Thunks

Suspend/Resume broken on Ubuntu

2010-01-03T21:59:00.002+11:00

Irritatingly suspend/resume stopped working around about the time I upgraded to Karmic. After going down a few blind alleys I eventually looked at dmesg and saw:

cannot set freq 16000 to ep 0x86

...which is due to a broken driver for the Quickcam 5000, which wasn't broken in Jaunty. Unplug the webcam and resume now works again.

ISP Filtering Live Report

2009-12-15T15:04:00.004+11:00

The ISP Filtering Live Pilot Report was just released. I've linked to an alternate URL since the site seems to be down. So what does it tell us?

"All six ISPs achieved 100 percent accuracy blocking the ACMA blacklist." Of 1000 URLs. Well, that doesn't seem very realistic since 1000 URLs isn't very many. But that's OK, there was another part to the test:

"In addition the above blacklist, filtering a wider range and volume of material to provide some level of protection to children using the internet." "In blocking additional categories of content all six ISPs achieved 78 percent to 84 percent accuracy when assessed against the test list of URLs compiled by Enex TestLab (Enex)." Hmm, that doesn't sound so good.

But at least there aren't blocking innocent material right?

"Testing was also undertaken against a list of content, prepared by Enex, considered to be innocuous and which should not be blocked by a filter. All participants experienced some level of over-blocking in this test (i.e. blocking of some legitimate URLs). All filters blocked less than 3.4 percent of such content."

Oh dear. Well, at least circumvention isn't an issue right? Because at least we can stop people from downloading prohibited material which is the point of the legislation (summary from wikipedia):

"A collection of both federal and state laws apply, but the most important are the provisions of Schedule 5 of the Broadcasting Services Act 1992 inserted in 1999 and since amended.^[6] Under this regime, if a complaint is issued about material "on the Internet" the ACMA is allowed to examine the material under the guidelines for film and video. If the material would be classified R18+ or X18+ and the site does not have an adult verification system, or would be refused classification, and is hosted in Australia, the ABA is empowered to issue a "takedown notice" under which the material must be removed from the site. If the site is hosted outside Australia, the site is added to a list of banned sites. This list of banned sites is then added to filtering software, which must be offered to all consumers by their Internet Service Providers."

From the report:

"A technically competent user could, if they wished, circumvent the filtering

technology."

Well... at least it didn't impact performance:

"Testing revealed that the three ISPs filtering only the ACMA blacklist had no noticeable performance degradation that could be attributed to the filter itself."

So it wouldn't actually stop someone if they were competent, and if you implemented a realistic sized list then you probably won't catch all URLs, and in any case you are probably going to get false positives. But at least it didn't affect performance. Although if you did want to actually prevent circumvention:

"As a general rule, there appears to be a relationship between measures to counter deliberate circumvention and impact on internet performance (i.e. stronger circumvention prevention measures can result in greater degradation of internet performance)."

Oh dear.

Social Networking

2009-09-29T12:12:00.004+10:00

Social Networking is certainly a big deal in the tech community, and it must be a big deal in the Internet community given Facebook's 300 million+ active users plus the innumerable other Social Networking sites, but somehow it has never clicked for me. I have a few blogs here and there which I update occasionally, but somehow it has never sucked me in.

There is definitely something important about the idea of Social Networking: humans are social animals and communication is an important part of our socialization. In fact our communities are defined by who we communicate with- from the detailed discussions with partners and work colleagues to the friendly nod to the-guy-at-the-bus-stop who's name you don't know but you see every day (except for rainy days when his wife drops him off at the train, which you saw one day when you missed the bus and had to walk to the train station).

And therein lies the rub: we are expert social animals and our everyday social communication is nuanced and subtle. We naturally track very subtle interactions with a range of different people (did the shopkeeper hear me discussing my tinea with my friend? oh yes, she made a comment about socks to the next person- wait no that was in response to a direct question about where they were- she still might have heard about my tinea) and we naturally present ourselves differently to different people in different contexts (yeah, this is a company offsite and I'm with the guys so we'll discuss getting drunk, but 24 hours later at work we won't have that discussion, except if we...).

None of that nuance exists in the online Social Network: we are rarely aware of who has read what, we don't present ourselves differently at different times (everything I've ever posted is available at any given time) and there is none of the everyday grouping of people which we do naturally and easily in the real world.

Social Networking as it stands today groups communication into: one on one private communication or public communication to everyone. And the tools to create more subtle groups (all the people within ear shot of my desk at work, the people I trust not to blab this sort of thing) are generally non-existent or far too much of a pain to manage. So at the moment social networking is basically a promiscuous process: hey everyone! here's everything about me!

That is effectively what it means to be famous. So if you want to be famous, Social Networking is currently an effective medium.

