked on a retarded person and posted the video to youtube: http://news.bbc.co.uk/2/hi/technology/8115572.stm - this one is simply preposterous. Going after the execs of a company who did nothing to aid, abet, condone or in any way facilitate the abuse in question is absurd, and if extended to other industries would mean that you could pretty much attack any company whose products happened to figure in a crime somehow. Kitchen knives, hunting rifles, golf clubs, even automobiles would seem fair game.
• Italy is going after "user generated content" sites like Youtube and wants to force them to register with the government if they wish to operate: http://arstechnica.com/tech-policy/news/2010/02/italy-preparing-to-hold-youtube-others-liable-for-uploads.ars
• And last but not least, this hit piece in the normally respectable Corriere della Sera: http://www.corriere.it/economia/10_gennaio_28/mucchetti_4de4be8a-0be8-11df-bc70-00144f02aabe.shtml
  - it's in Italian, but the gist of it is that Mr Mucchetti really has it in for Google because they operate out of Ireland in the EU, whereas he believes they should be registered in Italy as a publisher, and subject to Italy's myriad rules, regulations, and, of course, taxes regarding publishing. Despite, well, not really publishing much of anything themselves. He mentions "tax evasion" charges that had been considered, because the Italian division of Google is not where the adsense revenue in Europe goes. I suppose he figures that since the ads are bought by residents of Italy, the money should somehow stay in Italy? He also huffs and puffs about Italy's antitrust laws, which, in the same piece, he admits were created with the express purpose of not touching existing companies (the market share limit was set higher than the share of the largest existing company). Perhaps he would do well to reflect on political schemes and carve-ups like that and think about why companies like Google go to Ireland, rather than Italy. He also makes some quick mentions of network neutrality, and rambles on a bit about how it's a battle between the "Obamanian, Californian, search engines" versus the telecommunications industry, in "the rest of the world and above all in Europe". And of course he uses a liberal sprinking of keywords like "globalization", "multinational corporations", and "deregulated" to attempt to paint Google in terms of being a big, evil company throwing its weight around. One wonders if there aren't more pressing problems with the Italian media industry, such as the prime minister owning a large chunk of it?

One way of seeing things is that politicians and businessmen in Italy noticed Google was actually making quite a bit of money, and even if they don't quite understand this internet thing, they want some of the loot. And while Google certainly is becoming big enough to be cause for worry and discussion, the moves against them in Italy do not seem anything like a rational response calculated to offset severe failures in the market. In any case, it will be interesting to see what happens. Maybe, after China, we'll see Google quit Italy as well?

I was interested to see Adnan Hodzic discuss life without evolution in the GNOME environment recently. I too use GNOME as my desktop environment (I sometimes toy with various tiling window managers before getting annoyed at something or other). My solution to the GNOME problem is to purge the gnome-desktop-environment package and instead my own local package gnome-desktop-minimal. This package is a meta-package which includes a smaller selection of GNOME packages, notably ignoring several that the gnome-core package would pull in such as eog - why install that when I prefer qiv or feh? If I believed we could agree on precisely which packages to include I would submit a bug to the gnome team "Please provide gnome-desktop-minimal" or similar. Still I suspect individual biases/preferences will make such a suggestion contentious at best and impossible to satisfy at worst. ObTitle: L on

I was trying not to complain about it, but now that the number of people asking me about it is getting bigger, my frustration got the best of me. So unfortunately I won t see you all there.

Cropping images with GDAL I am working to get a better integration between Meteosatlib and GDAL. A nice aspect of GDAL is that it allows to create read/write drivers around two functions: Open and CreateCopy. Open opens a datset read only, and to implement that all you have to do is to implement read access to your data using the GDALDataset interface. To implement CreateCopy, all you have to do is to create a new file reading information from a GDALDataset; then call Open on it. This means that there is no need to support incremental updates, and that all the data required to create a new file is readily available. This simplifies matters a lot. GDAL provides some interesting image manipulation functions, that can work over just these two calls. The way it does it is by exploiting the concept of a virtual dataset, which wraps an existing dataset but changes some parameters on the fly. This means that you can wrap a read only dataset with a virtual dataset that transforms it somehow, and then pass the virtual dataset to CreateCopy to save the transformed image in a format of your choice. On top of that, for many transformations you do not need to create your own virtual datasets, but you can use the functions provided by the VRT GDAL driver. One can learn a lot on how to use VRTDataset by reading the source code for gdal_translate. By doing that I could come up with this code for cropping an image, which is an interesting VRTDataset example.

