Posts by Trevor_Pott

Re: Just one thing

Government anything is typically late, over-budget and doesn't work. Unless that government project is about oppressing their own citizens. Then it's delivered early, under-budget and is creepily efficient.

They can never make something that lets you renew your driver's license quickly, or distribute smart cards so we can all vote online, but damned if they don't get license plate recognition and automated speed fines working in a matter of weeks after it's legalized.

Re: @Trevor_Pott "They've reversed the polarity on the tachyon inverter"

I'm actually really disappointed noone got the scope of the reference. Yes, it was a TNG joke. The Enterprise-D didn't have a tachyon generator, let alone a tachyon "inverter". (And a reversed polarity inverter should just be a normal generator.) The Enterprise-D consistently had to monkey with the main deflector to produce tachyons in anything other than very small quantities, and having at any point something that "inverts" the operation of the main deflector would be a very bad thing for that ship.

It's worth noting, however, that the Defiant regularly produced both Tachyons and Chronitons without having to involve the main deflector. Presumably the Enterprise's need to send some red shirt to get melted by a plasma leak in the Jefferies tubes every time a tachyonic something or other was required meant that generators for those were added as standard equipment. </giant nerd>

That said, hat tip to Alister for picking up on the "reverse the polarity of the neutron flow" which was indeed front of mind when I started crafting the sentence.

I disagree. You're making a classic mistake thinking that defense is essentially a checkbox-following procedure of "doing the right things". It's not. You need to think messily. You need to think of new and innovative ways to break your own design.

The stuff that shelfware exists to solve is only a small part of the problem. It is nothing more than the beginning. Good defense, like great offense, requires totally orthogonal thinking. And yes, it will and does require implementing things that don't exist as off-the-shelf products. Open source or not.

Attacking is easier than defending. It is easy to train attackers. It is not so easy to find attackers who have flipped over and trained to defend.

A good defender needs experience attacking. A good attacker does not need experience defending, though they could do with studying the shelfware.

Re: Bah!

That may be the nicest thing anyone has ever said to me, thanks!

Re: A question on hashing

Hi Steve,

Hash collisions are a pain. These can be avoided in any number of ways but usually boil down to using a hash space so large the chances are exceptionally unlikely you'll ever encounter one. (See here: http://www.backupcentral.com/mr-backup-blog-mainmenu-47/13-mr-backup-blog/145-de-dupe-hash-collisions.html) Of course, this doesn't eliminate hash collisions which is why a almost nobody uses simple hashing anymore.

Each company has their own take, but a scary number of deduplication approaches really are just "hashing with some kludges to solve minor problems" (like collisions), where "solving" doesn't mean "eliminating entirely".

Actually coming up with really novel ways to do data efficiency didn't seem to become "a thing" until the latest round of the storage wars started about 5 years ago. For many government and enterprise buyers this can be a problem. They only buy from the legacy vendors, unless the new ones go through years of proving out. This can - and does - leave some ops teams in a bit of a pickle trying to deliver the full range of modern functionality. (Or, at least, trying to delivery it at prices that compete with public cloud computing, which isn't as feature rich or as secure, but it is the reality of what they are compared to. Damned if you do, damned if you don't.)

That being said, kludges aside, things like hash collisions can happen. So don't take your 10 year old SAN with it's primitive deduplication features and try to shove 1PB worth of data into it by "just adding some shelves". The bigger the space the more likely you're going to run into problems with implementations.

Mostly, if you refresh every 5 years you shouldn't have to worry. Also, if you use one of the new technologies that isn't hash reliant, you shouldn't run into these issues.

At 100TB you should be fine. Even at 500TB you probably run no risks worth mentioning. But if you are planning at any point to have a storage space larger than 1PB using one of today's storage technologies then demand to speak to a nerd and start asking exactly these sorts of questions of them. It isn't that hash collisions themselves are a likely risk, but you start getting into the scale where implementations don't quite match the theory and it's worth asking the questions.

And don't just look at what you plan to load your storage system with today. Think about the total life cycle of the system. 5 years in production and 5 years as an archival unit. Will you expand it to 1PB in that time? If so, be relentless in your inquisition. If you don't get the warm fuzzies from the answers you are given, move on.

There are lots of storage companies to choose from.

