It cannot be to Google's detriment, in court, that they obeyed the law of China when they did business there.
Try telling the US government that as it's legal to enrich Uranium in Iran, you weren't committing a crime when you sold them those centrifuges.
Anyway, Your argument falls quite simply: Google have the choice to not do business in China.
I look at this and wonder is google will use the common carrier clause
Unfortunately, in their willingness to censor when told to by any repressive government they want to do business with (China and India notably), they've proven they aren't a common carrier. Google can't be the Chinese govenment's bitch and also claim to be a common carrier.
For a billion dollars, it's much cheaper to just buy Viacom....
``Much cheaper'' in the sense of ``twenty five times more expensive''. Viacom's market cap is about 25 billion. To carry out a hostile take-over, which is what you're suggesting, would cost thirty five million or more.
Have you seen any ISO 9001 certificates?
The idea of going ISO is to be able to certify and advertise you compliance.
There is no 97% compliance certificate!
The management systems, starting with 9000 and spanning things like 14001 environment, 18001 safety/health and 27001 security, have audit as an integral part of the process. That's not true of the other ISO standards: there isn't a process for having your `ISO' C Compiler certified, and there isn't an audit process. There are test suites, but no certifications.
You can't fly planes like that manually, becuase they are inherently unstable. Even non-stealth aircraft have this property, in order to make them more sensitive in roll. A civilian plane will self-centre from small roll inputs, and you have to overcome that effect to actually roll. The stealth aircraft are such weird shapes, for which aerodynamics come second to radar cross-section, that the designers don't even have the choice.
The best bogus experiement I ever heard of was a defence of Glenn's shuttle flight: research on the effect on older people of 90 minute light/dark cycles. Leaving aside the actual applications of this research, which is cheaper: a shuttle flight, or a Motel Six room with a 90 minute switch on the lights?
Most space `science' falls into Feynman's comment that he'd heard they did science, but it never appeared in journals he read. It's mostly work to provide the systems that permit them to do more work to provide the systems, etc. All the talk of exotic zero-g materials is spurious: if you could lift water into orbit and transform it into gold using only your bare hands and the power of zero gravity, it would still cost more than gold is worth. It's just possible that materials whose cost per gram is more than gold (some exotic drugs, perhaps) might be economic, but there's never been the slightest explanation as to why zero-g would help make them.
Low voltage is more dangerous than high voltage, for a given power consumption. If the concern is heating in the duct, as it often is, 10KW@48V is a completely different animal to 10KW@415V three-phase. I'm always amazed at how flimsy cables are in 110VAV territories, compared to standard practice in 240VAC areas, and I can only think that back in the day, someone convinced themselves (wrongly) that you need cross-sectional area in the conductor for the volts, not the amps. Which in turn is why Americans complain about how weedy electric cookers are, and don't generally boil water with electric kettles: I can have 7.5KW into an electric cooker (30A, 240V) or 3KW into a kettle (13A, 240V), while Americans can't get 70A into a cooker.
Heating effect in a given cable rises by the inverse square of the voltage for constant power, which is why power transmission is done at 25KV or more. I don't think there's much OLE on railways in the USA, but it would be interesting to see what voltages they selected for that.
Chris Keates isn't quoted directly on `legislative control', you'll notice, so w don't know what she said. Having an indirect knowledge of both the union and the person (my mother was a regional vice-president of the NAS/UWT, and both my parents know Chris Keates, who lives locally, pretty well) it seems highly unlikely that they're calling for what people are complaining about. My father, tech-savvy and politically/.-compatible on most YRO topics, has a high regard for her, so I doubt she's some swivel-eyed fascist.
A search of the NAS/UWT website for the keyword `Internet' reveals no policy that could remotely constitute a call for censorship, and without a conference resolution a union general secretary has absolutely no authority. There's the usual guff about cyber-bullying, in a document that any union-watcher will recognise as the tortured prose of a composite (pronounced, bizarrely, to rhyme with website) motion, but even that is just a call for web filtering in schools.
Unfortunately,/. runs on its own timescale and fact checking takes too long, but I'll try to find out from Chris Keates what she actually said. The absence of direct quotation makes me suspect it wasn't what people think.
