Planet Compsoc

Fred: Ogg/Vorbis on N810

2008-08-19T20:59:29+00:00

There's easily findable tutorials on how to build the vorbis codec in scratchbox; however, you don't need to :)

Just add the chinook extra's repository, then as root, apt-get install gstreamer0.10-plugins-ivorbis (integer-only decoder). This package works fine on diablo.

kobrafang: My Trip to Leicester

2008-08-17T22:03:52+00:00

I went to Leicester this weekend. I had a very good time with some very good people. There are a few things to note.

I spent most of my time in a place called Kirby Muxloe. The focal point of this little village is undoubtedly its half-finished castle. A castle that remained half-finished, despite being rebuilt. The restoration work for the castle didn’t include extending and finishing it…

I also got to wander around Leicester City Centre, in a limited radius from the train station, and I was driven around it briefly, yet for long enough to see that Leicester is massive. But even more important than that is that this nice wonderful city managed to make me second-guess my preliminary evaluation of the attractiveness gradient of the country.

When I went to Scotland, it was the final piece of evidence I needed to form a basic trendline between London and the northern border of England. This trendline showed that, approximately, the average attractiveness of girls decreased the further north one travelled.

Well, that didn’t work out so well, since Leicester is further north than Coventry and Birmingham, but there was a really good proportion of reasonable-looking girls. So, my trendline probably isn’t vertical up the map. I will keep you posted when I work it all out, though. Maybe one summer I will be able to tour some of the country’s more prominent towns and cities and take more scientific readings.

Most importantly, however, was the fact that there was no god whatsoever in our weekend. We had no rain during the hours we were outside with the barbecue or walking the dogs (and planning to storm the castle). We had an amazing night with cheap wine and card games, peppered throughout with good-natured jokes about those who couldn’t be there. Come morning, there was an almighty fried breakfast created (I would go as far as to say it was the biggest breakfast ever) and everyone had an excellent, incident-free time.

Rock the fuck on!!

Fred: Maemo scratchbox on amd64 multilib

2008-08-17T15:21:28+00:00

Regarding TSDgeos' post, if you have working 32-bit multilib, you just need "linux32 sh installer-goes-here.sh". If you don't have that binary in your distribution, the source is in the slamd64 source archives among other places. Thanks to Andi Kleen for writing this incredibly useful little launcher back in 2002 :)

It's also useful for various games installers.

Fred: Akademy

2008-08-17T11:49:24+00:00

Firstly a big thank you to Trolltech^WNokia, and all of the team who organised the event - probably the best one I've been to so far.

As for what I did there:

Drank lots of Belgian Beer :)
Learned about new stuff in Qt, KDE, and Linux graphical stuff in general - in particular, Zack Rusin's talk on Gallium3D seems very promising
Added MPRIS support to YANIHP - so it now support the same D-Bus interface as Amarok, VLC, Xmms2, and others
Added a tray icon to YANIHP
Received a free N810 from Nokia...

...which led to:

Surprisingly few changes were needed:

Added a compile-time option to make it so that left-clicking on the tray icon pops up the context menu, instead of hiding/showing YANIHP - I'll later make this apply to all context menus, where appropriate
Added a compile-time option to hardcode data and music paths - otherwise YANIHP uses QDesktopServices::storageLocation - which seems to think that /usr/share/mime/data/ is a perfectly suitable location to store non-root user data on the N810...

Here's how to get it running:

Install taglib (1.5), libqtcore4, libqt4-sql-sqlite, libqt4-phonon on the device, and the corresponding -dev packages in scratchbox
Also install cmake in scratchbox
Build YANIHP in scratchbox, as normal, with the following options:
- EMBEDDED_FIXED_DATA_PATH=/home/user/.yanihp - this changes where the database, coverart, and lyrics are stored
- EMBEDDED_FIXED_MUSIC_PATH=/media/mmc2/Music - this just changes the default location, the user's asked if they want to change it
- EMBEDDED_USE_FIXED_PATHS=ON - enables the above two options
- EMBEDDED_LEFT_BUTTON_CONTEXT_MENU=ON - makes it so that left clicking on the tray icon shows the context menu instead of showing/hiding YANIHP
- WITH_Mpris_PLUGIN=OFF - not strictly neccessary, but I doubt you've got use for it
Copy it across :)

There's a few caveats:

It's only currently usable with the CleanLooks Qt theme - with small font sizes and so on, not with GTK support.
Phonon-gstreamer doesn't appear to support manually switching between gstreamer audio sinks - in particular, there's no mp3 codec as such - instead, you need to connect your audio source to "dspmp3sink" instead. Alternatively, you can use another phonon backend.

