Forgot your password?

Comment: Tape is dead (Score 1) 193

by ledow (#48465639) Attached to: Is LTO Tape On Its Way Out?

Tape is dead in most small business already. There's just no point in it. Putting all your eggs in one tape-format basket means that when you have a fire, not only do you have to spend days getting your data back, you've also got to pay a small fortune for the machine that can do that (if those drives even exist any more).

Extended NAS systems are what I see taking over from tape. Automatic network replication to a copy that then gets marked read-only, forwarded and verified to other NAS further down the line, etc. Offsite backup can be as simple as taking a NAS home for the smaller outfits, or mirroring to an off-site NAS or cloud storage.

Sure, one NAS is not "backup" on its own. You don't want to be able to overwrite written data automatically, But if you have enough of them, they become very reliable at a cheaper pricepoint than handling tape backups. This means you can buy more of them, and buy "more" reliability.

If your data is mirrored this way, in enough places on devices that don't automatically modify themselves to the very latest data, then you have backups. The old-farts of IT like to tell you otherwise, and bring up "RAID is not backup". But, sorry, at scale, and managed as such, it actually is. Especially when you have a RAID of RAIDs that you manage and cycle and secure properly

And buying something of that scale is something that a small business can do much easier with spinning disks than they can with tape. For the cost of the last tape drive and initial tapes, I can buy half a dozen NAS devices of larger size that - in a pinch - I can run VM storage from directly (iSCSI, etc.) if I want.

Restore times are better. Capacity is much better. The automation is much easier. Networking is so much easier. Restore is so much more "obvious" if your network guy suddenly dies.

There are concerns, but so long as you are considering them, they are rarely barriers. Most places do not need to keep data for dozens and dozens of years, and offline retention for most drives is actually stupidly impressive. Those that do generally cycle backups onto new media anyway, so nothing's changed there.

But drives are scaling SO MUCH faster than tape that it's stupendous. Hell, tape is often within range of SSD storage (not that I'd trust retention on that without a lot of redundant backups until the tech matures properly). And you can literally pop down to the local computer shop and buy a NAS that will backup your entire small business network - something you can't really do with tape. Do that a couple of times a year and you have a backup that's probably superior to whatever you had before.

I work in schools. The previous school had tape but it was a "last resort" backup. We never restored from one in 5 years despite two server failures. Always, there was a NAS or even just a drive that had the data available which could get us up and running NOW, this second, rather than 8 hours down the line. By the end of that employment, we'd basically abandoned tape use as anything other than putting one in the fire safe next to the drive backups (which were mirrored to encrypted cloud storage). The IT audit I had when they were trying to justify pruning every member of staff? Couldn't find a fault in the backup strategy and was able to see more copies of the data than they actually believed necessary.

The school I work for now - they employed me because their last guy didn't backup (to tape or anything else) and they lost the entire data and had to send live drives off to a data recovery firm. My prime concern has always been not repeating his mistakes, and showing that their data is safe. I NAS everything and there isn't a tape around.

It's pointless to worry about tape and expensive drives when you can just slap a NAS in each building, mirror to them as appropriate, mirror that to an off-site NAS, mirror that to the cloud, etc. The extra space available means you can store more historical copies in the case of a rollback.

And even from personal experience - every drive I have ever had eventually ends up getting copied onto the bigger, newer drive I just bought and takes up less than 10% of it. As such, I have brand-new drives that are still hosting entire copies-of-copies-of-copies-of-copies-of-copies of my entire hard drive from 10 years ago. And each copy is there because it's come from the copy I did to a bigger drive. And then copied that to a bigger drive. I still have the original disks, but I'll tell you where I go to when I want a file I know I have and want to make sure it will work first time.

Properly managed NAS is, as far as I'm concerned, vastly exceeding the definition of "good enough" for backups for everywhere I've ever worked.

Comment: Re:Race (Score 1) 1023

by ledow (#48457137) Attached to: Officer Not Charged In Michael Brown Shooting

A veritable symptom of the heart of the problem.

If there aren't even numbers of all races, someone, somewhere will say it's "unfair". If there aren't proportional representation of all races, someone else will say it's unfair. And so on.

At no time does it occur to anybody that saying "another race can't represent justice to my race adequately" is JUST as racist as anything else.

Comment: Re:Don't fear the singularity (Score 1) 433

by ledow (#48448129) Attached to: Alva Noe: Don't Worry About the Singularity, We Can't Even Copy an Amoeba

Well put.

