Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
For the out-of-band Slashdot experience (mostly headlines), follow us on Twitter, or Facebook. ×

Comment: Re:The Nature of Central Banks (Score 1) 323 323

by ledow (#50013459) Attached to: Greek Financial Crisis Is an Opportunity For Bitcoin

Ah, this would be the Iceland that "had to obtain emergency funding from the International Monetary Fund and a range of European countries in November 2008". And also the Iceland whose economy is "small and subject to high volatility".

The Iceland whose GDP is worth less than what the UK spend each year on weddings alone. The Iceland whose debt to other countries is actually more than 100% than that pittance of GDP.

With 3 people per square kilometre and less than the population of a medium size town in the UK (or any one single London borough).

Sorry, pal, you can make all the claims you like. The ONLY counterexample you provide is actually doing no better than anyone else, and is on a scale so small as to be statistically useless anyway.

I'm not a banker or economist, by the way, just a mathematician.

And when the Icelandic banks crashed, other countries had to compensate savers who had been using them as the Icelandic banks had zero actual protection for their customers at all. All that teaches you is that people WON'T invest in Icelandic banks because they just lose their money if it all goes wrong.

Sure, there's a point at which you have to let the banks fall over to save other things, but that's true of anything - even Greece today. We're choosing to let them collapse rather than extend more and more bailouts to them. It's just a question of scale.

An country that's got the population of Pittsburgh and the GDP less than a UK mobile phone network's entire worth is - pretty much - a nonsensical thing to extrapolate to the world economy.

Comment: Re:I Wish Mine Had Been Blocked (Score 2) 23 23

by ledow (#50002487) Attached to: Samsung To Stop Blocking Automatic Windows Updates

Or, like EVERYONE tells you to - backup your damn machine. P.S. If your backup doesn't get you back to exactly where you were last week, it's not a backup, just a bad data copy.

Also:

https://4sysops.com/archives/d...

However, for years, people have mocked my decision to NOT have auto-updates turned on. I only press update when I know that my machine is backed up, there's a fix I need to deploy, and I have the time / willingness to do it.

No, my machine doesn't have viruses etc. (I've had precisely one in my life and that was from a demo copy of Sin on a PC magazine coverdisc - which shows you how long ago that was!) because I abide by simple security practices that mean Windows doesn't NEED to run lots of random third-party executables to do what I want.

There's a reason that MS *can't* block WSUS for business users being used to stop automatic updates for Windows 10. Because we'd tear their fucking heads off. Windows updates have caused shit like you describe since their introduction. Sure, most people won't notice, but if it only happens to 1% of computers regularly deploying updates the chances are that none of your friends will have had those problems. But similarly, with the same odds the chances are that in any large deployment AT LEAST one machine will fuck up from automatic updates every month. Fuck adding that to my IT burden.

In work the other day, one of my users was accidentally given a brief window when they could receive updates from Windows Update instead of WSUS (I'd accidentally pulled them out of the client group on WSUS while looking for a test machine). In that short opportunity, it took it upon itself to update from 8 to 8.1, thereby breaking the finance software that we use permanently. Additionally, the desktop now gets a crash in in a mp4 video dll every 10 seconds that you can't stop crashing without reverting the update associated with it. Seriously, no newer patch fixes it or I'd deploy it in a second. And I had to give them RDP to a plain Windows 8 machine to finish their finance stuff temporarily while I revert their config.

Seriously, automatic system-level updates without user interaction is the most stupid fucking idea in the history of bad ideas, not to mention not being able to PERMANENTLY say no to a particular update, and having NO proper way to system restore to a point before the update applied and stop it (in the majority of cases - I've yet to see system restore do what it promises but I've dealt with lots of users have accidentally restored their personal laptops back to factory settings or unrecoverable states using it!).

If you work in IT and haven't yet realised this, I really pity you. Servers, internet-facing services, maybe but there you have the tools to deal with this crap and STILL shouldn't be blindly pushing updates anyway.

Unmanaged clients that aren't eligible for WSUS because they are home-use? Back those fuckers up and turn off automatic Windows Update.

Comment: Assange. (Score 2) 212 212

by ledow (#49994147) Attached to: France Could Offer Asylum To Assange, Snowden

Fine.

A) He can't get to France without stepping out of the door.
B) We arrest him the second he does that.
C) He stands trial for skipping bail etc. (unfortunately, his life in the embassy is prima facie evidence of guilt in that case, no matter the mitigating circumstances).
D) He serves whatever sentence he gets for that (hard to imagine he doesn't get one).
E) Then we're required to honour any EU warrant that was issued.
F) Then he's either out of UK hands, or able to go to France freely anyway.

After that you can discuss whether or not asylum in France is justified - methinks that the political climate may have changed somewhat by then (in which direction, who knows)?

