How not to build a 32-bit CPU

Ah, the 386. I remember when the first Compaq 386 machines came out (I believe at 20MHz, not 33MHz, though Wikipedia tells me that slower 386s were available), and my fanboyism disliking the fact that it took the performance crown for desktops back from the 8MHz ARM2 Archimedes machines (probably as measured by dhrystone, although BASIC may have been involved). I'm sure my copy of Structured Computer Organization contains some comment about Intel "finally making a decent CPU", but unfortunately I'm a few thousand miles away and can't check - anyone got the red edition?

The 486 was a bit nicer as designs go, excluding the slight problem of getting everyone to optimize code in a way that was pessimal for Pentiums. I still wish IBM hadn't decided to use the chip from their printers and had gone with the 68000 series from the start (and if we were going to end up with thirty years of compatible machines foisted on us by Windows, we'd ended up with anything less crufty than x86), but at least it meant that near and far pointers weren't always obligatory...

I'm confused...

How did we get from Samsung and LG being fined 1.47 billion Euros to them having a total of about 345 million Euros, of which Samsung's was only 151 million (and smaller than Philips - who are also fined more than LG - and Panasonic)? Even collecting the four mentioned companies together, there's half a billion Euros unaccounted for. Or did I hopelessly mis-read it? Not that Samsung are unused to being fined a billion, but it'd be nice to know what's going on. (Disclaimer: I'm employed by Samsung, this may hit my salary...)

Ope do

Ah, that takes me back. Not that I ever completed it as a kid. I really must go back to it. I still tend to think "ope do" when opening doors, which is a bit worrying now I come to think about it.

After this, the Lord of the Rings game was a big disappointment, mostly because it barely worked (maybe I had an iffy tape). Although that's how I got my copy of Fellowship of the Ring, so it turned out all right in the end.

People who don't read up...

It's very publicly been known as having a soft G since its instigation, and the request of the owners. You may have been pronouncing it with a hard G for all that time, and it's more recently become explicitly accepted that this is okay, but historically the hard G was plain wrong, if common. These days I don't bite anyone's head off over it (though I still twitch whenever my colleagues say it with a hard G), but I'm not going to take people claiming that a soft G is wrong.

For the record, Linux historically came with an indication that Linus preferred it either to be pronounced as he did (sort of Leenuss, as I recall), with a secondary preference that one should pronounce it as one would pronounce his name (in my case, natively, Lie-nuss). He didn't like people attempting to pronounce it like his name and getting it wrong, as in "linnux". I believe he has since changed his mind on this, not least because "Linnucks" is so common. I still say "Lie-nucks".

Oh, and Risk-Oh-Ess, for what it's worth.

Hawtness

"All this hardware oddness begs one question, of course: why would low-power ARM-based chips such as Apple's A series, fabricated in increasingly smaller and therefore increasingly less power-hungry manufacturing processes, need to resort to such a complex, space-wasting scheme as fan-based cooling?"

Well, my elderly HTC sensation gets nice and toasty even playing Angry Birds. Run something with actual 3D requirements and there's no way to avoid heat being generated. ARMs and embedded graphics cores are very efficient, but we're still talking multiple >1GHz processor cores and a lot of graphics, usually in a fingernail-sized chip, and they're not magic. Unless Apple are getting their performance by fabbing with superconductors and Josephson Junctions, of course.

Digital babelfish, how I miss you

Some remembering the earlier days of the internet may recall one of the first popular public digital translation services, babelfish.altavista.com, since moved to babelfish.yahoo.com, and now redirecting to Bing translate. Altavista, of course, started out as altavista.digital.com, set up by Digital/DEC, and some of us remember when it was the search engine of choice.

So, no - digital babel fish is not closer. Microsoft have moved Babelfish several step away from Digital.

That said, I'm often astonished that Google Translate manages to produce something that's not a coherent sentence, let alone incorrect. I always assumed that something in the implementation of these things must understand some rules of grammar which ought to make that kind of problem tricky.