I think there are lots of people who don't want to be famous. Or at least there are lots of people who like more control over who knows what about them.

The alternative online is hoping for obscurity. Just a few friends or family read your blog/tweets/facebook and hopefully nothing happens to change that.

Its a shame that better options don't exist because increasingly the people you communicate with are distant and electronically connected. And those everyday electronically connected people don't have some of the everyday real world modes of communication: John looks hungover- best not to ask him any questions to day, Jane is spending a lot of time talking to Peter- I wonder if something is going on?

That's why tweets and status updates and blog posts are important and simultaneously look like junk. They are day to day pings of communication.

Hey! The-guy-at-the-bus-stop got dropped off by a different woman! I wonder what's going on there?

In Defense of Google Books

2009-07-13T13:45:00.003+10:00

In my continuing series of "Yeah what he said" posts:

http://www.thebigmoney.com/articles/money-trail/2009/06/23/defense-google-books

So I have to be clear: I work for Google, but I don't work on Google Books and I don't speak for the company. This blog is my personal reaction as a reader. I love reading and I'd really like to be able to get access to the back catalog of human thought. So I get very upset when I read articles in the popular press that decry or vilify the book deal because it is an easy headline. At high level it looks like it is a win for all parties.

Maybe I'm just coming from the wrong perspective but it seems like a good deal from an author's point of view. Author's can opt at any time to change what information about their book is presented and Google Books otherwise acts as an additional distribution method. And even better, books can continue to be available and in distribution even after they are no longer being printed in dead tree format- that seems like a big win! There seems to be some concern about pricing but I find hard to believe Google gain anything by ripping off the customer (who won't buy) or the author (who will withdraw the right to sell). Did I mention that the authors can actually make money out of a book that is no longer available in dead tree format?

Publishers seem to profit to me too since for books that ARE in print they'll get extra promotion. Book stores benefit since in print books will point to a book store where you can buy it.

There is even a concern that Google will have some kind of "monopoly" on information. Well, given that Google has now forged the prototype agreement it should be much cheaper for someone to go down this path again breaking the 'evil' monopoly. But this actually makes information that would otherwise not be accessible open to readers again. And for that it should be applauded and supported.

Google is not beyond reproach and like any large organisation its actions should be inspected closely. But many of the critiques of the Google book deal that I've read so far don't offer much thoughtful analysis and merely cloak the deal in FUD. As a reader I would be angry to see the huge back catalogue of out of print books snatched from my grasp without very good reason. And so far, I haven't seen any good reasons.

ChromeOS

2009-07-13T12:48:00.003+10:00

I agree with this article:

http://www.businessweek.com/the_thread/techbeat/archives/2009/07/googles_chrome.html

I don't think ChromeOS is an attempt to spit in Microsoft's eye. ChromeOS simply reflects a real user behaviour which is 'I just want a light laptop to browse the web with- I mostly read email, go to ebay, watch youtube videos, lookup wikipedia and I'd like to do that while sitting on the couch.'

I've had people ask me if this kind of hardware existed and I've said "not yet, but it will come". ChromeOS enables this new hardware in the same way that Android enables a new generation of mobile phones.

Yes, Google can justify this for business reasons- more people using the web tends to mean more income. But my observation is that internally the real justification for most projects in Google is that this is good for users and an interesting technical project- and incidentally it also has a business case. (I don't have any special knowledge about ChromeOS from internal documentation- I haven't got around to looking.)

Australian Private Fibre

2009-01-28T15:40:00.002+11:00

I don't know if this has been considered by the Australian government but I've been wondering about an interesting use for some of this financial stimulus money. In short: subsidised private fibre to the node.

The Australian government provides a subsidy for individuals to plumb fibre optic cable from their house to the nearest node or junction. The tricky part is the legislation which controls how the fibre is connected and used at the node. But oddly enough, this is a problem the Government already has.

The Broadband Fibre to the Node program has a huge problem: Telstra owns the nodes. The current non-Telstra contenders all assume that somehow they'll be granted access to the nodes. But the nodes are owned by Telstra. Government intervention will be required one way or another- either to un-privatise the nodes or force Telstra to allow access.

But another option would be to create this infrastructure anew using modern infrastructure. And the timing is surprisingly good: Passive Optical Networks make fibre infrastructure cheaper and we happen to want to spend money right at the moment.

The current Keynsian Economic thinking calls for a stimulus to the economy: sensible Government led investment prefereably in productive assets. In an ideal world this investment would have long term benefits, be socially equitable and have broad geographic impact. Oh, and we'd like it to happen sooner rather than later. The classic stimulus is building roads. The 21st century equivalent has got to be Broadband networks.

I'm not suggesting that this is an easy suggestion. Maybe it has already been thought through and dismissed. But if not- well it is certainly worth the thought.