Comment: Low-latency (Score 3, Informative) 45 45

"Low-latency"

Yeah. Right.

At absolute best*, with no processing time, buffering, contention, sharing, delay or retransmission whatsoever through the entire process, with optical switching all the way along, with routing direct to each users and end-point, with not a single blip or anything else, that's going to be more delay on top of normal Internet latency.

Fast, yeah I can't argue that one way or another. But that's about volume, not delay. If you turn on a tap (faucet?) in the US and then put your head in the other end of the hose in the EU, it doesn't matter how big the hose is or how much water is coming down - it will still take a long time for the water to arrive. When it does, of course it can be high-pressure, huge volume down a ginormous hose. But delay will still make it useless for telephony, streaming, and a range of other purposes.

I'm all behind the concept, but don't claim low-latency as if it could possibly compete with any other technology out there - my mobile phone barely get 100ms delay to even default gateways).

(* Even LEO is 190km up. A round-trip from that to a base-station to a 0ms Internet back to the satellite back to the ground is going to be:

4 x 190km = 760,000m
Speed of light is 299,792,458 m/s.
3ms or thereabouts?

Maybe tiny in theory, huge in practice because none of the above theoretically-ideal-scenarios actually exist.)

Comment: Re:What's the score now? (Score 1) 77 77

by ledow (#49984205) Attached to: NVIDIA Begins Supplying Open-Source Register Header Files

Could be used on more operating systems doesn't translate to more sales.

And it's not "without them having to do anything". Just patent/copyright-auditing the proprietary driver they had in order to open-source would probably wipe out any extra sales they gained alone. Let alone ongoing maintenance, catering for all the Linux kernel changes as they try to get it accepted into the kernel for several years, bus-changes, new versions of CUDA / OpenGL translations etc.

Honestly, look into the costs. They wouldn't actually make that much at all, and it would cost them dear to try. It's not as simple as "let's just push our existing codebase to github", and certainly not if you want anyone to make any sense of it, and certainly not if your codebase changes nearly every week (I've done 20-30 driver updates to my laptop over the last 2 years, for a single chipset that's hardly used any more - god knows what kind of churn they have in their source code management).

And for a handful of geeks on a minority operating system that has things like Steam, etc., yes but they're not going to RUN OUT and buy a new nVidia card just to use those games. They either buy new regularly anyone to play them on Windows, or they will re-use what they already have.

Some places it makes sense. But here, I have to agree with nVidia that it probably doesn't make any business-sense at all.

Comment: Re:What's the score now? (Score 1) 77 77

by ledow (#49983921) Attached to: NVIDIA Begins Supplying Open-Source Register Header Files

I'm an Open-Source advocate, don't get me wrong.

However, they are under no obligation whatsoever, so why should they? What advantage do they get from opening them? What's going to be the thing that will make them want to open their drivers? What's going to outweigh potential patent etc. risks?

Because, as far as I can see, they gain basically nothing. They might get a "good news" article or two but it won't increase their sales significantly at all.

Are we still in the era of hoping that huge multinational companies will do complicated, expensive, liability-affecting things for us out of the goodness of their hearts if we complain enough?

I would love them to, don't get me wrong, but I can perfectly see why they - and others - don't. There's no advantage to a network-card manufacturer not having their network card drivers in the kernel. They don't do anything secret, they operate on well-defined protocols, they all pretty much do the same things, and you can't even start up a computer properly nowadays if your network card isn't supported from the first minute. So the open-source code is next to nothing anyway.

But graphics card drivers? What's in it for nVidia? Will they sell more video cards? No. Can you not boot your machine without an OS driver? No. Can you just use the proprietary drivers? Yes. Is card X that operates at 10billion IPS almost identical in operation to card Y that operates at 100billion? No. Not even close.

And then you have to have OpenGL / Mesa / CUDA etc. drivers, APIs, libraries, etc. All this doesn't affect most kinds of hardware but for graphics - one of the fastest moving technologies - it does.

So I can't blame them. And I can't think why they should beyond political idealism. And I can't think what the OS community could do to change that.

If OS ruled the world and could decree such things and nobody bought things that weren't OS-approved, sure, we could bully them into submission. But we don't work like that.

So what can we, as a community, provide to nVidia to convince them to open themselves up to potential patent lawsuits and huge developer expenses on a regular basis?

I can't think of anything.

Comment: Re:Free? (Score 2, Informative) 85 85

by ledow (#49980153) Attached to: Microsoft Brings Office To Android Smartphones For Free

"Advanced features" like being able to write in a blank document, or change a character in an existing one.

Basically it's view-only unless you pay.

That's not so bad for smartphone, but absolutely no different to the tablet versions at all (which are effectively useless beyond being a free document viewer, which you can get thousands of).

Comment: 104Mb (Score 4, Informative) 85 85

by ledow (#49980071) Attached to: Microsoft Brings Office To Android Smartphones For Free

104Mb download just for Word on its own.

