Slashdot Mirror

1. - the healthcare (when someone tries to kill you and almost succeeds, you don't want to be bothered about health insurance, co-pays, and any ridiculous limits on hospital stay or benefits!)... I have excellent healthcare through my employer, but my wife couldn't afford the premiums (she was a contractor for quite a while, and the costs for personal coverage are ... shameful... IMHO. Fortunately, my employer allowed me to have a 'domestic partner' covered, so I could get her on my insurance pretty easily. I still get the feeling we exploited a loophole to get medical cover, though, and frankly, that sucks. Where I'm from, "free" (read: low-cost and deducted from your pay-check so you don't miss it) excellent healthcare is a right, not something you worry over.
2. - education. My physics B.Sc. was free, I think I owed about £1000 on my bank overdraft at the end of the three years - pretty much all of that being beer money :-) and I actually got *paid* by the government to do a PhD. My US-born wife ended up with ~$70k of student loan for her two degrees (3 if you count the JD/MBA as two degrees). All paid off now, but that's still a huge difference.

Compare and contrast that with the US, and I can see your point. On the other hand, CA has nicer weather. I doubt I'll retire in this country (we'll probably go back to the UK) but it's a nice place to live, even if I do have to be sexually molested every time I come into and out of the country.

Simon

Funnily enough, I was in a motorbike accident myself before I came over to the USA. Nowhere near as bad as your own (I was very lucky) but I was in hospital for a couple of days, rode there in the ambulance, had police and fire trucks called out to the scene etc. There was no charge, and it didn't cross my mind that there would be...

Any (every ?) government gets a lot of flak for pretty much anything it does - you can't please all the people all the time and all that, but at the end of the day, they're not trying to make a profit. Any private institution has to run all the same risks, spend all the same money, and also make a return on the investment. Normally I'm fully behind this as a great motivator for the company concerned, but when the easy option is to simply screw the "customer" in order to turn a profit, I'm not so sure.

In any event, the point of my post wasn't about people like you and I, with good medical insurance coverage. It was because I don't believe *anyone* should be concerned about medical coverage, even if that costs me something. That, I think, is a big cultural divide between the US and the UK on this matter, not just the public/private debate.

My fiancee is in fact more-qualified than I, she has a JD/MBA. However, she is still paying off student debts (another thing I didn't have to worry about in the UK, but that's another rant altogether :) and has only got a position as a long-term contractor; she would have had to pay her own medical insurance without any company aid, which (even with her income) is simply ridiculously expensive. If a well-educated well-to-do person can't afford medical insurance, something is rotten in the state of Denmark...

As far as the argument that you don't trust the government because of its past performance, it seems you do trust an insurance company, despite all evidence to the contrary of how they behave when you need them to pay up. Anyone who's been involved in a car accident would probably attest that (a) they screw you if they can, and (b) they screw you later by increasing your premiums, even if they somehow didn't manage to screw you via (a).

On top of that, Medical insurance agencies have come up with (c), a new evil: "recission". This is where they go back through your file looking for any possible (no matter how tenuous) excuse to retroactively cancel your insurance (even after payment has been initially made), leaving you with the huge bill that you might even have thought was already paid, and no possibility of getting any medical insurance in the future. I read of a case where a fall by the pregnant mother cancelled a policy by the adult daughter when the daughter developed vision problems at age 27.

I'm sorry, but that just sucks. Really. Really. Really sucks.

Simon

Apple does NOT (repeat that, NOT) prevent people from developing open-source applications. The FSF's rant was just that, an uninformed screed directed at a company that doesn't play ball with the FSF's politics.

Proof: Read the first link. I downloaded some source from the 'net, I compiled it, I modified it and compiled it again, then I installed it on my phone and it works just fine.

I had an email exchange with the author of the FSF's rant, and pointed out his errors. I think he and I still disagree, but to not even acknowledge the possibility that FOSS s/w is just fine and peachy on the iphone is intellectually dishonest. Not that that will stop the crazies from apple-hating... [sigh]

Simon

Apple does NOT (repeat that, NOT) prevent people from developing open-source applications. The FSF's rant was just that, an uninformed screed directed at a company that doesn't play ball with the FSF's politics.

Proof: Read the first link. I downloaded some source from the 'net, I compiled it, I modified it and compiled it again, then I installed it on my phone and it works just fine.

I had an email exchange with the author of the FSF's rant, and pointed out his errors. I think he and I still disagree, but to not even acknowledge the possibility that FOSS s/w is just fine and peachy on the iphone is intellectually dishonest. Not that that will stop the crazies from apple-hating... [sigh]

Simon

Something that no-one seems to pick up on is that the ad-hoc certificates basically allow full open-source distribution without any recourse to the app-store. See my blog post for more details.