GDALDataset* crop(GDALDataset* poDS, int xoff, int yoff, int xsize, int ysize)
 
        VRTDataset *poVDS = (VRTDataset*)VRTCreate(xsize, ysize);
        // Copy dataset info
        const char* pszProjection = poDS->GetProjectionRef();
        if (pszProjection != NULL && strlen(pszProjection) > 0)
                poVDS->SetProjection(pszProjection);
        double adfGeoTransform[6];
        if (poDS->GetGeoTransform(adfGeoTransform) == CE_None)
         
                // Adapt the geotransform matrix to the subarea
                adfGeoTransform[0] += xoff * adfGeoTransform[1]
                        + yoff * adfGeoTransform[2];
                adfGeoTransform[3] += xoff * adfGeoTransform[4]
                        + yoff * adfGeoTransform[5];
                poVDS->SetGeoTransform(adfGeoTransform);
         
        poVDS->SetMetadata(poDS->GetMetadata());
        // Here I also copy metadata from my own domain
        char **papszMD;
        papszMD = poDS->GetMetadata(MD_DOMAIN_MSAT);
        if (papszMD != NULL)
                poVDS->SetMetadata(papszMD, MD_DOMAIN_MSAT);
        for (int i = 0; i < poDS->GetRasterCount(); ++i)
         
                GDALRasterBand* poSrcBand = poDS->GetRasterBand(i + 1);
                GDALDataType eBandType = poSrcBand->GetRasterDataType();
                poVDS->AddBand(eBandType, NULL);
                VRTSourcedRasterBand* poVRTBand = (VRTSourcedRasterBand*)poVDS->GetRasterBand(i + 1);
                poVRTBand->AddSimpleSource(poSrcBand,
                                xoff, yoff,
                                xsize, ysize,
                                0, 0, xsize, ysize);
                poVRTBand->CopyCommonInfoFrom(poSrcBand);
                // Again, I copy my own metadata
                papszMD = poSrcBand->GetMetadata(MD_DOMAIN_MSAT);
                if (papszMD != NULL)
                        poVRTBand->SetMetadata(papszMD, MD_DOMAIN_MSAT);
         
        return poVDS;
 

This function wraps a dataset with a virtual dataset that crops it. Just pass the resulting dataset to GDALCreateCopy to save it in the format that you need.

At last night s #koha community hand-over meeting, it was felt that the LL/PTFS transfer could easily delay community control of the previous koha website for 3 months, so HLT should register a fan site for community members to put the latest Koha news and information on the web. So welcome to Koha-community.org which will mainly be about updates about new releases, support listings and the manuals until the previous website reflects the community again. Well done to all involved for getting it online so quickly!

It was easy to understand how to activate and set ACLs and I also understood the mask, which defines the maximum rights a user might get through ACLs. However, what I don't understand is, why this mask is dynamically calculated and set with every chmod (or equivalent) access... Perhaps I should describe my problem a bit more: For our (web-) developers Linux workstations I need to grant them some additional rights to the configuration files. As they always get no tasks and need to install new software (solved via sudo) and change the configuration, I so far thought the easiest would be to grant them write privileges on /etc (safe some special files like /etc/shadow of course). So I thought I could use ACLs to grant them rwx-privileges. Using default-ACLs (which are inherited when creating new files) it would be easy, if they e.g. install an own Apache web server, as the new directory and the new files inherit the default ACL. However, due to the mask being set to r-x, our developers still don't get write permission on /etc/apache2/foo. I tried to set a default mask, too, which seems to be inherited. But as soon as the package management does the equivalent of chmod 740 the mask gets recalculated and my ACLs don't work any more. Appearently, ACLs are not the solution to my problem :( But how do I solve my problem? And why is the mask always recalculated?