Unprecedented?

In this sense?

http://www.theverge.com/2012/6/1/3056490/sharp-caac-igzo-498-ppi-display-prototype

No magic

"Instead of capturing the output of each individual pixel separately – as sensors normally do – the trick is to combine the output of groups of individual pixels into a larger pixel."

Which is in fact exactly what happens on most cameras if you save a lower resolution image than the default provided by the sensor. And, indeed, it normally improves the noise handling. I've no idea why everyone is making a big deal about this, other than that Nokia can justifiably claim that the sensor resolution is not detrimental. (Now, arguing about whether Nikon should implement a "small raw" mode on their high-end DSLRs is another matter.)

"This technology allows the 808’s sensor to capture as much light information as much larger pixels and sensors would."

The fact that the 808 has, for a phone (and most compacts), a chuffing enormous sensor and a fast lens means that it can capture a lot of light information. The amount hitting the sensor is the amount hitting the sensor. The amount per pixel is small, but DxO do their noise tests normalized by total image area, and exactly the same concept applies. So "[fewer] larger pixels", yes - "larger sensors", no. So much FUD about downsampling...

Anyway, nice bit of kit. I might pick one up as a compact camera once their price drops to clearance levels. Not so tempting as a phone, though.

And still down-playing themselves?

So, assuming this is actually done with image processing (and yes, it's normal to capture a larger frame and calculate a shift within it), why have they decided to go with "optical" image stabilization? Optical solutions fix camera shake, but do nothing to handle subject movement, like the bouncing Nordic woman, which is why professional sports photographers still have large aperture lenses. In extreme cases, as here, keeping the subject static in the frame would result in the background bouncing around, due to the change in perspective from the moving camera position (something Canon have tried to fix in a stabilized macro lens, but not for riding a bike).

There *is* a lot of research into stabilizing/removing blur independently from separate bits of the image - some was presented at SIGGRAPH this year, and Adobe explicitly stated that their work on this was the reason that they'd not yet released their camera-shake-removal technology (demoed recently) in Photoshop. But "optical" it's not.

If it's image processed, you may as well own up to it. Of course, if there really a stabilization element in there, I take it all back - but it's quite possibly not the best solution.

These things always look good in demos. I'll reserve judgement until a real world test, although I don't think my DSLR is going anywhere.

Film/film ambiguity

So they're keeping their "commercial film" (is this motion film?) but getting rid of the consumer stuff? (I assume they don't mean commercial vs consumer photographic film.) And they're keeping their ink jets; really?

I hope someone picks up the slack. Otherwise, I'll be off to stock up on Portra - to go with my stocking up on Velvia after Fuji discontinued that. They're not making it easy to buy into a 5x4 camera system...

And the rest?

Interesting review. To me, a lot of the sample images are focussed in odd places (not on the eyes, at least); I don't know whether that's the AF system or the photographer. I've mostly heard bad things about the video on the D4, apparently because the downsampling algorithm introduces unnecessary softness (the 1:1 crop mode is okay, however, but why use a D4 for that?) The D800, which probably just throws lines away, is apparently much sharper. Interesting that there's no mention of the new AF switch position, which is already annoying me on my D800E. I'm not sure that mapping the AE-lock to pressing the joystick is an improvement, either. I'm sure grid lines were available in the finder on the D3s - I've always used them on my D700 (and they help me keep my horizons straight).

This is obviously a camera with a purpose: it's a work horse for high speed, low-light shooting. Journalists and sports shooters will love it. It's not a consumer camera, or a megapixel monster, so comparing it to a 'blad (or a D800) is pointless, as is talking about the price - the right image can sell enough newspapers to justify it, and the price is still lower than many pro lenses (and almost exactly the same as the 200-400 f/4 shown in the review). Consumers shouldn't feel they're missing out if they can't afford one and have to settle for a D3200 instead; each tool to its own place. So no more "how much?" comments, please.