Anyway, even if she called for the Internet to be closed down tomorrow, she's the general secretary of a minor NUT-affiliated trade union. That's not exactly a position of major political power, is it? All sorts of people believe all sorts of things: being able to execute them is rather a different proposition.
The BBC transmitted ideoforms (which were nothing like a secure as is often claimed, but that's another story) to the French resistance via transmitters aimed at France, yes, but in 1943 it was no secret that the British were keen to help the French resistance, and it was no secret that there was a war on. Today, the list of countries in which the CIA have agents is itself interesting intelligence.
Perhaps British understatement isn't the flavour today. ``I'm not sure I approve'' just means I haven't thought about it terribly hard. To get from there to a rant about the desire to snatch copyrights from the hands of starving children seems a bit of a reach. I agree with evey one of your points. I'm not sure why you felt the need to assume I didn't.
My uncertainty comes from the following scenario:
I release my product under the GPL.
Various people beta test, patch, enhance and generally improve my product, believing it to be GPL'd.
I then dual-license as a commerical offering to those that want it.
Now yes, you could argue that the people submitting patches knew the score, or that if they had reservations about possible license changes they should have made appropriate legal moves. And certainly this pattern of events is neither illegal nor (for most people) immoral. But the point is that the value I would be selling is partly due to the improvements that other people donated, and I'm not sure I approve. ``I'm not sure'', note, so try to avoid the hyperbole of stacked assumptions starting from my being sure. The issue isn't the main author's work, it's the intangible value created by the work of others.
Mind you ``Revenue? GPL enthusiasts can't even spell it'' is a tempting cheap shot.
Hans Reiser says that he actually makes some money selling the right to use his file system without telling anyone else that they're using it.
(Yes I know, but the corporate world is weird.)
I've got one of the machines that is covertly Reiser4 under the hood, I believe. I can see why a vendor would want to keep it quiet, too.
Firstly, the admission that you don't own the filesystem (in the sense of employing all the major contributors) is a worry for many customers.
Secondly, if you want to put your own secret sauce into the filesystem (perhaps hooking it more intimately into your product's volume management, or providing a shortcut API into your block level IO, or doing extra things for fast failover between control units, or whatever) then you don't want to have to pass this stuff out GPL'd.
And finally, if you want to use an otherwise-GPL'd filesystem linked into a non-GPL real-time executive like VxWorks (no relation to VxFS, confusingly) or QNX, having a non-GPL version of the filesystem probably saves everyone a lot of lawyers bills.
I'm not sure I approve of this as a GPL enthusiast --- hey, I had code on the Emacs 17.61 tape! --- but as a customer I don't think I care too much. You don't get to have much oversight of the components used in products you buy unless you're entering into the wild world of source escrow, and buying a non-GPL'd version of a GPL'd product is no different to the OEM buying something completely closed, and in many ways better (I still get the many-eyes thing, up to a point).
In a couple of years IPv6 will be main-stream because of MS, and we all know the benefits from using the upgraded protocol.
And for those of us who don't know the benefits, even though we've been told every year for the past five that ``next year will be IPv6!'', what are they?
There are a couple of shops, in Germany and I'm sure in other places as well, which happily sell prepaid phones without proof of your identity.
In the UK there's no requirement to even pretend to take ID. You can buy a phone for cash, a SIM for cash, top-up minutes for cash, etc, etc.
It's a classic case of `unintended consequences': the assumption circa 1995 was that the big criminal issue was theft of service. At the time the only way to get a mobile was on a contract: you needed a bank account that would take direct debits, and some proof that you weren't totally sketchy with regard to credit. The only way the operators were going to build out their userbase in that environment is to redefine sketchy, and even then it's not going to get anyone under 16. So in the new world, sketchy would be redefined upwards (ie anyone the slightest bit dubious is refused credit), but pre-pay is universally available. Shazam: instantly you have a massive increase in mobile phone userbase. Fraud drops, because (a) GSM cloning isn't as easy as ETACS cloning (b) your contract customers are better risks but (c) most importantly the prices are falling and phone fraud isn't really worth the candle.