Agile Business Analysts? I have recently been fleshing out my thoughts on the role of business analysts in an Agile team. This is something I have historically addressed on a case-by-case basis with clients, but when thinking about it last week at a client site, I realized that I do have a general take on how the role and responsibilities of a business analyst change when a group moves to using an Agile process. Per the usual, I ll be using Scrum as the example process where needed. The Classical Roles Historically the business analyst has had two primary roles. In a process where cross-functional communication and collaboration are minimized, these roles are essential. What little communication that takes place between the business and the software development teams purporting to serve them passes almost exclusively through one or more business analysts. I view the business analyst in a phased development approach as having two roles; Translator and Gatekeeper. The Translator The business analyst is primarily tasked with taking the needs of the business and translating them into a written document that is then handed to the software developers and testers. This role s necessity is based upon some fundamental assumptions. First, programmers and testers are on the whole, too socially retarded to actually talk to the business persons and find out what they want. Second, the business is primarily viewed as being unable to focus its thought long enough to tell the programmers and testers what they need. Since a key value of an Agile approach to software development is to have individuals and their interactions be the driving force rather than processes an tools (see The Agile Manifesto), the obliteration of this role is a primary objective. The Gatekeeper Business analysts are also viewed the gatekeepers for requirements in classical, phased software process approaches. This is a somewhat unfair role, since they rarely have the authority to prevent business or technology changes, like a meter maid trying to stop an armed terrorist. It s even more unnatural since change always happens, so it s like the meter maid being assigned to an anti-terrorist unit with no additional training, authority, or weapons. The Agile Roles Some of the more extreme practitioners of Agile methodologies say that business analysts have no role in an Agile team. I disagree, provided the role of the business analyst is allowed to evolve. The roles I ve seen them take on as valuable members of an Agile team are Facilitator, Historian, and Journalist. The Facilitator A high degree of collaboration and interaction take place between business and technology in Agile Software Development. The business analyst can serve a critical need in these interactions, facilitating communication between business and technology and making sure that critical areas are being covered in an interaction. This can be a more fulfilling experience for an analyst, since they often talk to one side, then go to other to pass on the information, all the while thinking, Man, this would be simpler if you were both in the room with me at the same time. The Historian The business analyst s documentation skill is excellent for capturing significant information that often gets lost in a team of purely technical people. I have seen some great uses of business analysts in this area. One is to document the system that the team actually builds (not the big, up-front imagined system that is covered in a phased design stage). Another is capturing key technological and architectural decisions and the context in which they were made, so that when a group revisits certain items asking, Why in the world did we decide to do that? , they have the means to be reminded or informed why a particular path was chosen. The Journalist The business analyst can also be the team s journalist, making sure the latest information makes it out to all interested parties. One creative approach I ve seen at a client site is a business analyst who has a project blog. He posts entries after each meeting between the business and technology, documenting what new user stories were created (even scans in the story cards!), a summary of the discussions held, and documents any key decisions that might have been made. Oh, and he provides an excellent executive summary at the top that s suitable for anyone s review, right up to the CEO. Tell me you wouldn t love to have that guy in your Agile team.

Getting to X After having prepared a bunch of libraries, the idea was to get X in shape. I d like to thank Julien Cristau very much for his support all along the way, since I knew nothing about X. Basically the plan was to get:

• xserver-xorg-core-udeb: X itself.
• xserver-xorg-input-evdev-udeb: for input, as you probably guessed.
• xserver-xorg-video-fbdev-udeb: for output, as you also probably guessed.

Not much to say about the former, it only pulled two additional libraries compared to what was previously udeb-ified: libaudit0 and libselinux1. Since security modules aren t the primary concerns in this exercise, --enable-xselinux quickly became --disable-xselinux and those dependencies went away. Both drivers were trivial to tweak, nothing really worth noting there. Once those 3 additional packages added to localudebs/ and to pkg-lists/local, the mklibs tool really wasn t happy with them during the rebuild of the netboot image, since the drivers are X plugins, meaning they need symbols which aren t defined in other libraries. Example:

No library provides non-weak EventSwapVector

Since the idea is to get something working rather than getting something perfect at once, d-i s packages/mklibs/src/mklibs.py (installed as /usr/bin/mklibs) was told to stop raising an exception, and to print a warning message instead. Rebuilding an image was then working again. It took some time, but finally it got noticed those packages were actually all installed in the initramfs, so there was no need to try and load/install them manually during the installation process, everything was there already, and it was sufficient to switch to VT2 to see the /usr/bin/Xorg binary. Some packages were still needed: xkb-data, and x11-xkb-utils. The latter also pulled libxkbfile1, but well, an additional library doesn t sound very scary. At this point, one has to try and run Xorg. The result was a black screen, and no means of getting back to another terminal, be it inside Qemu or VirtualBox. Troubleshooting X First idea: adding the openssh-server-udeb package to the image, and start sshd from there. But then one has to go through the missing config file, tweaking privilege separation, generating a host key, setting a password, etc. And for whatever reasons, it wasn t possible to access it. That s where the DebianInstaller/NetworkConsole page on the Debian Wiki came into play: it includes a few steps to get a working sshd within d-i, painlessly. So, X was actually starting properly, but getting a black screen is now the default. One wants X -retro to get the well-known old behaviour. From udevadm info --export-db, it looked like input stuff were missing, and indeed, udev-udeb doesn t include /lib/dev/input_id and /lib/udev/rules.d/60-persistent-input.rules. Since scp seems broken, wget from the host did the trick followed by chmod +x for the input_id file, otherwise that doesn t work so well. But still, there were some missing evdev stuff. And indeed evdev wasn t to be seen in lsmod, nor available to be modprobe d. That s where it became a little dirtier: wget and udpkg on mouse-modules-2.6.30-2-amd64-di, followed by depmod to make sure, and then modprobe was OK. A little udevadm trigger run to ensure udev was fully aware of the new situation, and restarting X led to a moving cursor there, as well as a working keyboard, since it was possible to switch between terminals! Prove it! Alright, I m proudly providing you with a screenshot: (Really, I ve never been so happy to see this fracking black-and-grey-grid screen. I kid you not!) If we forget about the very few network-console-related questions answered to get an sshd running, please note that the installation process is still just starting, the very first debconf question is still unanswered. Now, patching udev again so that udev-udeb includes the two mentioned files was almost easy (details aren t important anyway), while pulling evdev.ko module automatically was trivial, just a matter of adding such a line to a packages list (I picked pkg-lists/netboot/amd64.cfg):

mouse-modules-$ kernel:Version 

Which means one can now, out of the box:

• Boot the ISO, select Install.
• Wait for the first question to appear.
• Switch to another terminal, and activate it.
• Run Xorg -retro there.

Next step Rework the Pango/Cairo/Gtk stack to use X11 instead of DirectFB.

The South Australian government wanted to force bloggers to disclose their name and post-code when commenting on an election [1]. According to Adelaide Now this included posts on Twitter and Facebook [2]. As expected there was a strong public reaction to this and Michael Atkinson (state Attorney General) stated that the law was not going to be enforced and that it would be retrospectively repealed after the election [3]. Which might have been the end of the issue, but he also said All MPs and all parties voted for Electoral law. Hope Libs, Greens, Family First, Independents etc will join us to support repeal which seems to be a clear claim that the Greens supported that legislation. The Greens are pretty good about freedom of speech issues so I immediately enquired as to what was going on. The following is from a Greens media brief issued yesterday by the office of Mark Parnell MLC (the Greens MLC in SA) which was issued before the new law was retracted: Attorney General Michael Atkinson must urgently clarify the scope of new electoral laws covering public comment on the internet, in the wake of concerns that the impact will be much wider than expected.

With only weeks to go before the election kicks off, there is an urgent need for the Attorney General to explain exactly what his laws are
intended to capture, said Greens MLC Mark Parnell.

When this was debated in Parliament last year, we were told that the law change would only affect electronic versions of a journal . Now there is
concern that the laws could extend further, restricting public commentary on media sites like AdelaideNow and ABC Online, he said.

When the Bill was debated in the Lower House, Michael Atkinson originally wanted to include ALL material on the web. However, the Government
backed down on this by the time it got to the Upper House, with Minister Holloway saying:
The intention is to limit the coverage of section 116 as it applies to the internet to electronic versions of a journal rather than any electronic publication on the internet.