And film? It does have its place, but its place isn't in low-light, high frame-rate, fast turn-around shooting. I have, and use (mostly for flowers that look better on Velvia than digital), an F5, but I'd be mad to try to take on a D4 (or D3) with it for the tasks it was originally designed for. Pros do still use film - but more often in a cheap camera for portability and travel, in a Leica for subtlety (those who haven't spent D4-money on an M9), in medium format for quality or in large format for the ability to fill a wall with a sharp image. Even my Pentax 645 isn't going to get much use now I have a D800, though a Mamiya 7 or a 5x4 (on my shopping list) would be another matter; for what they're good at, they'll smoke a D4 - but then so will a D800. Photojournalist use of film died with the D1 (and especially the 1D), which is why the F6 is so different from the F5 and aimed at prosumers, and nobody's updated a film camera with a modern autofocus system. Shooting through rose-tinted spectacles results in poor images. Someone was using large format at the Olympics to get some interesting images, but I'd be astonished if any pros were using 135 film, at least in an SLR.

Dragging to external devices

For what it's worth, I don't really know *what* to expect when it comes to dragging files around. I established in the 90s that Windows had different behaviour depending on the source and destination location, decided that this was one of Microsoft's usability nightmares (I have USB drives that don't behave like removable devices and systems that come up with their drives in a different order; worrying about the default behaviour is the last thing I want), and I've been right-mouse dragging files (and selecting copy or move from the menu) ever since. Oh, and occasionally I use Ctrl-X/Ctrl-C and Ctrl-V. But then I was brought up on Acorn systems, which let you choose whether you wanted a copy or move by which mouse button you were doing it with, so trying to second-guess based on the device type was never a problem.

Not that this makes the Chromebox any more usable from the sounds of it, but "not like Windows" doesn't mean "broken". That statement is really quite often true.

I'm vaguely tempted

My latest version of a DataLink has got to the stage where I'm probably never going to program it again, the strap is mangled and the battery is dying. I was on for a Pebble (eInk appeals, although if the previous product was still made I might have been tempted by that too), but they're not out yet, and I'm struggling to find other smart watches that are actually being made - at least, any smarter than a DataLink. I'm less interested in syncing with the phone than running my own apps.

I'm not quite clear with this whether users can write their own stuff for it (need to do more reading). Knowing the resolution of the screen would be nice, too.

Good old British weather

I was up at 4:15 (after watching first contact on a web cam before going to bed - good work Hawaii, not so impressed by the guys from the continental US who were filming each other and not the actual sun), ready to go out and meet the sunrise. But I saw the weather and stayed in bed.

Then I got up around 4:45 just in case, and watched some more live streaming from Oz and from Hawaii. And it stayed cloudy, but with the occasional thinner bit.

About 5:45, after Hawaii had reported third contact and while Norway were showing parallax, I realised the sun might actually hit the house. So I ran upstairs, ignoring all the exciting telescope stuff I had with me, and pointed a (stopped down) lens right at the sun. Lo, the sun had a tiny bite out of it (I was slightly nearer third contact than fourth), and I have photographic evidence (and hopefully not a knackered camera) proving I was looking. It turns out that clouds work as an extremely dangerous alternative to a solar filter. Of course, if it had been sunny, I would have had more time to project an image onto paper.

I've heard the "refractors good. reflectors bad" argument before, but I'm very confused as to why. I'd have thought (enclosed) refractors are more likely to overheat than an open reflector (incidentally, *all* Dobsonians are reflectors, by definition). The only reasoning I can think of is that reflectors tend to be faster (shorter focal length per aperture), but that's not strictly a reflector vs refractor difference. Anyone care to educate me?

Lander in BASIC?

I'm not sure why people are saying that Lander was written in BASIC. To the best of my knowledge, it was pure ARM code, and I'd be astonished if it went anywhere near Acorn's triangle drawing routines. BBC BASIC on an Archimedes is fast (and the Arthur, although not RISC OS, desktop was written in it), but it's not *that* fast. Minotaur, one of the first commercial games launched for the Archimedes (alongside Zarch), *was* written in BASIC, however.