However, for crims, losing the ability to sell phone time to Pakistan for half the standard rate is a small bump. Suddenly having access to completely anonymous, mobile, non-suspicious (as compared to hooky PMR equipment) point-to-point communication is like heaven on wheels. But once the genie is out of the bottle, how do you deal with the problem when every eleven year old has a phone (my ten year old daughter claims to be the only child in her class without one: I think she's exagerating, but not by much)? That forces you to permit cash top-up from people with no ID.
They don't `always' release to business first. This case is because had Vista not shipped to holders of certain Software Assurance contracts written around the time XP MS would have found themselves on the receiving end of some nasty lawsuits from people with deep pockets. A lot of SA agreements were sold expiring Dec 31 2006 promising to cover whatever Longhorn was called that week: that was part of the appeal of the SA. Microsoft got the revenue and the lockin, now people want their software. By shipping Vista in 2006, no matter how non-functional (_no-one_ with a business large enough to have SA is going to be deploying Vista before Christmas), Vista slips into the SA.
The computer industry needs to face up to the fact that computers are now `good enough'. For most current desktop purposes --- email, word processing of small documents, web browsing, running corporate applications (usually client/server) and so on, a 2006-spec PC will do the job. There's not been a compelling feature in desktop Windows since NT 5 --- witness the reluctance for Windows 2000 shops to move to XP --- nor in Office since 2000. Except for providing toys for your younger employees to play with (a dubious benefit), why would any shop with >1GHz machines running NT>=5 and office >=2000 want to upgrade? How would you show the cost/benefit?
My wallet is actually a US one, book at Brookstone. As I like to think of myself of an international man of mystery, liable to be summoned to save the world on any continent, it doesn't just contain UK currency (tens and twenties: unless you're a criminal or a cashier you probably don't see fifties, and I doubt I've seen one in ten years). There's $10, 20 Euro and 5000 Yen in there as well. All the notes fit. I've seen US wallets which don't take UK notes, but here's a clue: the depth of the wallet might possibly be designed to match the depth of the notes.
To me, the conclusive proof that American notes (except for the $20, I think) are indistiguishable is that US cashiers always, always say ``from ten'' or whatever as you hand the money over. That's unheard of elsewhere.
Unlike with game consoles, where you have the option to tie spiffy new games to spiffy new consoles, it's not at all obvious there _is_ a market for `next generation' MP3 players. Cost? They are cheap enough that the only people who can't afford them can't afford anything, and there are weird off-brand USB stick with a headphone port for that market anyway. Capacity? A Nano, or even a shuffle, is enough for people who charge their player with new tunes every week. A 40G or more player is enough for all the the most obsessive to carry a significiant percentage of a lifetime's music purchasing with them. Size? Maybe a nano-sized 40G player might have a market, but it's niche. Video? Already available.
What compelling feature could be offered in a year's time to make people change player? Especially as, as each day goes buy, every tune bought for ITMS is locking people deeper in to the Apple platform?
I can think of things that niches might want --- Bluetooth headphones, say --- but the idea that any vendor can come to market and displace Apple from the MP3 market is a reach. The barriers to entry --- specifically, the synonymous use of `iPod' with `MP3 player', the fact that Apple have consistent marketing, branding and ITMS from $70 to $400 and the massive `dock' ecosystem for cars and other accessories --- are hard to overcome.
And, as several people have pointed out, the channels for products like this are channels that microsoft don't own, into markets where their branding isn't strong. Products have to earn their shelf-space, and Microsoft can't leverage Walmart. And I suspect that the key branding in the mass market isn't Apple, it's iPod.
It's incidental to this discussion, but in actual fact gcc is a particularly bad case to pick for HURD-like FSF behaviour. gcc sprang pretty well fully-formed from rms' head. Techniques were absorbed from other compilers, sure, but those of you not around in the mid-eighties can't imagine how seismic the arrival of gcc was. Within a few versions it was fantastically better --- in correctness, in compile- and run-time performance, in diagnostics, in every way --- than the prevailing (mostly pcc-derived) compilers. I recall using it to compile cTeX using something like 1.0.7 and being simply blown away by the performance. It was fantastic on 68k and on VAX, which were the only architectures that mattered at the time, and it did a perfectly decent job on NS32K.