Simon

I did this as well - an (old) picture of the dish is here - that was back when I lived in London - who needs a back yard, anyway ? :-)

Simon.

Perhaps you didn't get much out of neural networks, but my PhD thesis was on the similarities between a kohonen network and relaxation-labelling equations. Part of it is up on my blog (I haven't actually got as far as that bit yet, but the groundwork is there).

A neural network (well, anything more complex than the single-layer perceptron anyway) is an arbitrary classifier. I'm curious as to why other methods are "much better". Unless you do an exhaustive search of the feature-space, all classifier methods are subject to the same limitations - local maxima/minima (depending on the algorithm), noise effects, and data dependencies. All of the various algorithms have strengths and weaknesses - in pattern recognition (my field) NN's are pretty darn good actually.

It's also a bit odd to just say 'neural networks' - there are many many variants of network, from Kohonen nets through multi-layer perceptrons, but focussing on the most common (MLP's), there's a huge amount of variation (Radial-basis function networks, real/imaginary space networks, hyperbolic tangent networks, bulk-synchronous parallel error correction networks, error-diffusion networks to name some off the top of my head), and many ways of training all these (back-prop, quick-prop, hyper-prop, batch-error-update, etc. etc.) I guess my point is that you're tarring a large branch of classification science with a very broad brush, at least IMHO.

Not to mention that this is all the single-network stuff. It gets especially interesting when you start modelling networks of networks, and using secondary feature-spaces rather than primary (direct from the image) features. Another part of my thesis was these "context" features - so you can extract a region of interest, determine the features to use to characterise that region, do the same thing for surrounding regions, and present a (primary) network with the primary region features while simultaneously(*) presenting other (secondary) networks with the features for these surrounding regions and feeding the secondary network results in at the same time as the primary network gets its raw feature data. This is a similar concept (if different implementation) to the eye's centre-surround pattern, and works very well.

If you work through the maths, there's no real difference between a large network and a network of networks, but the training-time is significantly less (and the fitness landscape is smoother), so in practice the results are better, even if in theory they ought to be the same. I was using techniques like these almost 20 years ago, and still (very successfully, I might add) use neural networks today. If it's a fad, it's a relatively long-running one.

Simon.

(*) In practice, you time-offset the secondary network processing from the primary network, so the results of the secondary networks are available when the primary network runs. Since we still run primarily-serial computers, the parallelism isn't there to run all of these simultaneously. This is just an implementation detail though...

My suggestion is that they deliver what they advertise. That would suit me. It's too late for Comcast now, at least for me, though. I was never a heavy user of P2P, but I was pissed off about losing iChat...

Simon

At least, from my perspective. I'm not a huge user of P2P, my ire is more directed at the violation of the principles that founded this 'internets' thing. If we let company-interests direct the future development of the internet, we may as well give up now.

What *did* annoy me, after the decision was taken, was that my difficulties with ichat over the last few months seem to be similarly down to Comcast policies.

I use iChat a lot to keep in touch with my family (all of whom have Macs, and 4-way video-conferencing can be pretty cool). There's several thousand miles between us, so this is one of the few ways we can actually see each other without major travel.

Until a few months ago, it all worked great. Now, I get less than a minute of great picture, and then everything breaks up. I was putting it down to transatlantic bandwidth issues, but then I tried it from work, and (lo and behold) had no problems whatsoever.

I pay (not for long, now though, the T1 arrives in 2 weeks) for the most bandwidth Comcast offer, and I cannot believe I average even 1% of that bandwidth. To have them limit me when I *do* want to use it, as a deliberate *general* policy of theirs, is infuriating. All I can do is cancel the service, and hope others do too. Eventually, hopefully, they'll get the message. Not everyone can cancel due to the monopoly they hold in some areas, but perhaps enough can to make a difference.

Now a T1 used to be a lot of bandwidth, but it's not so much any more (1.5Mbit/sec is pretty poor by advertised-bandwidth standards). I'm willing to trade off the small time-periods I actually can use that advertised bandwidth for the reliability of always having the smaller amount - it may not work for everyone, but it works for me :)

And so, Comcast lose another ~$200/month. Hopefully part of a trend, because won't anyone think of the network ? [grin]

Simon.

They're doing it for more than just file-sharing. iChat, for example, is suffering. I can no longer video-conference my parents across the Atlantic. At first I just blamed it on bandwidth issues, but it only seems to be a problem when I initiate the connection from a comcast network (home). Others are seeing the same.

Lotus notes is similarly affected. It seems that if you transmit small bursts of packets, Comcast give you the bandwidth you've paid for. If you start streaming data, or the volume of data goes over some (low, ichat only takes ~30 secs) limit, they attack the connection, injecting fake RST packets to the data-stream at both ends. I've decided to bin comcast. Lower (but reliable) bandwidth is far preferable to me.