Although I work in IT, the stuff I used to do is no longer profitable - because there are tools that allow everyone to do those things nowadays. Similarly, the ubiquity of commercial devices means that IT is no longer demanding huge resources, nor that people are unfamiliar with how to manage the machines.

Granted, there's always going to be someone doing it wrong somewhere, but there are entire schools running networks and banks of iPads without a single technical member of staff on-board. I know, I used to work for some of them. I have a job in setting them up, but staying around is harder to justify for the smaller schools. Eventually, that kind of creep will move into the teaching staff and before you know it someone is managing a thousand machines alongside their day job.

Even IT isn't safe from his, as a profession. Coders are being ousted for app-creators and more simple languages, etc.

We are likely to hit a point where there aren't job left first, yes, but we're also more likely to be self-sufficient by then and - thus - probably not require much in the way of work also.

How the monetary system will cope is an interesting question, but over 100 years ago we were working fields, growing our own crops, and that was us done. If/when most things are automated - specifically food produciton - and the manual jobs aren't present, we're going to have a lot of well-fed idle people, and how they will entertain themselves becomes a real problem.

Comment: Re:AI and real intelligence are the same (Score 1) 433

by ledow (#48448067) Attached to: Alva Noe: Don't Worry About the Singularity, We Can't Even Copy an Amoeba

This is the most common misconception, and the point of this article.

The brain is not doing what it does because of scale. And the scale is HUGE. You'd need millions or billions of machines all capable of joining to 10,000 others directly, and able to break and remake connections all the time in order to come CLOSE to simulating a brain of any significance.

But that won't get you AI, and certainly won't get you intelligence. That will get you a computer built on neural networking principles. It's like suggesting that if we just ordered carbon atoms in the right way, we could make a human. Well, yes, technically. But to do so is unbelievably more complicated than making a human waxwork out of carbon atoms. To make it "alive" from there is orders of magnitude more difficult again.

AI isn't close to actual intelligence. It's an incredibly impressive, powerful tool - no doubt - that is capable of extremes of computation. But it does not hold several vital factors in the quest for intelligence, and certainly not those for "life". Every AI project you've seen is "written". It's crafted by humans. It has to be told HOW to think. And it's scope in doing so is extremely limited.

Of course, there's nothing stopping us making a program that can make programs. But if you've ever seen the results, they are incredibly disappointing. Almost like they would work if only we could put several billions years of evolution into them and write off vast portions of the results as not viable.

Huge amounts of computing power can do things that we used to have to use humans for. But they cannot, and are decades or even centuries off doing so, do some much simpler things that even babies take for granted after a month or so.

Scaling up what we have now has been done. It's part of the driver between cloud computing, Internet2, and a lot of other scientific firsts that we pulled off in order to do them. What you get is a mess. Missing billions of years of priming the networks really hinders any sort of effort you do. Scaling up just makes it a bigger mess, and even harder to analyse, coax, or "design".

The problem is that we lack any way of expressing "do what you want to do". to a machine. AI has it's limits, no matter how cool it looks when you first dig into it. And we've been battling those limits for decades because, if it was just a matter of throwing computing power at them, Google wouldn't need to manually filter websites or mis-categorise images - it would just get it right.

Comment: Re:people drop their phones :( (Score 2) 199

by ledow (#48444287) Attached to: Corning Reveals Gorilla Glass 4, Promises No More Broken IPhones

If you have a naked phone, what do you expect?

Fuck, I drop mine at least one a month onto something solid. Of course if it hits a stone, or the edge of a rough surface, it's going to scratch or shatter.

Put it in the most basic of cases so the force (not the sharpness) goes to the screen and it's fine. I have never, in my life, broken or scratched an electronic device like that.

And, honestly, yes, I've had some doozies! When you phone cartwheels down a set of marble staircases in a hotel, and smashes so hard every component falls out, you think it's game over. Pick it up, put it back together, all works just fine.

What phone? Galaxy Ace (the cheapest junk you could buy at the time), S4 mini, etc..

Electronics don't survive mishandling. But a four-foot drop onto concrete is nothing. Absolutely nothing. Your pen survives it. Your USB stick survives it. I've seen laptops survive it (but that's mostly luck, admittedly). But your remote controls don't shatter into a million pieces when you drop them off the sofa. I've seen plates and bowl survive worse unscathed.

It's all a matter of dampening and removing the sharpest points. It takes one, tiny, shard of stone a few mm tall to be the pressure point that smashes your screen. Put it in the cheapest case from Amazon, it's covered with 2-3 mm of foam or board, no more pressure point.