Re: Probably a necessity

@Doctor Syntax

"I think I can see your problem. You're not really using vi, you're using vim which stands for "vi improved". And all those improvements allow distros to play silly buggers with config. Old vi has been available since V7, which included the code for ed became open source. Old vi is confusingly called new vi, nvi. Use that instead."

Bingo. Except some distros map vi to vim and others don't. Some put vi under a new name, others don't. There's no consistency, and it requires learning who's done what that's different from the next guy. Which was sort of my point.

Also: re Red Hat 6.x still being "real Linux"...I agree. Unfortunately, support won't last forever, and nobody has taken up forking it and continuing on with a great distribution.

IMHO, Red Hat's arrogance has killed Linux. It's time now for all of us to learn BSD, for better or worse.

Re: Probably a necessity

@Uncle Slacky

I've been getting behind Slackware more and more recently, but honestly having trouble with BSD. BSD is just different enough that it's actually quite frustrating to use. It looks like Linux. It feels like Linux. But every time I go to use my decades of Red Hat muscle memory something doesn't quite work right or some command isn't there. BSD is something I learn in small doses in between a lot of cursing.

Re: Probably a necessity

I've bunged my protest money in their direction. I haven't seen anything like a practical product yet, however.

Re: Probably a necessity

Not really. Red Hat has changed all this. Consider for a moment that minimal install of RHEL 7 doesn't even come with ifconfig! Starting and stopping services is completely different (fuck you sideways, with a rusty tractor, systemd!) and the names of many basic services have changed.

It's not enough to know some basic POSIX stuff. A distribution of Linux is way more than just basic POSIX stuff. It's names, file locations, configurations of basic software and more.

Editing a text file in Red Hat-based distributions is simple. You use vi. And all you really need to know is "i" for insert, ":q!" to quit if you screwed up, ":wq" to write and quit and you're good. There are, of course, other useful commands in vi, but you can administer a system with just those.

Move over to Ubuntu and...wait...vi doesn't behave the same. it's configured differently and doesn't respond to commands the same. Well, shit.

Now, we can go on and on about the whys and wherefores and even the how these things are made to occur. It doesn't matter. What matters is that the differences between Linux distribution versions can be (and often are) as big as the differences between any two Windows versions. What's more, different Linux distributions have drifted so far apart that they now represent a bigger real-world gap in administration than did the NT line to the 9.x line back in the day.

Linux isn't unified. And what you know in one doesn't quite port to the other. That's bollocks.

If II know Windows really well I can probably pick up $Linux_Distribution_1 as easily as if I knew $Linux_Distribution_2 really well. It has nothing to do with which OS you learn first. It has everything to do with understanding the fundamentals of how and operating system works, and why it works.

But knowing the fundamentals won't stop you from having to study, and study hard. Because the UI - text based or GUI, it doesn't matter - changes all the fucking time in Linux. Just like it does in Windows.

The difference in Linux is that you have choice. If you want to change the UI and set it up so that it behave as you expect and how it is burned into your muscle memory you can.

Or could. You know, before Systemd.

Fuck systemd.

Re: "News"

@Big John

Now you're just being purposefully obtuse.

Hillary is referred to as the Democrats' Dick Cheney because they are both complete sociopaths, war mongers and believers in rule by fear. They also both really like the idea of a police state.

Re: It's just like a bad French remake of the US 2001 bullshit

@Big John: you're an idiot. There is essentially unlimited evidence that guns in the hands of the proles doesn't prevent crime. In fact, in the few batshit crazy places that practice this crime is not only higher, it's more lethal. It's more lethal because not only does more violent crime occur, when it does the criminals do more damage.

Not only that but on the very rare occasion where civvies packing heat pull out their pieces to threaten the criminals with guns they almost always cause additional civilian casualties.

A gun in your pocket makes you feel powerful, nothing more. It doesn't make you able to stop crime. Training and some actual fucking courage do that.

You want a great example of how to stop crime? Look to those trained folks who stopped that attack on on a train in France a few months ago. They handled a touch situation unarmed. Because they had training and courage.

Guns don't give you courage. Not being a terrified pissant hiding behind a need for a feeling of metallic power gives you courage.

Re: They did their dash with me

"Easy test. Go watch about two dozen videos on YouTube that are out of your normal viewing habits, like "epic plane crashes" or somesuch. Then delete your viewing history. Then note that YouTube is making suggestions for you based upon what you just watched. How's it doing that if you supposedly deleted your viewing history?"