(Speaking of BASIC, I'm not sure about this "press reset twice and you get the program back" thing. Not in my memory you didn't. I also remember the reset switch being on the back, right next to the headphone socket, where it was easy to reset the device when plugging headphones in. It was still on the back on the RiscPC, even though the power switch was on the front. Never understood that...)

As for comparisons to Virus, I believe Virus had some more enemies - it got harder faster than Zarch - but it was also noticably less pretty; for example, there was no depth cueueing of the background (in Zarch, everything got darker towards the back of the screen). I'd assume that the Amiga version used the blitter, although since there was an ST port I can't guarantee that. I've never played the Spectrum version, but it's high on my "most preposterous port" list. Lander, of course, didn't have all the enemies, let you blow up on the launch pad, and didn't clip the front edge of the landscape properly - but it was very good at training people to play Zarch! (I still maintain that I ought to be able to fly a helicopter, should I ever need to, because of this game.)

Part of the exciting colour scheme of Zarch was that it could use the 256 colour mode, back when the best PCs had original VGA graphics. The 256 colour mode had 16 palette entries (that most people didn't touch) and the rest of the values derived from them; the default mapping was an effective perceptual HSV scheme, accessed from BASIC by 64 base colours (setting the top two bits of each channel) and four levels of "tint" (setting the bottom to bits of all three channels at once), giving you fine grained luma control. It might not have had all the colours of HAM6, but it was pretty effective. Despite a brief foray into VU-3D on the Spectrum, it was probably Render Bender that taught me to think in 3D graphics (and now I work in graphics professionally).

Oh dear, spoilers

I'm only half way through the second book, but I second the concerns that things seem to be slowing down a bit - sad if it gets worse in later books. One of the things that hit me about the first book was that quite a lot happens in it; I'm less surprised that the TV adaptation worked than that they managed to fit most of it in. On the other hand, at least some of the characters come across without the "comedy dwarf" tweaks that Jackson felt the need to make to LoTR.

The thing that really put me off the first series was that some of the early acting - even by relatively experienced actors - seemed incredibly stilted. Some of it was very good, but Littlefinger's need to orate everything really grated; even Lena Headey (who's been good in plenty of other things) didn't seem very comfortable. It's possible it got better as the series progressed, or maybe I just got used to it.

I'm also not all that impressed with the need to age the entire cast (because it stops some of them acting their age and the shock of what they go through is lost a bit), but I guess they'd have had problems broadcasting it if there was really the requisite amount of child nudity.

Here's hoping that the pay cheque persuades the author to find a way to finish the story.

Optical track pad

"With the same rounded corners, chrome vanity band and optical track pad you would be hard pressed to tell the two apart."

What optical track pad?

Otherwise... meh, SD resolution, not for me - but that doesn't mean it doesn't have a market.

Actually, a little harsh

I'm grateful that Kodak kept making film so long; I've still got some in my fridge, although I admit that my film shooting these days is more usually with Fuji or Ilford products; I may regret their troubles more when I eventually get a 5x4.

There's no doubt that the film market shrank radically the moment cameras went digital; despite my first paragraph, I do most of my shooting with a DSLR, and that's not going to change (except, with small probability any time soon, if mirror boxes go the way of the dodo). I actually started with DSLRs, unless a Polaroid camera I had in primary school counts, and added film to my repertoire.

I can't imagine Kodak not seeing it coming, but the question is what they could do about it. According to dpreview, Kodak have made 144 digital camera models, some of which are still current. For a company which had more to do with chemicals than optics, electronics or consumer goods (at least in recent years), that's not a bad effort - but it's not surprising that Canon is the company most visible in the desirable compact and DSLR space and that Sony's electronics combined with Nikon and Pentax's camera design is taking much of the rest of the spotlight.

