Forgot your password?

Comment: Re: Finally! (Score 1) 502

by ray-auch (#47588147) Attached to: Judge: US Search Warrants Apply To Overseas Computers

But getting the Irish involved is irrelevant. The prosecutor in this case does not want the actual hard drive with the emails. What would the Irish courts actually *do* in this case?

The EU has data protection laws. If the data is personal data and held in Ireland then Irish DPA applies and the data can only be sent outside of Ireland in accordance with the DPA. The Irish courts would give permission (or not...) for the data to be sent elsewhere outside the protection of the DPA, simple.

Would they send back an email saying "you have our permission to tell your US company to go ahead and retrieve their data from the comfort of their US offices?"

It is not their data to do with as they wish under the DPA, in fact most people would consider it is the customer's data - I certainly consider my emails as my data even if they are on ISP servers.

I don't buy the argument that the data is in Ireland.

Then you have MS on perjury.

MS actually makes a big marketing thing about control of data localization in their cloud, precisely so it can sell it as compliant with the DPA in the EU. A lot of MS cloud business in the EU depends on this, and that is why MS will fight it on principle - they will not care about this one individual's emails at all, but the amount of cloud hosting business they stand to lose if a US prosecutor can simply order them to breach the DPA in the EU is enormous.

Comment: Re: Finally! (Score 2) 502

by ray-auch (#47585505) Attached to: Judge: US Search Warrants Apply To Overseas Computers

The Irish police will say "why are you bothering us? Ask an (Irish) judge like you usually do (via MLAT)". It is the judiciary that need to be involved first. Funnily enough that is also what former Irish attorney general and various senior EU officials have said - the correct approach is to use the MLAT, which is there for exactly this purpose. With a warrant from an Irish court the Irish police will be happy to help, should the Irish operators of the Irish data centre need a hand pushing the button in Ireland.

But the US doesn't want to use MLAT, it wants to end-run around it because they run into annoying problems like having to justify their request properly and whether or not their search is a search or a fishing expedition and stuff like whether the Irish user, in Ireland, has actually done anything illegal in Ireland.

Coincidentally,the UK police don't like MLAT process either and have said so in submissions regarding judicial limits of UK RIPA act. They run into annoying things like the US first amendment stopping them getting details of US users posting stuff on US servers that can be read in the UK and is not legal in the UK. If the DOJ succeeds in bypassing the MLAT process with this trick then some UK police forces will be right behind them ( and probably others elsewhere too). At least one of the amicus briefs points that out.

This will open up a big can of worms if it stands, extradition and MLAT treaties are a necessary process to protect both corporations and individuals from conflicting legal jurisdictions. Microsoft is not (just) saying "we can't, it's in Ireland", they are saying "we are not allowed to under the law of the country where the evidence is". Enabling the international rendition of data at the whim of any country's law enforcement doesn't physically extradite anyone, but it could easily end up making international travel a whole lot more interesting and risky. Be careful about exercising your first amendment rights to hate speech if you ever want a European vacation or business trip. Be aware that some EU jurisdictions are actively trying to outlaw anonymity and pseudonymity online...

Comment: Re:Technical solution to a social problem. (Score 2) 98

Seconded. Except >20yrs and HPUX rather than SunOS.

Police ourselves - yeah sure we did. Act like adults ? er, nope. I figured out several ways to crash machines from console, if someone logged in remote and started using all the resource, I'd crash the machine and move to another. X was completely unsecured in those days but they installed a graphical login. Fake login windows, key loggers, fake error windows (make the guy on the better workstation think it's crashed so he moves off it), check.

Best X trick back then was obviously the ability to put up a window on someone else's screen when the tutor was standing behind them, topless or nude pictures were good (bitmap - time before jpeg existed)... I guess the only thing that's changed now is that the available selection and quality of such images has increased a little. Happy days.

Comment: Re:9 States automatically increased (Score 1) 778

by ray-auch (#47496517) Attached to: States That Raised Minimum Wage See No Slow-Down In Job Growth

Actually, that is the interesting bit.

Of the 9 states that increased automatically, 8 are in the top half of the growth league - conversely 3 out of 4 of the states that voted an increase are in the bottom half.
(link to chart:

The thing about that correlation is that there is actually a plausible mechanism for causation - predictability. When a business decides to invest in hiring more people, you want to try and work out the costs, the potential for profit and the risk of failure. If you are hiring at or near minimum wage, then the level of minimum wage comes into that business plan. If any increase is automatic then you can simply add it in to costs (based on inflation assumptions which you need elsewhere anyway) - it will go up, but that can be accounted for. If any increase is _not_ automatic, then it becomes an unknown in your costs, a _risk_, dependent on politicians. Even in states that did _not_ raise it but _might_ have, that unknown, political risk, may have adversely affected investment decisions.

Businesses like a predictable environment more than anything.

If my suggested effect is real, then there is also a flip side - it will work like this in an expanding economy, but if the economy is contracting then having an automatic minimum wage increase could conceivably accelerate firing decisions - minimum wage jobs might be lost faster because of the known future costs.