It's associated with your IP address. Google builds two simultaneous profiles: one attached to your user and one to your IP. This can be tested simply: log out of your user and see if the suggestions are the same. For me, they aren't. I start seeing stuff better tailored to my wife, or to the business account that I use for research.

Re: They did their dash with me

I see quite a lot of duplicitous language, such as the oft-used myth that "computers do it, so it's all OK" - oh really? And who programs those computers?

Who cares? I've programmed lots of things that crunch data I don't actually see. In fact, I've programmed things that crunch sensitive data, extract just enough information to establish trends then jettison the original sensitive data.

Why should I assume there is a nerd manually examining my data? I don't even assume Microsoft does that. (Though I do assume they give all my data to the NSA, since they're so chummy.) What I do assume is that Microsoft has algorithms that hunt through everything looking for information that is of commercial competitive value. Microsoft conducts industrial espionage, of this I am 100% certain. I have been given no reason whatsoever to believe that Google does.

Also let's not forget it was Google who was fined $22.5M by the FTC in 2012 for explicitly developing and deploying a method by which it could track Safari users despite the tracking prevention being enabled

It was Google who got CAUGHT doing this. I am 100% positive that Microsoft have done the same thing, and more. They are every bit as keen on tracking everyone all the time as Google, Facebook or Apple.

"The other thing they are not upfront about is your ability to opt out of Google Analytics. You know what their advise is? You should INSTALL something to opt out - why not simply follow the Do Not Track flag that most browsers now have, or would that make it too easy for people to drag Google again into court if it did regardless?"

This is one of the few valid complaints that I agree with. From Google's point of view DNT is about *advertisers* tracking you, so they see Analytics as a separate thing. (And frankly, I'm more than happy to ally website operators to see some basic data about me, but don't want advertisers to track me. That's just me, however.) I don't think that it should require a cookie to be placed on your system. That said, we don't currently have another mechanism to differentiate between the two types of information. That isn't an excuse - Google could be leading the charge to create one - but I can at least sort of understand their point of view, even as I disagree with it. I emphatically don't think it's malicious.

That's the thing, I guess. Google's entire income is based on building a profile of you and selling advertisers the ability to put an ad in front of the right audience. They spend a lot of time trying to walk a tightrope between what is okay to do to build that profile and what is not. Google is also huge. I can believe that some groups go overboard while others would not approve of the same actions.

Microsoft, on the other hand, seems to make decidedly hostile choices which are all the more galling because tracking you is such a small fraction of their income. Microsoft violates your privacy and compromises your data sovereignty because they CAN. Not because they have to.

With Google, we know what we're getting. They're selling us nothing so we're the product. With Microsoft, we're paying spectacular amounts and they're STILL screwing us over, privacy-wise. It's not okay.

Google are not the good guys. I acknowledge this.

But Microsoft are worse. They lie repeatedly about how very much they are the good guys, then behave just as badly - frequently worse - than those they deride. Give me someone overtly evil over a well trained conman any day. At least I know where I stand with the openly evil fellow.

Re: They did their dash with me

@AC RE: Google TOS

Good catch.

Re: "The main issue I have with Google is that NONE of their actions have inspired any trust."

But they haven't really done anything to break it either. They've been pretty up front about things. They've made some mistakes, but there has been no protracted campaign of lying or denial that I know about. (Except possibly the WiFi snooping thing.) They've been pretty up front that they slurp data and use it to advertise at you.

Microsoft pretty much seem to lie about everything by default, then backpeddle slowly until they reach a point where the screaming has died down to a dull roar. That's what politicians do when trying to pass bills people don't like. I don't want that from my software company.

I mean, okay, I don't really want Google-style building a psychological profile of me either, but the choices seem to be "a company that will spy on you and charge you a lot to do so", "a company that will spy on you and charge you nothing to do so" and "a company that will spy on you while denying they do so and then blame everything on you when it's proven that they are lying."

So I guess I'll go with the one that doesn't charge me to spy on me? Least shit option of a bunch of really not very appealing options.

Re: They did their dash with me

*shrug*

Okay, Google are horrible at collecting data. That's a fair point. But...