I'm not uploading packages, as I've got no idea how to make a "good" debian package.

Bucko: Re: Prisoner's Dilemma

2008-08-15T07:17:41+00:00

Seems no-one responded to the simpler problem either. I guess the solution is to produce programming related problems instead of logic/maths related ones.

Well, here's the answer: There's 8 possible arrangements of hats. Assuming true random distribution, just by writing down the possibilities we see that: If you see 2 hats of the same colour, the chances are that yours is the other colour (3/8 instead of 1/8). So upon seeing two hats of one colour, you should guess the other colour - and if you don't, you should pass. This gives the team a 3/4 chance of success, since in every arrangement someone sees two hats of the same colour. The key is that you're actively betting against the two "all hats the same" possibilities.

There's a clear counter to this strategy, though - just give everyone the same colour hats. However, we can alter the above strategy to involve a uniform random element to prevent this. Call our people X, Y and Z and write a hattage as a 3-tuple with 0 and 1 as the colours. Then the strategy above is just betting against the two opposite hattages (0, 0, 0) and (1, 1, 1). You can just arbitrarily pick a hattage and its opposite to produce a new strategy, so for example picking (0, 1, 1) and (1, 0, 0) we find that the strategies are as follows:

X: If I see two of one colour, I pick that colour.
Y: If X has 0 and Z has 1, I pick 0. If X has 1 and Z has 0, I pick 1.
Z: If X has 0 and Y has 1, I pick 0. If X has 1 and Y has 0, I pick 1.

The possible hattages are:

(0, 0, 0): X wins, Y, Z pass.
(0, 0, 1): Y wins, X, Z pass.
(0, 1, 0): Z wins, X, Y pass.
(0, 1, 1): All three lose.
(1, 0, 0): All three lose.
(1, 0, 1): Z wins, X, Y pass.
(1, 1, 0): Y wins, X, Z pass.
(1, 1, 1): X wins, Y, Z pass.

Thus, once again we have a 3/4 chance of winning! Thus if the tablers secretly agree on a random hattage and its opposite before they are given hats, they will win more often than you'd expect.

This kind of strategy can be extended to higher numbers of people but since no-one was really bothered by the 3 person case, I won't detail the higher numbers here. It's available on the interweb anyway.

I'll try to come up with some interesting computational problem for my next one.

Bucko: AmeCon

2008-08-15T06:59:01+00:00

Oh, hi there.

kobrafang: The Monarchy is Archaic

2008-08-14T01:19:08+00:00

It’s very historical and fun and tourist-attracting, but it’s difficult to take the monarchy seriously anymore.

The Queen is purely a figurehead anyway. Things are decided by Parliament; I’m sure we all know that. There really is no point in having a Queen, other than it makes printing and minting money that much easier, since you never have to make a decision about what’s on the back of the coin.

The trouble is, we’ve come entirely too far to do anything about changing it. It would require nothing short of a flat-out revolution to change these things, and as far as I know, people are too frightened, lazy and contented to revolt these days.

If times were bad, everyone would love a revolution. But currently, the situation is that people are perfectly happy with the way things are. Everyone has their little complaints, but by far not enough to do anything about them (other than complain). Things are “ok” how they are, and that seems to be all that most of us need.

Of course, one revolutionary isn’t going to make a huge difference. We need lots and lots of people to kick things off, and I suspect that such a thing won’t happen in the Western world, ever. If nuclear war were pending then it might galvanise people into action. But nothing short of the threat of impending apocalypse is going to work.

Technocracy

But what if it did? Wouldn’t that be great? The space-race, nuclear war, chemical war, electronic war (Die Hard 4.0 style) or some other modern cause for panic that’s faster than global warming starts, and everybody goes crazy. What will happen? Well, if we’re lucky, people will turn to the scientists to protect them. If the populace gets desperate enough, they might turn to something they don’t understand to salvage their world. Presently, we are more than happy to elect leaders based on their persuasive words, not their persuasive qualifications. Tony Blair studied jurisprudence at Oxford. The theory and philosophy of law. What I want is a prime minister who studied physics or maths. Maybe the world would turn to men of science in its time of desperation.

With more and more parts of our lives being influenced by technological advances, why are we not searching for officials who excel in scientific fields? Get the right people and they can run the world like a well-oiled machine, not the festering ball of corruption and politics and war and religion that it is now. When we get to the fucking moon, do you think the astronauts are going to let just anyone come and live in their perfect new paradise? Fuck no! They’re going to let the war-mongers and the politicians and the bureaucrats and the kings and queens and the presidents all kill each other on Earth (or, as it will be referred to then, the Doomsphere) with their nukes and greenhouse effect while the Moonians (or maybe lunar-tics?) will instate a High Technocrat, a man of (computer) science and logic, to be the head of the state.