It struggled with x86 (because of segmentation) and took a long time to catch up on SPARC (because of things like deferred branch and register windows, which RTL and the instruction scheduler struggled with) but by the time those architectures mattered the whole industry of somewhat hacky single-target C compilers was dead. Then Tiemann came along with g++ which, although it had some other issues about standardisation (which was nascent at the time for C++ anyway), blew cfront into the weeds within minutes.
There are FSF-sponsored projects which have been bloated, late and technically dubious. gcc isn't one of them. To my mind, the GPL is a fantastic intellectual and moral achievement, and probably rms' magnum opus. However, if he still cares about being seen as a memorable programmer, gcc is the place his repuation should be measured. emacs, yes, it's impressive, but the original TECO one is in many ways more groundbreaking than GNU emacs. For the latter, rms had Gosling's implementation --- whose redisplay was used up until about v16 --- and perhaps more importantly Bernie Greenberg's, later maintained by Barry Margolin, Multics Emacs, which was written in MacLisp. Not in C with an embedded Lisp-lite: Greenberg wrote in it MacLisp and made it usable on mainframes (I used it on Multics for about five years: an amazing piece of work). But there was nothing of the scale and scalability of gcc. It's a great piece of work.
They're also shit. Why don't you just buy a Dension Icelink? That way you can select playlists with the CD changer buttons, have the display work, have RDS TA work properly to interrupt (and pause!) your tunes, have the steering wheel controls work properly and generally behave like a proper music system.
Also, for a lot of people, cassette-shell adapters are a better choice. They don't pause the tunes when you get a traffic announcement, but they do at least allow the TA to come through.
There are no plans to move FM radio off 88-108MHz. I think the penny dropped that (a) DAB car radios are almost unheard of, even as factory-fit on new purchases and (b) unlike Televisions, for which set-top-boxes are perectly plausible, for radios there are just so many of them, most of them portables, that it's politically unacceptable. And ~3m isn't as useful a band as UHF TV anyway.
The government is forcing the UHF TV shutdown because it thinks it can then auction the spectrum. The theory is:
Spend millions promoting shutdown of analogue TV
Spend millions providing free STBs to the poor, photogenic and otherwise tabloid-friendly.
Lose a few marginal seats as locals find that their TVs don't work.
Put spectrum up for auction, saying ``it'll be as valuable as 3G licenses!''
???
Profit! To spend on a war that no-one wants.
You know how politicians live in the past? Brown's lot still think that Wired is pretty damn edgy, and are getting their financial advice on technology from Kevin Kelly.
I seem to recall that one key difference was that Microsoft gave away (near enough) development kits for Windows, reckoning that lots of applications was good for them. Whereas IBM sold development kits as a profit centre, figuring that developers should pay for the privilege.
Certainly today, MSDN can't possibly cover its costs directly. But Microsoft's great asset is all the ISVs who develop to their API. I didn't follow Apple `back in the day', but I get the impression that their current supportive attitude to developers, and the fact that they ship compilers and debuggers with every machine they sell, is a new posture. If that's true, and Microsoft stood alone in supporting a software ecosystem around their product then, and the very words choke me as I type them, that might be quite a major reason for their success.
He himself came out with "My enemies pronounce it 'Croulley' hoping to treat me foully, my friends pronounce it 'Crow-ley' to show that I am holy'".
Yeah, now you mention it I recall that quote, which sounds pretty convincing. But I'd be fascinated to hear a recording of Crowley's voice --- they exist, I believe. Leamington Spa, Malvern School, Cambridge at the turn of the century would be an interesting stew of influences.
The German beta-symbol can be, and is, replaced with ss. The umlauts can be, and are, replaced with a following e. Canada and France definitely, Germany I think also have the property that accents aren't used in runs of capitals, signwriting or other non-continuous text. So in fact, of the G7 countries, only Japan has a significant problem.
Amusingly, when the issue is non-UTF8 character sets, or censorship, or anything else that upsets the non-Western countries, they start shouting threats like ``Turkey will start its own top-level domains'' or ``Iran will disconnect from the Internet''. Which I'm sure is terribly impressive in UN-type meetings where we're supposed to pretend that all countries' opinions matter, but in the real world is an entirely hollow threat.