1) They aren't collecting the data at the operating system level in a means that can't be turned off and firing it back to the mothership. (ASOP does collect data at the OS level, but does actually honour privacy commands, and in the latest versions doesn't come all creepified out of the box.)

2) They have clear, simple controls for your privacy all in one place (though admittedly online) and the ability to nuke out everything they ever stored on you.

Caveat A) That is associated with your Google account.

Caveat B) You do have to trust they are actually deleting it...though they haven't really done anything yet to make me think they don't. Eventually. I am curious as to the time before the core data (and not just the association with the identifier) is deleted.

Caveat C) Google's creepy snooping is pretty well understood so there ate lots of browser plug-ins to stop them.

3) Google probably aren't "selling your data" to advertisers. That's actually silly, from an economic point of view. They are probably selling advertisers access to individuals who meet certain criteria. This gives Google the control. The advertisers have to constantly go to Google in order to get their data in front of your eyeballs.

It doesn't make sense to think that advertisers get a great big spreadsheet that says "Trevor Pott, Age: Feck Off, Likes: Fishkeeping, Gardening, Global domination", etc. Simply handing that data over to the advertisers means they don't need Google any more.

Depending on the details of this, and on Google's restrictions and controls about making what amounts to my complete psychological profile available to governments, my competitors, enemies, exes, leprechauns and so forth, I am mostly fine with that. (I realize not everyone is.)

Advertisers want to put an ad in front of me and Google is going to (mostly anonymously, at least as far as anyone other than Google is concerned) try to make sure the best ad gets in front of me? Okay. Since I choose when and where I view ads (thanks, Adblocker, Ghostery, etc), I am okay with that. Mostly. There are niggles related to Android.

Where I get squiggly is things like "invading my e-mail because you think a source of mine might have snuck me insider information" (see: Microsoft) or handing my info over to the spooks. Those are things I don't like. Here, I am more comfortable with Google's handling of these issues than I am Microsoft's. Apple is somewhere in the middle. Apple seem to have good intentions, but shitty execution.

So...I don't know, is Google really the Big Bad All Seeing Evil? It really depends on two things:

1) Exactly who can see how much of the psychological profile of you that they construct. If it's basically just bots and scripts, we're cool until the singularity.

2) When they say they delete it, do they actually delete it?

I wish Google would provide answers. Or Microsoft. Or Apple. Or anyone. And answers we can trust. Maybe from a non-US subsidiary. Because national security letters could let HQ lie.

Re: "why not try for devotion?"

And here, ladies and gentlemen, is devotion. The willingness to honestly and earnestly believe that trust doesn't matter. The ability to convince one's self that anything which gets in the way of belief is simply irrelevant.

Witness it in it's glory. Simply believing that such a thing as an "acceptable privacy cost" exists is the first step. But to sidestep issues of trust around end user control over their own devices, applications and operating systems as irrelevant? That's devotion.

Re: "why not try for devotion?"

Maybe we aren't. Everyone has different things that matter to them. Microsoft is catching up in some areas (containers, for example) but is at par in most places and completely wrecking the competition in others (hybrid on premises/hosted/public self service virtualization a.k.a "hybrid cloud").

Let me be perfectly clear here: while Microsoft may be behind on individual features with individual products very few companies have the hybrid infrastructure, automation and orchestration capabilities of Microsoft. Fewer still make those capabilities as easy to use.

Openstack is possibly more comprehensive than Microsoft's stack, but is harder to use. (Which, given how horrible System Center is, is saying something.) ZeroStack is a startup that removes a lot of the installation and configuration headaches for Openstack and provides a decent user experience, but they do so through a SaaS management plane riddled with data soverignty concerns that they don't seem to care to acknowledge, let alone address. There are a handful of other Openstack offerings that might be useful one day, but aren't quite there yet.

Yottabyte has a hybrid cloud that isn't as feature rich as Microsoft's yet (they're missing the app store) but is otherwise fucking spectacular to use. Sadly, they're too small to be a real threat quite yet.

VMware has all the pieces of the puzzle, but they cost a mint, are crazy hard to install and even harder still to use in concert. (How the hell do you make something more miserable than System Center, I ask you?) VMware does the basic virtualization quite well. They even do basic hybrid cloud in an almost usable way. But they are completely worthless at automation and orchestration. Which is why they only have a handful of customers for those software layers.