With the premium products made by big names, Kodak - whose cameras have never exactly been the M3 or F-series of their generation - could only really try to compete at the cheap end, and I suspect they didn't have the manufacturing capacity to create small plastic boxes as cheaply as the bigger companies in China. Even if they did, that market must be feeling the squeeze now everyone owns camera phones. Other than "something else", I don't know what they should have done.

Still, maybe I should get my hands on the DCS-14n that I want as a back-up to my D700 before the collectors start putting the prices through the roof.

Oh, and to add to the history lesson in digital photography, Bryce Bayer (of the pattern) worked at Kodak. I wonder which name will live on longer?

Call me a pedant, but...

"along came Quake, the first OpenGL application"

Er. No. I couldn't actually tell you what the first OpenGL application was (probably a demo included in someone's implementation of the OpenGL1.0 specification, which dates to 1992), but I'm damn sure that something was written before Quake (1995), if only because there were a load of SGI IRIS GL applications to port.

It's true that Quake was the first application that popularized OpenGL support in consumer devices, though.

Z88 - so near...

I used a Z88 when they were new, and bought myself one a few years back with a view to having something I could use for typing and notes. Long battery life, famously quiet keyboard to type on, lovely bit of kit. I used it about twice, because it was huge in comparison with my Libretto (whose hinge never, thankfully, broke) and therefore never got taken anywhere to justify the inconvenience of transferring content.

One of the tablet/keyboard combos might be a better modern equivalent, although swype is actually almost decent enough on a mobile that you really need to be writing something big for it to be worth something with a full keyboard. I've yet to try the bluetooth keyboard route (they were horrible under Windows Mobile), but I'm certainly going to.

It doesn't solve the "I want a battery that'll last me going on a writers' retreat in the woods" problem, but you could get a lot of spare battery (or solar charger) alongside a mobile without getting to the bulk of a z88.

There was a time when I was involved in a long chat about the "Psion FX" - something in the Psion 5 form factor, but with an Atom in it. I think this got killed by the (x86) netbook fad, which the discussion predated, although a netbook is really not the same thing.

Cheap enough and small enough (but how do you make a keyboard and screen small without having a hinge and ending up with a Vaio P?) and I'd still find the idea tempting today.

Argh.

"A lot of people have talked about it being bad for innovation but I think the ability to enforce patents is a core aspect of what drives people to invest the huge sums of money that they do in R&D,"

Something that's making me seriously consider getting out of the software industry (preferably in the direction of academia) is how much nonsense this is. Patents were invented so that the few people who had enough spare time on their hands to solve problems would share those solutions, furthering society. We now have an industry of getting on for millions of people who are employed solely to solve a limited set of problems. The problem will have needed solving by someone else long before a patent is granted, so the disclosure advantage is no longer outweighed by the exclusivity.

Patents were meant to allow the better mouse trap to be available to more people. Now everyone is making mouse traps, one person invents the best one, and everyone else spends their time realising that they couldn't do it the right (and often obvious) way and have to invent an inferior workaround instead. This is immensely frustrating if you want to put your heart and soul into making the best possible mouse trap, and if you're building a machine out of thousands of mouse traps, it's immensely frustrating to have to worry about this at every turn - patents rarely cover something that took a long time to invent, and software engineers spend their time building large projects out of lots of these small problems.

Has anyone worked out how much bandwidth and storage space would have been saved if IBM hadn't had the arithmetic coding patent that meant everyone's JPEG implementation uses Huffman? (Admittedly, only a little per image, but there's a big multiplier there.)

There was an interesting article on Ars Technica a month or two back discussing a report that had noted the effect of patent disputes on company market values. It put the cost of patents to the software industry over the last twenty years as $500bn. I'm not surprised.

Hello new phone... oh wait.

I've been waiting for a 720p phone for a very long time (partly for text quality, partly I was spoiled by a 3" WVGA screen in the past, partly because video is less mangled by it). I've no problem with the size - my Sensation isn't that much smaller, and easily fits in a pocket/my hands. PenTile is a big disappointment, though - it may put me off.