Were some random non-UTF8 country to make interworking with the rest of the Internet harder, it would be cutting its nose off to spite its face. For the G7 countries (yes, G7, not G8), the value of Internet connectivity to random minor countries is minimal. The value to those countries of Internet connectivity is large. Do US users care if Uzbekistan is on the Internet? No: it has zero impact on 99% of them, minimal impact on 0.9% of them, etc. Do people in Uzbekistan care about being able to access Google, Wikipedia, Amazon, CNN, the BBC? I rather think they do.
No one likes pointing out to random minor countries that their presence on the Internet is far more in their interest than it is in anyone else's. But that doesn't make it any the less true. So, in general terms, the choice they're getting is ``largely anglophone, largely UTF-8, or nothing''.
Slashdot Mirror
ian
Most space `science' falls into Feynman's comment that he'd heard they did science, but it never appeared in journals he read. It's mostly work to provide the systems that permit them to do more work to provide the systems, etc. All the talk of exotic zero-g materials is spurious: if you could lift water into orbit and transform it into gold using only your bare hands and the power of zero gravity, it would still cost more than gold is worth. It's just possible that materials whose cost per gram is more than gold (some exotic drugs, perhaps) might be economic, but there's never been the slightest explanation as to why zero-g would help make them.
Heating effect in a given cable rises by the inverse square of the voltage for constant power, which is why power transmission is done at 25KV or more. I don't think there's much OLE on railways in the USA, but it would be interesting to see what voltages they selected for that.
ian
A search of the NAS/UWT website for the keyword `Internet' reveals no policy that could remotely constitute a call for censorship, and without a conference resolution a union general secretary has absolutely no authority. There's the usual guff about cyber-bullying, in a document that any union-watcher will recognise as the tortured prose of a composite (pronounced, bizarrely, to rhyme with website) motion, but even that is just a call for web filtering in schools.
Unfortunately, /. runs on its own timescale and fact checking takes too long, but I'll try to find out from Chris Keates what she actually said. The absence of direct quotation makes me suspect it wasn't what people think.
Anyway, even if she called for the Internet to be closed down tomorrow, she's the general secretary of a minor NUT-affiliated trade union. That's not exactly a position of major political power, is it? All sorts of people believe all sorts of things: being able to execute them is rather a different proposition.
ian
The BBC transmitted ideoforms (which were nothing like a secure as is often claimed, but that's another story) to the French resistance via transmitters aimed at France, yes, but in 1943 it was no secret that the British were keen to help the French resistance, and it was no secret that there was a war on. Today, the list of countries in which the CIA have agents is itself interesting intelligence.
ian
My uncertainty comes from the following scenario:
-
I release my product under the GPL.
-
Various people beta test, patch, enhance and generally improve my product, believing it to be GPL'd.
-
I then dual-license as a commerical offering to those that want it.
Now yes, you could argue that the people submitting patches knew the score, or that if they had reservations about possible license changes they should have made appropriate legal moves. And certainly this pattern of events is neither illegal nor (for most people) immoral. But the point is that the value I would be selling is partly due to the improvements that other people donated, and I'm not sure I approve. ``I'm not sure'', note, so try to avoid the hyperbole of stacked assumptions starting from my being sure. The issue isn't the main author's work, it's the intangible value created by the work of others.Mind you ``Revenue? GPL enthusiasts can't even spell it'' is a tempting cheap shot.
ian
-
Firstly, the admission that you don't own the filesystem (in the sense of employing all the major contributors) is a worry for many customers.
-
Secondly, if you want to put your own secret sauce into the filesystem (perhaps hooking it more intimately into your product's volume management, or providing a shortcut API into your block level IO, or doing extra things for fast failover between control units, or whatever) then you don't want to have to pass this stuff out GPL'd.
-
And finally, if you want to use an otherwise-GPL'd filesystem linked into a non-GPL real-time executive like VxWorks (no relation to VxFS, confusingly) or QNX, having a non-GPL version of the filesystem probably saves everyone a lot of lawyers bills.