Simon

I recently wrote about this ...

Short version: Genuine Vista crapped out on me, screwed up a huge download (twice!) and initially refused to realise it was genuine. Only after installing an Active-X control (God, I hate those) did I manage to get it working (and it only offered that solution the second-time-around).

A sufficiently bad experience that I just deleted the windows VM and installed Ubuntu on a VM instead. So, yes, MS screwed me out of the $300 or so for the 'Windows Vista that is licensed for VMs", but it's the last thing I'll ever buy from them. Anyone want to buy a (used once) GENUINE copy of Vista ?

I don't pirate software. I don't see why I should be inconvenienced (at full price) because MS can't find their backside with either hand - if you're going to deny fake vista installations, then MAKE SURE THE DAMN SOFTWARE WORKS. PERIOD. NO IFs BUTs OR OTHER EXCUSES. [rant over].

Simon, disgusted with MS's attitude.

... and the theory behind what I was doing is up at my blog. Or at least most of it is (all of it modulo time constraints. It'll all get there eventually).

Back in the day (almost 2 decades ago), I was using video rather than still images (which allowed me temporal information as well as spatial information) but I recently wrote a simple application to just use the spatial information to find me images "most-like" a source one. The original goal was to train the system and then try to leverage a semantic processor from the trained system. It worked reasonably well (sometimes astoundingly well) on the database I had (some 300,000 images downloaded from keyed-searches on google images).

As Hitachi said, the key is to develop a matching system within a higher numerical dimension. One of the missing pages on the blog (I'll get there!) is how to evaluate the usefulness of any given feature (=dimension) of a region of an image. With this, one can approximate a numerical value for the information being relayed to the recognition-system using that feature, and therefore establish its worth as a feature.

When you know what you're looking for (your feature set) *and* the value of each of those features to your recognition system targets {man,boat,grass,house,...} you can create reasonably useful discriminators and rule-systems based on those discriminators. Note that the discriminators and the rule-system can be given to the system as a-priori information, but most of them are created and destroyed automatically *by* the recognition system as it evolves. It sounds complex, but really it's a bunch of simple ideas applied one after the other.

Simon.

Re "Robin Hood and Friar Tuck" - that was the first I'd heard of it, but I have a similar tale, though in my case it could be more accurately described as "Robin Hood and the Sheriff of Nottingham" :-)

Simon

So, I tried to reply to this, once it was being marked as 'insightful'... But the new /. text-formatter sucks dead bunnies through thin straws for preformatted text. So, here's a link to my blog-post instead.

Summary:
    - There's more to security than a firewall.
    - Linux is a fine server OS too, no disagreement there.
    - OSX has a better client than SWAT.
    - Disk performance on OSX is fine.
    - OSX doesn't need a "guru" to administer.

Simon.

These are some of the greatest questions ever asked - Are we alone ? Is there anyone/anything like us in the rest of the universe ? Would it be possible to communicate with an entirely alien species ?

Quite apart from the Wow! signal (so I guess they found something after all), there's a world of difference between the Seti@home distributed computer program, and the SETI institute - a collection of individuals who have SETI-capable telescopes . The SETI institute is not at all connected with SETI@home, and it is they who are 'seti', or at least they have the greatest claim, having been 'SETI' for years previously...

It's not actually hard to make a radio telescope - get a big dish, an LNA (low-noise amplifier for the signal), a microwave receiver, and a PC (windows or linux). Oh, and lots of space for that dish :-) Total cost is ~$2000 if you buy everything. Ebay is your friend regarding getting stuff cheap, though :-) It cost me significantly less than that... So, get searchin!

Simon.

The data I had for hostip.info was 1-pixel-per-kilometre. This new data is twice that resolution, and if I combine this new data with the soon-to-be-open-source 3d engine in Flash, I think it'd be really cool. Geolocate yourself or anyone else by their IP, then zoom around that location in 3d :-)

Sounds like a fun thing to put together - maybe this weekend for the 2D stuff, and as soon as the 3D engine is open-source, I'll include that :-)

As always with this sort of thing, it's getting hold of the data that's the hard part - kudos to those giving it away free to research, education, and me :-)

Simon

In the event of a massive cyberattack against the country that was perceived as originating from a foreign source, the United States would consider launching a counterattack or bombing the source of the cyberattack, Hall said. But he noted the preferred route would be warning the source to shut down the attack before a military response

The last java applet I wrote geolocated you on the globe (by your IP address), and did an 'enemy-of-the-state' style zoom-in to your city. In retrospect, I wish I'd written it in flash - it's just annoying to see a grey box for a few seconds while the applet VM initialises itself. At least in Flash, you can put a "loading..." animation up quickly.