I vote: me.

kobrafang: Sick State

2008-08-13T19:08:46+00:00

The days of advertising your product are long gone. Recently, unless you can offer up-to-date comparisons of your product with your competitors’ product, you don’t stand a chance.

And then, of course, there are those who don’t actually have a product on the market, but specialise in the comparisons. The trouble is, there are so many comparison websites out there, which one should you use? They all have their advantages and disadvantages. So how do we know the comparison site we’re using is the right one?

If you can’t see where I’m going with this, you are a very generous reader to slow your brain whilst reading and receive the full force of my point.

Why doesn’t somebody just launch a comparison website comparison website? You know, comparethecomparisonmarket dot com or comparisonsupermarket.com?

If now, you’re thinking “well, why stop there? Won’t we then need comparisons of the comparison websites of the comparison websites?” then you are absolutely right. There’s no end to the potential distance we could get from companies selling actual products and services. It’s fashionable to not sell anything at all. At best it could be called the sale of convenience, but I refuse to believe that comparison websites are as accurate as checking all the insurance companies “by hand” and getting the prices and features yourself. And sacrificing accuracy for convenience is just bad practice. Television advertisements cost a fortune and you only really see well-known brand names and corporations purchasing advertising time, particularly in prime-time evening viewing. So that means these comparison websites are making enough to be able to firstly afford the adverts, and secondly to justify roping in more people (meaning that, somehow, these bastards are making money out of nothing).

Just checking a few of the websites I know off the top of my head, comparethemarket.com and moneysupermarket.com had adverts. Not very obvious or intrusive ones, nor were there many on each page. Gocompare.com did not have any adverts, neither did Tescocompare.com, and uSwitch.com’s didn’t come out until a few pages into their site. So that’s not where they get their revenue from. Then, of course, it twigged.

uSwitch.com has agreed deals with some suppliers across all our services to receive a small commission payment when a customer chooses to switch or apply for a product through us.

uSwitch.com

Each insurer pays us an introductory fee per sale. We don’t charge you in any way - not even a penny, and unlike some other price comparison sites, we do not charge insurers to advertise on our site, they only pay us when a sale has been made.

GoCompare.com

And there we have it. It’s plain, old-fashioned salespeople commission. Only salespeople have been replaced by faceless websites, because you can’t doubt a website. You can doubt a salesperson, and each salesperson has a given persuasive power. They’re not certain. They’re flawed. But a website? That can be an engine, something incorruptable.

But it’s all lies anyway, so fuck it.

Blood God: Spooks Code 9

2008-08-11T17:00:34+00:00

So, last night saw the first couple of episodes of the new BBC drama, Spooks Code 9. I’ve always quite enjoyed watching the usual Spooks series (even if only to play the game of which lead character is going to get killed off next), so I figured I’d give this new one a shufty.

The premise for this series is that a nuclear bomb has gone off in London during the 2012 Olympics, and the series follows a set of agents recruited in the aftermath trying to do what the Spooks lot usually do, stopping terrorists performing their nefarious deeds. The episodes are shorter than those of the original series, clocking in at 50 minutes rather than an hour, and accordingly there seems to be less that happens in the episodes.

So far the plots in the episodes haven’t been particularly enthralling, although they have set the wheels in motion with the requisite long-running plot. Unfortunately this, along with a number of other things in the first two episodes, have been rather clichéd.

I’ll probably still end up watching the other episodes in the series (if only because I’ve set my BT Vision box to record the series) and perhaps it will get better, but I’ve not really got high hopes after this first pair of episodes.

kobrafang: BioMars

2008-08-11T01:33:11+00:00

I think it’s about fucking time someone put a fucking biodome or something up on Mars. I mean, I know it’s freezing up there and everything, but come on, you lose nothing by blasting a few thousand potted plants covered in bacteria or fungus or a bunch of amoebae up there just to see what happens. Get a big old bunch of solar panels and rig them up to some heaters, I don’t know, you’re NASA for fuck’s sake, you’re supposed to come up with these awesome ideas.

All we have are pictures of Mars. Well, observing is all well and good, but if I know humanity, we’re much more inclined to poke our fucking noses into stuff and interfere than we are to sit back quietly and watch. Or, fucking hell, even on the moon. People can walk on the moon, it gets plenty of sunlight and face time with mother Earth, just make a massive plastic dish and attach the bastard to the face of that bland rock. DO IT!!