I have launched phones (accidentally) across entire school playgrounds. Not once have I broken one, except once the plastic on the battery catch went loose and I had to pay about 1GBP to replace it.

Phones used to have raised edges, the screen would be the last thing to contact the floor. When you have a phone where the front is entirely glass, edge-to-edge, nothing is going to save you if you drop it. Except putting a wrap around it.

I blame Apple "design" again - yeah, looks pretty. Totally fucking impractical, however, and unfit for purpose. Gimme a 2mm raised edge around it and I'll never have to replace the screen. Fuck, just unpacking iPads and iPhone from the box can be a hazard because their "design" teams didn't think to put fucking fingerholes in the packaging. You either have to shake the thing upside-down or tear your brand-new box. I know, I unpacked 200 over the summer for the school I work in. It was a damn nightmare.

Apple's takes "design" to mean "looks pretty". I take it to me "is a good engineering way to make this device that makes it look pretty as well as be user-friendly". Stop making phones with edge-to-edge glass if you expect people to use them in the real world. I'll happily pay the cost of an Apple device for a Samsung device that has a completely rubberised raised exterior.

Comment: Re:I have been wondering about WhatsApp (Score 1) 93

by ledow (#48427015) Attached to: WhatsApp To Offer End-to-End Encryption

I've heard of WhatsApp for ages.

Most of my contacts are on Whatsapp.

But I didn't start using it until my Italian girlfriend introduced me to it - because texting internationally via WhatsApp costs you 63p a year as opposed to nearly that per text!

All the ex-pats and foreigners that I know seem to be the biggest users of it.

Comment: Re:my takeaway (Score 1) 219

by ledow (#48425911) Attached to: Lessons Learned From Google's Green Energy Bust

This is my biggest problem with "green" things.

Sure, we can make changes. But what impact will the changes have and, compared to what will happen otherwise, is that better for us or not? If the changes enforced by the new ideas actually cost us more than just carrying on anyway, or gain us nothing, we're really just wasting time.

As such, I often think that all the "renewable" debate is taking too long. Can't we just pump that money into fusion and be done with it? That would give us a kickstart to having more "clean" energy than we know what to do with and THEN we can start putting serious efforts into getting the hell off Earth or compensating for the damage we've done (and, literally, we could start expending energy directly and purposefully to reverse the changes we're making if we have that much energy just lying about to be consumed for the next few hundred years).

I've read a few writeups now where they over-egg the situation, blanketing the whole of my country with green power (wind, wave, tidal and solar) for instance in one of them, and then working out what difference it would make. The answer is quite often little.

The lack of long-term insight into the COSTS of doing what needs to be done don't give me confidence in the research. Sure, we can reduce consumption, reduce pollution, reduce all kinds of metrics, at a cost. But quite what would that mean for our way of living, and how much would that cost. How many people die because they can't put the heating on if you increase the cost of electrical production (or decrease the available power at certain times) by X%? It never seems to be included in the research.

Green tech solutions to climate change often rely on complete worldwide co-operation, full funding for research, global deployment and testing, and all resources pushed into making it work as well as possible. The cost of that is rarely studied to the same detail as the potential benefits.

If the cost of stopping the sea level rising is a million deaths worldwide because of energy shortage, increased costs, raised taxes, job losses, or whatever, is that better or worse than even abandoning countries and low-lying areas entirely? I have no idea. And unfortunately, it always seems that no-one else does either.

Comment: Re:Bad sign. (Score 2) 219

by ledow (#48425855) Attached to: Lessons Learned From Google's Green Energy Bust

It's all a problem of advertising hype.

Old isn't "old".

It's tried.

Whenever someone says they want to throw out the "old", especially in computing terms (e.g. init systems, *cough*), I mentally substitute those words. And when I'm not immediately keen to jump on board, I get mocked.

Until the project flops, that is, or the reinvention of the wheel, or the having to sacrifice functionality, or the realisation that two systems are needed, or whatever.

There's a reason that large companies "extend" their existing products, rather than replace them. They "build upon" or "enhance", they don't rip out and start again.

Sometimes, yes, it's needed to start again. Linux printing went through enormous flux and had to be reinvented. But always be suspicious of "new" until it's proven itself better than "old" and is also old enough to be considered "old".

You can see it throughout computing. We used to have shared computing and terminals, then everyone got their own PC powerful enough for everything, then we tried to move people back to shared terminals over many years (thin clients, etc.) and made the same mistakes as why we abandoned those ideas in certain use cases (and yes, it's often more about use-cases, than about the technology as a whole), and now we're learning that cloud probably isn't the best idea for everything and so on.