Blogging: content generation made easy

2008-12-02T23:53:00.002+11:00

Blogging has made content publishing a much easier process. The technical challenge has been reduced to a sign up process and typing out a document. Just as importantly blogging provides a conceptual framework for thinking about publishing.

The technical changes are obvious: no need to worry about html and css, or creating DNS entries or webserving (or any other mix and match combinations of hosting). The signup process for blogging is now more or less: what would you like to call this blog and what sort of colour scheme would you like?

Less obvious is how much the conceptual process has been simplified. My Space and Facebook have been very succesful in part because they have provided a framework for what to say as well as how to say it. They have lowered the barriers to entry by constraining the up front choices required to get started. Blogging is slightly less socially constrained but still provides some of the simplifying assumptions: I'm here to write about this topic- which could be me, or it might be about sewing or Belgian politics. But I don't have to think much about the site design, or access permissions, or social networking until I want to- I can start by posting and think about the rest later.

Mediasaurs

2008-11-12T11:29:00.002+11:00

I've not really read any Michael Crichton (RIP 4 November 2008) so I can't say whether I am a fan or not but I was fairly impressed by just how prescient his article Mediasaurus was (via Techdirt). Most of what he said is now pretty much established fact (outside the media industry)- the only exception being his suggestion that people will become more prepared to pay for information:

More and more, people understand that they pay for information. Online databases
charge by the minute. As the link between payment and information becomes more
explicit, consumers will naturally want better information. They'll demand it,
and they'll be willing to pay for it. There is going to be - I would argue there
already is - a market for extremely high-quality information, what quality experts
would call "six-sigma information."

Actually, more or less the reverse has happened. Information is an infinitely reproducible good which means the price of information tends to zero (as the cost of distribution goes down). That doesn't mean the cost of assembling the information goes down. This is as true of high quality information as it is of low quality information.

The tricky part of getting good information is finding it.

The short answer here is that Google and wikipedia have made whole classes of information readily discoverable- the cost of discovery is cheap. So if the information is free somewhere that becomes the price of the information everywhere. Without a price fixing mechanism and collusion competition forces the price of information to zero.

Aside from that relatively minor quibble he's got a pretty impressive hit rate given that he wrote the article in 1993. The funny part is the media companies still don't seem to have got it.

In defense of blogging

2008-10-14T18:00:00.004+11:00

"Blogs are full of crap."

I've heard this quite a few times over the last few weeks. This is especially ironic since I'm about to start working on Blogger- so I guess I'm partisan- I've generally managed to argue the complainants around to a different perspective: they aren't crap, they just weren't written for you.

Actually I tend to start the argument by saying there are two classes of blogs. The first group is made up of the utterly fabulous well published blogs that you read even though you might not realize they are blogs. Things like engadget or "I Can Has Cheezburger?" or digg (really, they are blogs). Or for something topical have a look at Paul Krugman's Blog (this year's Nobel prize winner in economics) or the NPR Plant Money Blog. Have a look through Wikipedia's List of Blogs and be amazed by how well written these blogs are. This kind of blog can be characterized by being a bit specialised. They have particular focus, they are generally written by experts or interested amatuers and the quality or writing and/or editing is a cut above the general mass of blogs. (Even I Can Has Cheezburger? is very consistent even if you don't like lolspeak- if nothing else the editing is effective.)

On the other hand the rest of the blogs seem to be made up of all sorts of dross: endless pokes and superpokes, mindless conversational noise and the dull minutae of boring lives. Its not the kind of blog you'd read, although maybe, its the kind of blog you'd write.

These sorts of blogs aren't written for you. They are written for friends of the poster and they are sort 'weakly' informational- intended to say "Hi, I'm alive, I'm doing some stuff." They are a way of maintaining community.

Community is about the people you communicate with in all the subtle ways we communicate. Not just explicit conversation but the observation of who is coming and who is going, the frilly underwear on a washing line or the car in the garage. Humans are naturally social animals and we automatically collect these observations into a model of the people with whom we interact. But our modern population is very mobile (we move to a different city for work, our parents retire to the sea side, our next door neighbour goes to University) and we lose these simple observational signals when someone is out of eyeshot. They are being replaced by more deliberate but equally insignificant communications- things like blog posts.

Piecing together the individual blog posts, pokes and chats gives us a shape of someone's life- in all its everydayness. And they largely hold no interest to someone who doesn't want to know the person. But if they are a friend who you'd like to keep in your extended community then the everyday minutiae are simple signal that they are alive and breathing.

The Google blog recently posted on this topic (Social Web: All about the small stuff) in what is a much more elegant discussion of this sort of idea.

Ultimately the complaint about blogs comes back to a more general point that Clay Shirky has so nicely made: its not a problem of content, its a problem of filtering. Or more crudely: if you don't like it don't read it- it wasn't meant for you anyway.