Wow. Seems like all those years of bloated coding are coming back to bite them.

Install office with very limited use on a mobile, and you lose half a Gig of internal storage on your smartphone and still might have to pay for an Office 365 subscription.

Comment: Re:So where are the CVE/Vuln reports for this?Oh,w (Score 4, Informative) 165 165

by ledow (#49976593) Attached to: Car Hacking is 'Distressingly Easy'

There have been public demonstrations, some televised, of certain models of modern car that allow you to change things like timings and injection sequences, via OBD, over Blueooth, using default passcodes.

I'm sure they're all patched now. Of course. No more will that ever happen again.

There's also been demos of being able to DoS certain buses in the car remotely and wirelessly, preventing everything from in-car entertainment to immobilisers from working, etc. using similar techniques.

These things are all out there. Go look. And that's just OBD. God knows what happens when you start tying in Wifi into the car speakers, joining that to the satnav for Internet updates, joining those to the car etc.

You can see cars on the market today, not even particularly unusual or modern ones, that pull in OBD information into the electronic dashboard which also doubles as a music interface and a satnav and a fuel gauge and a Bluetooth phone interface and everything else. It's not at all hard to imagine that such things haven't covered every single possible hole where information from one can leak to another.

And anything OBD-writing is potentially dangerous. As in "blow up your engine" dangerous. Most older OBD systems are nothing more than read-only technical data. Newer ones do more to allow flashing, firmware updates, and even modification of settings that control emission levels (e.g. fuel injectors, exhaust re-introduction pumps, etc.). Add that together and you have one big mess waiting to happen.

There's a reason that you don't buy mod-chips for your engine nowadays that you can swap out to pass emissions test and then swap back to get the "sports performance" of your car. Because they don't need to swap the chips physically any more.

Comment: Re:IPv6 (Score 1) 287 287

by ledow (#49976535) Attached to: IT Pros Blast Google Over Android's Refusal To Play Nice With IPv6

So what you're saying is that all ISPs have to support IPv6, they all have to do so in a standardised (or EVERY POSSIBLE) way, and there's no way to do anything until they get off their butt.

That's what 6-in-4, and the various tunnels were made for, because the ISP's aren't getting off their backside because if they support 6rd but your router was made before that and so doesn't support it, then as far as the users are concerned they don't support IPv6 at all.

But even there, that's FOUR WAYS to do the same thing. All involving third-parties.

What about what *I* would like to do to combat a third-party not supporting IPv6? What if my router didn't support ALL those protocols independently and completely? What if my ISP never adds IPv6, how do I get on the IPv6 network even with all the above?

When you have so many different and competing standards, some EXPRESSLY designed so that the ISP doesn't have to be IPv6-ready, and STILL there's so much choice that your router has to support them ALL in order to claim IPv6 in any significant way, then you're onto a loser.

I don't care about PnP. I care about it being able to be done. But I'm an IT guy. I don't need it to be PnP and I can sort that out for my users. But not without all the ISP's we use onboard, not without explicit support for all the protocols (What if my ISP changes from DHCPv6 to another method? Can they still claim IPv6 compatibility even if my hardware no longer works?), not without having to know how all the standards and protocols work, and not without having to do all the legwork.

With IPv4, you have basically two options - DHCP which is a way of automatically plugging in all the information you would require for the alternate, which is a list of static addresses of various services. With IPv6, there are six, seven, eight protocols that all need different levels of information and co-operation from your ISP, assign different kinds of IPv6 addresses to you, (6rd, or 6in4, or local IPv6, or global IPv6? Who knows?) all work differently, may or may out just route out via a 6in4 address to the wider Internet via any route they like, rather than being provided by your ISP directly, etc. etc. etc.

It's a damn mess. And I have a router that I bought specifically to do this, have enough knowledge to set any or all of them up, could easily sign up to even tunnel provider available, and you know what - I can't be bothered because of the hassle of all that junk.

My websites and external servers are all IPv6 and accept mail over it on a daily basis. I just set up a static, it routes, off we go. My hosting providers provide NONE of the above automatic configuration services. My ISP provides NONE of the above and won't get out static IPv6 ranges.

What you have is a complete deadlock and mess until someone picks a standard and sticks with it. Because if I were an ISP, I'd just say "Sod it, I'm not going to provide ALL those methods and then be accused of missing one out, so I may as well provide none and let the user worry about it". And that's exactly what ISP's are doing.

You know what my solution was? I set up an OpenVPN link to my external servers and just talk via IPv4 to it, sending pure IPv6 through the VPN for a globally-routable IPv6 address that I've reserved for that purpose. It was easier to set it up myself using YET ANOTHER way of doing it than faff with any of the services available, supported or not.

Chemist who falls in acid will be tripping for weeks.

Working...