A journal was narrowly defined as a newspaper, magazine or other periodical . Now I don t have a great objection to a law that demands that journalists identify themselves when commenting on an election, and I think that most people would not care about that. It seems that Michael Atkinson is repeatedly changing his claims to try and match popular sentiment. He has got form for this sort of thing, his past achievements include censoring the censorship debate about an R18+ rating for computer games [4].

• [1] http://etbe.coker.com.au/2010/02/02/precision-vs-accuracy-in-identifying-people/
• [2] http://www.adelaidenow.com.au/news/in-depth/labor-gags-internet-debate/story-fn2sdwup-1225825708827
• [3] http://www.adelaidenow.com.au/victory-atkinson-loosens-gag/story-e6frea6u-1225826104175
• [4] http://blogs.theage.com.au/screenplay/archives//010915.html

I have just installed an old Three mobile phone with 3G broadband for my parents home network access for the reasons described in my cheap net access in Australia post [1]. The first problem I had was that the pre-paid Three SIM just wouldn t work at all. I ended up phoning the Three support line and had a guy guess at which version of Windows I was running, after guessing every version of Windows from the last 10 years and Mac OS/X he finally asked what OS I use and then told me that Linux isn t supported. I said I HAVE TWO SIMS FROM THREE, ONE WORKS AND THE OTHER DOESN T, IT S ON THE SAME PC WITH THE SAME 3G ACCESS DEVICE, THE PROBLEM IS WITH THE SIM OR THE SERVER NOT MY OS . When the support guy discovered that one sim was pre-paid he said that there is a configuration difference, instead of an APN of 3netaccess for post-paid (contract) you have to use 3services for pre-paid. There are a bunch of web pages describing how to get Three 3G broadband working on Linux in Australia, some say to use 3netaccess and some say 3services. None of the pages I read stated correctly that 3netaccess is for when you are on a contract and 3services is for pre-paid. I ve submitted a suggestion for the Ross Barkman s GPRS Info Page (which seems to be the best reference for such things) [2]. After getting the pre-paid 3G SIM working for net access from the Huawei E1553 USB 3G modem I was unable to get it working from my LG U890 mobile phone. I never figured out how to solve this problem, I left my parents with the SIM that is connected to my $15 per month contract plan for 3G net access and am now using the pre-paid SIM for my own use. Of course this means that as I m using a SIM registered to my mother and she s using one registered to me I ll surely have some problems getting the support center to help me with problems in future. I found that the 3G net access got better reception when the phone was higher than the computer, so I used a USB extension cable to allow it to be placed on a shelf above the computer. The extension cable also allows it to be easily unplugged and plugged in again I ve already seen one situation where Linux got confused about the state of the USB device and replugging it was necessary to solve the problem. I was using Debian/Lenny. Here is my chatscript for connecting to Three with my 3G modem on a pre-paid SIM which also allows roaming to Telstra (I haven t tested whether pre-paid allows roaming, I ve only tested Telstra roaming with a contract SIM): ABORT 'BUSY'
ABORT 'NO CARRIER'
ABORT 'ERROR'
'' AT
OK ATQ0V1E1S0=0&C1&D2+FCLASS=0
OK 'AT+COPS=0,0,"3TELSTRA",2'
OK AT+CGATT=1
#OK AT+CGDCONT=1,"IP","3netaccess"
OK AT+CGDCONT=1,"IP","3services"
OK ATDT*99**3# Here is the ppp configuration for connecting via the USB 3G modem. For use as a permanent connection you want to also include persist and maxfail 0 :
/dev/ttyUSB0
230400
noauth
defaultroute
logfile /var/log/ppp.log
connect "/usr/sbin/chat -v -f /etc/chatscripts/three" For connecting with an LG U890 mobile phone you need to use ATDT*99***1# as the dial command and the device is /dev/ttyACM0 .

• [1] http://etbe.coker.com.au/2010/01/15/cheap-net-access-australia/
• [2] http://www.taniwha.org.uk/gprs.html

Biella passed this flyer around on the Debian-NYC mailing list. If you are in New York City this Friday, you won't want to miss this Eben Moglen's talk.