I'm not sure I approve of this as a GPL enthusiast --- hey, I had code on the Emacs 17.61 tape! --- but as a customer I don't think I care too much. You don't get to have much oversight of the components used in products you buy unless you're entering into the wild world of source escrow, and buying a non-GPL'd version of a GPL'd product is no different to the OEM buying something completely closed, and in many ways better (I still get the many-eyes thing, up to a point).ian
ian
ian
It's a classic case of `unintended consequences': the assumption circa 1995 was that the big criminal issue was theft of service. At the time the only way to get a mobile was on a contract: you needed a bank account that would take direct debits, and some proof that you weren't totally sketchy with regard to credit. The only way the operators were going to build out their userbase in that environment is to redefine sketchy, and even then it's not going to get anyone under 16. So in the new world, sketchy would be redefined upwards (ie anyone the slightest bit dubious is refused credit), but pre-pay is universally available. Shazam: instantly you have a massive increase in mobile phone userbase. Fraud drops, because (a) GSM cloning isn't as easy as ETACS cloning (b) your contract customers are better risks but (c) most importantly the prices are falling and phone fraud isn't really worth the candle.
However, for crims, losing the ability to sell phone time to Pakistan for half the standard rate is a small bump. Suddenly having access to completely anonymous, mobile, non-suspicious (as compared to hooky PMR equipment) point-to-point communication is like heaven on wheels. But once the genie is out of the bottle, how do you deal with the problem when every eleven year old has a phone (my ten year old daughter claims to be the only child in her class without one: I think she's exagerating, but not by much)? That forces you to permit cash top-up from people with no ID.
ian
ian
ian
To me, the conclusive proof that American notes (except for the $20, I think) are indistiguishable is that US cashiers always, always say ``from ten'' or whatever as you hand the money over. That's unheard of elsewhere.
ian
What compelling feature could be offered in a year's time to make people change player? Especially as, as each day goes buy, every tune bought for ITMS is locking people deeper in to the Apple platform?
I can think of things that niches might want --- Bluetooth headphones, say --- but the idea that any vendor can come to market and displace Apple from the MP3 market is a reach. The barriers to entry --- specifically, the synonymous use of `iPod' with `MP3 player', the fact that Apple have consistent marketing, branding and ITMS from $70 to $400 and the massive `dock' ecosystem for cars and other accessories --- are hard to overcome.
And, as several people have pointed out, the channels for products like this are channels that microsoft don't own, into markets where their branding isn't strong. Products have to earn their shelf-space, and Microsoft can't leverage Walmart. And I suspect that the key branding in the mass market isn't Apple, it's iPod.
ian
ian
Also, for a lot of people, cassette-shell adapters are a better choice. They don't pause the tunes when you get a traffic announcement, but they do at least allow the TA to come through.
ian
The government is forcing the UHF TV shutdown because it thinks it can then auction the spectrum. The theory is:
- Spend millions promoting shutdown of analogue TV
- Spend millions providing free STBs to the poor, photogenic and otherwise tabloid-friendly.
- Lose a few marginal seats as locals find that their TVs don't work.
- Put spectrum up for auction, saying ``it'll be as valuable as 3G licenses!''
- ???
- Profit! To spend on a war that no-one wants.
You know how politicians live in the past? Brown's lot still think that Wired is pretty damn edgy, and are getting their financial advice on technology from Kevin Kelly.ian
Certainly today, MSDN can't possibly cover its costs directly. But Microsoft's great asset is all the ISVs who develop to their API. I didn't follow Apple `back in the day', but I get the impression that their current supportive attitude to developers, and the fact that they ship compilers and debuggers with every machine they sell, is a new posture. If that's true, and Microsoft stood alone in supporting a software ecosystem around their product then, and the very words choke me as I type them, that might be quite a major reason for their success.
ian
ian
Were some random non-UTF8 country to make interworking with the rest of the Internet harder, it would be cutting its nose off to spite its face. For the G7 countries (yes, G7, not G8), the value of Internet connectivity to random minor countries is minimal. The value to those countries of Internet connectivity is large. Do US users care if Uzbekistan is on the Internet? No: it has zero impact on 99% of them, minimal impact on 0.9% of them, etc. Do people in Uzbekistan care about being able to access Google, Wikipedia, Amazon, CNN, the BBC? I rather think they do.
No one likes pointing out to random minor countries that their presence on the Internet is far more in their interest than it is in anyone else's. But that doesn't make it any the less true. So, in general terms, the choice they're getting is ``largely anglophone, largely UTF-8, or nothing''.
ian