Mythtv and DVDs

2008-07-13T20:28:00.002+10:00

Mythtv supports the playback and ripping of DVDs. Under ubuntu the mythdvd package supports playback of dvd, ripping and transcoding. (There used to be a separate myth transcoding package which mythdvd supersedes.)

Mythdvd assumes the dvd device is /dev/dvd, however the udev rules in Ubuntu don't necessarily generate a /dev/dvd link to the cdrom which has lead to a number of questions on Ubuntu forums about being unable to playback dvds.

Under ubuntu the /dev devices (nodes and symlinks) are automatically generated under udev, so even if a link is made by a user it will be lost on reboot. Instead, what you'd like to do is make sure udev generates its own symlink from /dev/dvd to the real device.

Udev (in Ubuntu) uses rules in /etc/udev/rules.d. The two pertinent rules sets are:

/etc/
/etc/udev/rules.d/70-persistent-cd.rules
/etc/udev/rules.d/75-cd-aliases-generator.rules

The cd-aliases-generator rules file automatically generates the persistent-cd rules file which is executed to create cd link aliases.

The short solution to this for me was to edit the persistent cd rules file and duplicate one of the entries which created symlinks

# This file maintains persistent names for CD/DVD reader and writer devices.
# See udev(7) for syntax.
#
# Entries are automatically added by the 75-persistent-cd-generator.rules
# file; however you are also free to add your own entries provided you
# add the ENV{GENERATED}=1 flag to your own rules as well.
# TSSTcorpCD-RWDVD-ROM_TSL462C (pci-0000:00:1f.1-ide-0:0)
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-ide-0:0", SYMLINK+="cdrom", ENV{GENERATED}="1"
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-ide-0:0", SYMLINK+="cdrw", ENV{GENERATED}="1"
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-ide-0:0", SYMLINK+="dvd", ENV{GENERATED}="1"
# CDRWDVD_TSL462C (pci-0000:00:1f.1-scsi-0:0:0:0)
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-scsi-0:0:0:0", SYMLINK+="cdrom1", ENV{GENERATED}="1"
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-scsi-0:0:0:0", SYMLINK+="cdrw1", ENV{GENERATED}="1"
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-scsi-0:0:0:0", SYMLINK+="dvd1", ENV{GENERATED}="1"
ENV{ID_CDROM}=="?*", ENV{ID_PATH}=="pci-0000:00:1f.1-scsi-0:0:0:0", SYMLINK+="dvd", ENV{GENERATED}="1"

Note that the last line was one I created by copying the previous line and editing the 'dvd1' down to 'dvd'. This exact file isn't going to work on your machine since the path identifier ("pci-0000:00:1f.1-scsi-0:0:0:0") is hardware dependent so likely to be different on different machines, but the principle is the same.

As far as I can tell the 'ENV{GENERATED}="1"' tag at the end of the line prevents the entry from being deleted.

For a better understanding of udev rules you might try this link.

Ubuntu Removable Drives and Media

2008-07-12T22:00:00.002+10:00

I have had a photo importing script that worked very well- automatically executed when plugging in the SD card from my camera, importing photos, rotating, filing in date folders and copying to a network drive. The automatic behaviour was kicked off by changing the default application in the Preferences control panel 'Removable drives and media' in Ubuntu.

This worked very well- until it didn't. I just worked out what went wrong and this post is basically Google fodder for any other poor unfortunates who may have been struck by the same problem.

Basically in Ubuntu Hardy (8.04) Nautilus was changed over to use GVFS (the gnome virtual filesystem) and GIO (Gnome I/O abstraction layer) it also took over the functionality of handling certain media types- including removable media like photo CDs, digital cameras and so on. You can find this from the menus in any Nautilus window: Edit > Preferences > Media. Sadly this functionality is significantly less functional for my purposes. Where before I could simply substitute the path to my import script to get similar functionality I had to:

As root create a new 'my-import.desktop' file in /usr/share/applications (example contents see below)
Open nautilus media preferences (from the menus in any Nautilus window: Edit > Preferences > Media) and select 'Ask What To Do'
Plug in the memory card and select the 'Open My Import Script' option and click the 'always use this option' checkbox

[Desktop Entry]
Version=1.0
Encoding=UTF-8
Name=My Import Script
GenericName=Import Script
Comment=Run my import script
NoDisplay=true
Exec=import_script %u
Icon=f-spot
MimeType=x-content/image-dcf;x-content/image-picturecd
StartupNotify=true
Terminal=false
Type=Application
Categories=Graphics;Photography;GNOME;GTK;

These bug reports led me to the solution. There is some discussion about a fix for this backward step in functionality- hopefully their feedback will lead to an upstream fix in Gnome or a standing patch in Ubuntu to return something like the old behaviour.