If you want to know more about the geolocation thang - details at my blog [no adverts :-]

Simon.

I created the hostip.info website a couple of years ago which is an open-source geo-location system. I've since passed it to others to maintain.

On their front-page right now is a geographic map of website traffic, which is a "public-beta" of a mapping system (upload IP addresses, get a map).

They don't actually appear to give out the code (from only a quick look on the site, that may not be correct), but this ought to be pretty trivial to do yourself if you wanted... The public API of hostip.info returns the lat/long of any IP address it knows, and plotting lat/long on an image of the world ought not be too hard.

I designed the whole system so you could download and set up your own geolocation database, slaved from the master at hostip.info with regular updates. Then you can query IP addresses locally and generate charts however/whenever you wanted...

Just a thought,

Simon

Perhaps the scientists (you know, the people who know ALL ABOUT how to get the best use from a telescope, the same people who designed it!) might just have taken that into account ?

The main constituent of atmospheric aberration is turbulence within the atmosphere. The atmosphere over the Antarctic is the thinnest in the world, it has far less turbulence because it's damn cold (heat = energy = motion of the gas), not to mention any massive heat 'spires' from human pollution.

You can use adaptive optics to characterise and therefore minimise the effects of the atmosphere - you shine a laser upwards, scatter off sodium atoms ~90km up, and use the measurements as inputs to actuators on the mirror segments approx 1000x per second. This can significantly remove the aberration if done correctly (you can use 2 adaptive systems, one natural, one artificial with a laser)

In any event, this is all old news, and there are existing telescopes using the technology. There have been arguments made before for the use of ground-based devices rather than space-based ones...

And yes, I do have an interest in astronomy, but of the radio kind rather than the optical variety - I picked all the above up from news channels...

Simon

Optical (ie: laser detection) SETI has been up and running for a while now (see Optical SETI overview for example). Drake ought to declare an interest though, since he's one of the investigators on the project.

It's a reasonable argument, but it's far harder to set up optical "listening" posts than radio ones. It cost me about 1000 uk pounds (WHY is the pound symbol banned from /. ?) to set up a SETI listening post, including all the costs from dish/low-noise-amplifier through receiver and PC. Setting up an optical one is waaay more expensive. Optics in general are far more expensive than radio components, and large-scale ones are extortionate :-(

The counter argument of course is that to detect laser light, the remote civilisation have to be pointing their laser at us, whereas with radio it doesn't matter since it's not a directed beam. Against that you have to offset the time-period over which transmissions of either kind could be made...

The chances of getting a radio contact may be a few orders of magnitude lower than getting an optical contact, but since the chances of me setting up an optical SETI station are precisely 0, the chances of getting 'the' signal with radio is infinitely greater than with optics, at least for me :-)

Simon

Do you know how finely tuned the receivers are ? I do.

Typically your receiver has a (physical, not in-software) narrow-bandpass low-noise amplifier that is tuned to a minute section of the frequency spectrum (say 1MHz range), specifically because you're trying to find a needle in a haystack - a galactic haystack at that! It's a noisy place out there; to design for the specific case you need requires a lot of optimisation of that case versus everything else - it's not just like scanning up and down the range... if you are not spot-on the correct frequency at the correct time, you'll miss the signal.

Simon.

I'd love to give it a go with my very own personal radio telescope (dish.jpg). Sure it'd be hard to point, and maybe not possible to receive anything at all, but I'd like to try :-)

Simon.

I thought about doing this for my own radio telescope. One day I might just get around to it :-)

The interesting thing about using an FPGA would be the speed of the FFT. With FFTW (the Fastest Fourier Transform in the West), I get approx 230 us (microseconds) per fft. Using an el-cheapo FFT, best efforts would be on the order of 15us, or approximately 15x faster... Bung several on a few PCBs, and you're talking super-computer speeds :-)

Say you use 16 FPGA implementations, that'd be the equivalent of 240 Athlon 1800XP's... With those sorts of speeds, you could do realtime chirp analysis for doppler effects on an incoming signal. That *would be cool* :-)

Simon.

In a word, yes. I have a WinRadio 1550e, which allows monitoring within the waterhole (~1.4GHz) which is where most amateur seti astronomers look.

I have a dish, which had to get signoff from the secretary of state before I could install it :-) The picture shows the width of the house, with the dish being approx 4m across...

Making an interferometer poses major problems with time resolution though - to merge all these amateur radio telescopes together would (a) take a huge chunk of bandwidth for each telescope (ADSL ain't enough...), (b) need excellent synchronisation between the telescopes, which almost all of us don't have, and (c) need the dishes to be steerable, which most of them aren't...

There is however a project argus doing the same thing with lots of individual telescopes. As soon as I'm happy with the s/w running on mine, I'll be a member of the group :-)

And no, no aliens yet :-)