<object data="http://d1.scribdassets.com/ScribdViewer.swf" height="600" id="doc_178351046781277" name="doc_178351046781277" style="outline:none;" type="application/x-shockwave-flash" width="100%"><param name="movie" value="http://d1.scribdassets.com/ScribdViewer.swf"><param name="wmode" value="opaque"><param name="bgcolor" value="#ffffff"><param name="allowFullScreen" value="true"><param name="allowScriptAccess" value="always"><param name="FlashVars" value="document_id=26266103&amp;access_key=key-ie370f25lousi01fmdj&amp;page=1&amp;viewMode=list"></object>

If you can't see the embedded object, go here or to this post's permalink.

Andrew Dowdell and Michael McGuire have an interesting article in the Adelaide Now about censorship in the South Australian election [1]. The South Australian government wants to force everyone who comments on the upcoming SA election to provide their name and postcode. Attorney-General Michael Atkinson said the law was all about honesty . However a law that forces someone to not comment as Anonymous and instead forces them to use a name that sounds like something that might appear on a birth certificate and a postcode is not going to increase honesty at all. I think it s much better to honestly say I m not telling you my name than to lie and claim to be revealing your name. When the government forces people to give precise but totally inaccurate information it seems that it s going to be bad for everyone. It s even bad for the people who want to be identified, if most people who comment on a blog post are anonymous then I d like to be distinguished as the person who uses their real name! In the vast majority of cases the effort of determining who is using their real name will not be worth the effort. If someone comments on my blog under the name John Howard from Bennelong I will be very suspicious that they are using a fake name. But probably the vast majority of the English speaking population of the world wouldn t immediately identify such a name as fake. If someone wants to comment on my blog and they don t have a published phone number then I won t have any good way of identifying their address. If their name and postcode match an entry in a phone book then they still might be faking it they could take a random name from a phone book. If I was to demand that people who enter blog comments provide their phone numbers then I would have to pay the expense of phoning them as well as dealing with cases of people who are away from home, should I delay a blog comment for a month until the author returns from their European vacation? I have had a moderate amount of experience in writing letters to the editors of newspapers, and I have only once had an editor phone me to verify my details. In all other cases I guess I could have fooled them if I wished. I believe that there are already precedents regarding libel, if I approve a blog comment (or fail to unapprove it in a reasonable amount of time if it meats the automatic approval criteria) then I could be sued for libel if the contents of the comment are deemed to be suitably damaging. So if I was to try to get the real names of people who make comments on my blog then it wouldn t make it any easier for an idiot who wants to sue also anyone who wants to sue regarding an Australian political issue will probably find me a better target than most people who comment on my blog (a significant portion of whom are from the US and have much greater legal protections of their freedom of speech). Extending such a law to US based services such as Twitter is just silly. The stockholders and employees of a US based corporation can freely laugh at Australian censorship laws. Also it s pretty stupid to have a global scope on such laws, as a Victorian why should I care about the South Australian state laws? Implementing laws that can be easily broken inadvertently and can never be enforced against anyone who cares is just pitiful and will result in the MPs who vote for such laws being the object of derision. Vote for the small parties and independent candidates. Both Labour and Liberal want to censor us, put them in the second last two spots on your vote card! Now this law doesn t take affect until the writs for the March 20 election are issued. I encourage Australian bloggers to write bad things about the Liberal and Labour parties after the writs are released which are not libelous and which don t include your postcode. If they try to apply the new law then your blog post gets wide attention. If you plan to be a professional blogger then you could consider the $5,000 fine to be an advertising fee. If you don t want to be known then you can use a US based blogging service. I m going to continue to write political blog posts whenever I feel like it and I won t be telling anyone where I live. I will rely on the Streisand effect to save me. Update:
The politicians in SA have surrendered, the law in question won t be enforced and will supposedly be repealed after the election [2]. It will be interesting to see whether they really do repeal that law.

• [1] http://www.adelaidenow.com.au/labor-gags-internet-debate/story-e6frea6u-1225825708827
• [2] http://www.adelaidenow.com.au/victory-atkinson-loosens-gag/story-e6frea6u-1225826104175