Comment: Re:Local testing works? (Score 2) 778

by ray-auch (#47496335) Attached to: States That Raised Minimum Wage See No Slow-Down In Job Growth

Even when people are supposedly more mobile, moving is a big thing for most people so they do not do it.

Here in the UK we had a 50% tax rate imposed on the very richest a few years ago. There were lots of stories about how this was going to drive away people who were successful abroad but in the end it made very little difference because while these sort of exceeding rich people might threaten to take their family somewhere else, but then when they talk to their wife and she refuses to move more than a 20 minute drive from her family and refuses to move the kids out of school and away from their friends.

as well as moving, people at that level can move their income elsewhere, pension it, or defer it to avoid the tax. Avoidance is not illegal (evasion is).

Key issue with the 50% rate is - did it raise 20% more money than the 40%, for incomes over 100k ? If not, then people _did_ move either themselves or their income, and the country's finances got less benefit.

HMRC reckons the income moved - - chart on P28 is very interesting, 25% fall in total declared income over 150k, on the introduction of the tax. Other stats: before 50% tax rate 16,000 people with income over £1M, after - 6000. Gradually increased to 10,000 in following years, but that is still 6k people with 1M+ income who went somewhere else (at least 2Bn in tax they would have paid at 40% rate, gone).

Comment: Re:"here on the Android side" (Score 1) 42

AC above beat me to it - my thoughts exactly.

Take designed-for-desktop/laptop-mouse+keyboard Linux stuff and bolt it together with designed-for-mobile+touch+one-app-fullscreen Android stuff on the same device.

Nope, can't see what could possibly go wrong, sure it will be entirely intuitive and seamless... after all Win 8 was.

Comment: Re: Black box data streaming (Score 1) 503

by ray-auch (#47489739) Attached to: Russia Prepares For Internet War Over Malaysian Jet

For MH17 it will tell us very little - big bang, plane breaks up, end of recording. Obviously streaming would mean that someone would have the data now, rather than be hiding it, but _who_ would have it ?

  - stream the data to the western world and any release would be an NSA/CIA fabrication according to Russia
  - stream the data to Russia and any release would be a KGB fabrication according to the west

Does that really help ?

For MH370, streaming might tell us what happened, but if we just had streaming location and knew _where_ it happened then we'd get even more info from wreckage and be able (with very high probability) to recover the black boxes anyway.

Look at AF447 - 2+ years to find the crash site, less than 2 days after that to find and retrieve the boxes (IIRC).

Comment: Re:Lies and damn lies (Score 1) 503

by ray-auch (#47486941) Attached to: Russia Prepares For Internet War Over Malaysian Jet

Or not, depending on context. See, stuff like "Russian sources now stating that the rebels do not posses such air defenses" could in fact be perfectly true, to understand just add "now" or "anymore".

"we do not have such air defences now - momma bear took our new toys away because we made big mess with them"

Comment: Re:Black box data streaming (Score 1) 503

by ray-auch (#47486203) Attached to: Russia Prepares For Internet War Over Malaysian Jet

Or why it is still one large black box, but not, say, 20 orange 64GB flash drives, which fall all over the place?

An aircraft costs hundreds of millions and 20 flash drives cost peanuts. And it is not a problem at all to write to several flash drives simultaneously.

There are obviously going to be problems wiring data (or certifying a wireless solution) and, more importantly, power to 20+ disparate locations (you want them spread out, right) on the plane, but the big one, IMO is the survivability standards.

Every "black box" is required to have a certain amount of fire, impact / shock, water-pressure etc. resistance. They are also required to have a ULB (pinger, for underwater location) with 30 days battery life (maybe soon going up to 90). All the casing to do that, and the pinger, and the chassis that is supposed to keep them together under the same test conditions, ends up weighing quite a lot (for an aircraft component). Inside it all, these days, is basically a small flash drive - bit higher than consumer spec for survivability, and hence more expensive, but that is all it is.

20 black boxes to meet current regs is too heavy and too expensive, so your problem is changing the regulations. 20 slightly less survivable black boxes that could therefore be much smaller lighter and cheaper sounds great - except for the risk that you then find them easily but the data hasn't survived.

That brings us to the biggest issue - the current standard black boxes are almost always found, so where exactly is the problem you are trying to solve. Real-time data streaming would solve a problem by eliminating the search for the wreckage entirely - arguably saving a lot of money in some cases. Except that not every insight comes from the black box, investigators conclude a lot directly from the wreckage - so I reckon you need to find it anyway.

Real time _location_ streaming (doesn't take much data) plus fully survivable black box is probably the best compromise, significantly shorten some search times - and that is what is going to happen.

Comment: Re:Not if you use the Virtuix Omni (Score 1) 154

I actually have experienced nausea in non-VR FPS games. Descent and the old Duke Nukem used to give me headaches and nausea on and off.

Descent had that effect on a lot of people, even where other games at the time didn't. I think it was due to the full six degrees of freedom movement plus zero-g and "which way is up" effect. I don't think the sentence in the summary about cockpit games applies - unless you are in zero-g in real life.

The world is not octal despite DEC.