In the best tradition: "Those who do not understand Unix are condemned to reinvent it". And it applies not just to Unix.

Comment: Voice recognition - AI (Score 1) 62

by ledow (#48425177) Attached to: US Intelligence Unit Launches $50k Speech Recognition Competition

Given my own personal experience with voice recognition, it's not a problem we can throw money at. We can throw money AWAY trying, but we haven't improved much in many, many years of trying.

I don't have a particularly poor speech, or unusual accent, and English-speakers all understand me - even foreign English speakers like the one I live with. But speech recognition has always been an absolute flop unless I want to learn how to talk to the computer, which is the exact opposite of what I want to happen.

Since the first days of Dragon NaturallySpeaking, it's never been worth the training time even if all I'm doing it dictating serial numbers, or product codes, using simple single letters spaced out in a silent environment. Telling the difference between "eight" and "A" is much more involved than just context matching on a rough FFT of my voice.

And, as has been pointed out, someone who can do this will get a damn sight more than $50k reward as the patents would be worth billions.

To do it properly, we're really looking into problems that are the equivalent of the higher functions of AI.

Comment: Re:TFA: "what happened to freedom of speech?" (Score 2, Informative) 306

by ledow (#48417407) Attached to: UK Hotel Adds Hefty Charge For Bad Reviews Online

For a country "without free speech", we do pretty well and I can get away with ten times more than I could ever do in the US.

The hotel were talked to by Trading Standards, and have immediately revoked the policy (because it was legally dubious right from the outset). They are currently being spoken to also about refunding this "fine" despite the idiots signing a piece of paper that says they agreed with it (which is also legally dubious). And there doesn't need to be any change in the law because already, by the laws that exist, including by default many EU laws that do include free speech, the area is more than well enough covered, thanks.

The reason it makes news is not because it's legal to do what the hotel did. It absolutely is not legal. It's because it's OUTRAGEOUS to even try, given the current laws. And they are quickly finding that out in more ways than just the Streisand Effect as they now have a lot of lawyers and government departments breathing down their necks.

There is nothing whatsoever in law that gives the hotel the right to do this, only the opposite, and no need for explicitly stating this beyond the existing laws. UK laws do not explicitly enshrine a number of things, like the "official" language of the country, the rights of free speech, etc. because they are just automatically entrenched in the law and the case law.

We don't have a "You have the right to say anything" law because we haven't needed one. You have pretty much the same rights as anywhere else in the EU, and a damn sight more rights than the US.

Remember the UK "super-injunctions" that supposedly stopped people talking about the very existence of another court injunction? It went down the pan because the media basically ignored it, made it front-page news for several months and then exposure of their existence meant they were dead - legally speaking - from that point. I can't imagine US media fighting like that for a second.

And the UK's defamation laws? We gave them to the US:

They've been through changes, and a number of high profile cases lately have resulted in changes, but asking someone who says you're a paedophile to prove so (and not be unchallengeable in court unless you can prove you're not) is not the end of free speech. And all those laws have been fixed for quite a while now.

You cannot, and cannot ever have been, successfully sued for your reasonable opinion, in any first-world country in modern times. What you can have been is defamed with absolute untruths and then had the defamer hiding behind "his opinion". That's always been true in any system.

Hence, as a Brit, I've never been one to hold back on forums, or otherwise. The threat to me is zero. I'm either clearly expressing an opinion or stating fact, and you cannot ever have been successfully sued for that.

The problem with the US is that they think they are a free country. However, whenever I've been there people are shocked at the opinions I express, the way I express them, and friends have honestly believed that I would get into trouble for expressing them. Yet, in the UK, if anything I'm considered quite passive.

The UK defamation laws give this place NO RIGHT whatsoever to block reviews of their business, nor to charge for them. Hence why the policy has been revoked on the same day and why government departments are "in discussions" with the hotel. That's English politeness for "We're currently explaining the law to them, and won't stop explaining it until we have to take them to court or they stop doing it of their own 'free will' ".

Comment: Re:Go back in time 5 years (Score 3, Insightful) 566

by ledow (#48416757) Attached to: Debian Votes Against Mandating Non-systemd Compatibility

Agreed. I don't get systemd. If people want to use it, fine. But, like Windows 8 taught Microsoft, giving people a one-click way of going back to the old-and-tested interface is always a) possible and b) sensible.

If systemd was really that good, I wouldn't need it foisted upon me forcibly, I'd be voluntarily choosing it rather than the default init on every distro I boot.

I think worse than pushing it on your users is this - saying you won't support the old way of doing for those that don't want to change.

All we need is one remote-root in systemd and people might start to think again.

Comment: Re:"Just pay extra..." (Score 1) 472

by ledow (#48409041) Attached to: Elite: Dangerous Dumps Offline Single-Player

In my day too.

But you can't really have multiple peer-run servers co-operating in an MMO without cheating taking place. And that's basically what they've turned off - you won't get any "alternate universe" run by gamers on their own servers here.

The days of MUDs, and even just community servers, died when cheating became profitable. Sure, you can run your own TF2 / CS / other servers, but they won't be a part of the official game unless they defer to Valve's central servers for weapon drops, etc. As soon as any kind of digital content is the preserve of only paying people, all control is taken from the community's hands - and you can see why that has to be.

P.S. I have, or still do, run many servers for CS, CS:CZ, Altitude, OpenTTD and a multitude of other games.

Comment: Re:Apparently "backers" don't understand the term (Score 4, Insightful) 472

by ledow (#48409023) Attached to: Elite: Dangerous Dumps Offline Single-Player

Agreed. However, crowdfunding for something and then completely abandoning the idea is only going to prompt ire.

You're still obliged, in law, to deliver what you promised you would. Sure, it's almost impossible to enforce that, but you can't go spending the money on holidays in the Caribbean nor can you use it to develop an entirely different game or product. People have had their projects shut down and been chased through the courts for failing to deliver on Kickstarter. It's not easy, but it's no different to any other payment. If you misrepresent what you're going to receive in return for someone's money, it's fraud whether it's an investment, crowdfunding, or written into a sales contract.

To be honest, E:D is my worst Kickstarter. I've contributed to a handful and they've all been great, whether for physical products, digital content, or whatever. I've got several rare beauties of games (I collect mathematically-interesting board / card games), good video games on Steam (including copies), video graphics hardware, all kinds from it.

E:D is disappointing, however, mostly because of the constant demands for more money and the complete under-delivery of the base product. I backed it out of retropathy, yet I have ZERO idea how it plays as yet. That doesn't bother me. But being told "Just X amount of money more and you could see how it plays!" every week in an email is really grating. I regret backing E:D just because of the lack of real return for the backers as yet, and the constant demands for more cash.

That said, it was such a pittance that I don't really care because I always follow your "rule": Never crowdfund with money you can't afford to lose.

Comment: "Just pay extra..." (Score 5, Informative) 472

by ledow (#48408995) Attached to: Elite: Dangerous Dumps Offline Single-Player

Elite Dangerous is a shower.

I'm one of the backers of the Kickstarter. I am absolutely TIRED of being asked for more money for every damn thing they do.

The number of paid Alpha's, premium content, several Beta's (Beta Premium!) is unbelievable and they seem to want to make me wait until the very day of release before I get anything out of my backing unless I pay more money.

Sure, I get a "reserved Commander name" and a couple of bits of digital content but I have seen nothing of the actual game in all that time except for the occasional screenshot. They have probably made more from the Beta's than they have from the Kickstarter, and every damn newsletter is "just another $15 will get you this...".

I've totally lost any interest and regret backing but, unlike some, I'm true to my word so have written off the money I've given them so far. I've truly not expected to see the game because every preview/screenshot/update still without any access by myself but with begging all the way through it just disappoints me further. If they are milking it that early, what the hell is going to happen in-game when they want to form the economies?

I'm honestly fatigued by the requests for money, which they are still putting in every newsletter. It makes me worry that any final game is going to die from budgetary shortages the second it's release because the begging is so intense.

Meanwhile, all I have to show for backing it is a cart with one item "bought" that I can't touch for another month or so and that's all I ever had.

Honestly? I'm sick of it already. And I haven't even got to play it. Given that it was one of the largest and most successful Kickstarter projects there was, I'm a bit disgusted by how much more they seem to want in order to let me see how it plays, even in a tiny demo.

It's gonna be an over-hyped flop, isn't it? Or crash and burn in the first few months when the servers can't be kept running due to lack of budgeting. And to leave it until NOW to tell people about the lack of single-player, while you're still pasting in 4K screenshots and plugs for various books written in the Elite:Dangerous universe (that doesn't exist yet as far as I'm concerned)? I just don't care any more.

The one Kickstarter project that I really regret backing.

"When the going gets weird, the weird turn pro..." -- Hunter S. Thompson