Cisco has a complete-ish offering based on Openstack, but they are still putting the pieces together into something cohesive. They'll be a threat in a few months, if they don't flub it.

IBM's Softlayer can be pretty awesome, but nobody can afford it.

LOL, HP.

Beyond this I could start putting together stacks of software and they'd amount to what Microsoft has, but they'd be almost a dozen vendors. Almost all of which are startups that may or may not be there in two years.

So, for my money, Microsoft have some great technology that is better than most of what is on offer elsewhere. At least on the infrastructure side. Unfortunately, they're liars and scoundrels. Even more unfortunately, they're terrible at UIs and don't even know it. They measure themselves only against other large software vendors, all of whom are also terrible at UIs. But there are a whole bunch of startups out there who are great at UIs, and they have some compelling offerings on the table.

The next few years will be interesting, but for Microsoft I think it will mostly be with them serving as an example of how to get everything so close to right but not quite get the cigar.

Re: "why not try for devotion?"

I don't know, from an end user side, you're probably correct. From an administrative side, however...their vision of hybrid cloud computing is actually quite compelling. Many of the technologies they've developed are quite good and they have provided perhaps the most complete and relevant offering in the space.

The problem is that their execution is entirely one sided and utterly lacking in trust. To be blunt: they got all the hard stuff (the technology) right and all the easy stuff (privacy, data sovereignty) wrong. They say one thing out one side of their mouth but do something different in practice. They break the faith as a matter of course.

For me, at least, that's what's so damned frustrating. I can make a convincing argument for why Microsoft have the best vision for the future of tech, the best technology on the table today and the best roadmap to get from today to a pretty kickass future. Unfortunately, taking advantage of this neato stuff means trusting absolutely in a company that has proven they aren't worthy of it and have zero interest in being worthy of it in the future. That's a huge problem.

I could see myself being devoted to the Redmondian vision. It's a really cool vision. In fact, it's so cool that I find myself wanting to believe in it, despite everything Microsoft has done. When you really learn how all the bits go together it's amazingly compelling.

But there are big huge problems in execution and in trust that shouldn't have to be there. They are there because Microsoft made a choice, and that choice was to pretend that privacy and data sovereignty simply didn't matter. Even users having control over their own systems didn't matter. (Let along users being able to license their systems in perpetuity, which Microsoft has become very against of late!)

Microsoft have a narrow vision of the world based on echo chamber use cases. They view the world only through the lens of USian large enterprises who are immune to data sovereignty concerns, can easily obtain insurance or legal indemnity against privacy concerns and are completely immune to boom-and-bust economics.

On the one hand, USian large enterprises are a huge revenue source. On the other hand, the majority of businesses in the world are SMBs and the majority of those are located in (and subject to) boom and bust economies. Microsoft made it's choice. It chose a niche to cater to and that niche isn't "the majority".

That is enough to keep a large company going. It isn't enough to see growth for that company. To see growth, they need to start at least paying lip service to those niches they wrote off as irrelevant years ago.

The question is: will the denizens of those niches fall in line? Or will they say "hold, enough"?

The vision is compelling. Unfortunately, it might also be fatal for those who choose to believe in it.

For the same reasons that through the ages various religions have mouthed pieties whilst engaging in the unspeakable: they want not only money, but power and, through the exercise of that power, control.

Unlike governments, Microsoft has been unable to obtain power through fear. It thus needs to obtain power through devotion. So it will say whatever needs be said to obtain that devotion, but do not under any circumstances expect it to practice what it preaches.

Both fear and devotion have a rich history of causing people to forget the past. Since Microsoft has failed at using fear, why not try for devotion? It has worked for other tech companies.

Re: Using toys as tools...

@Vic

Being a pilot doesn't seem to have gifted you with reading comprehension. Pity.

That said, I wish I was a "bullshit artist". I've met a few. Those who are good enough to properly earn the title "artist" seem to be quite wealthy. I think I'd rather like being that wealthy.

Unfortunately, you see, I'm really not all that clever. I learned a long time ago that telling lies means you have to track those lies. It makes life terribly complicated. With my ADD that's a really bad plan. So I tend to stick to the truth as much as possible. For varying values of truth, of course. Human memory, for example, is highly fallible. See: eyewitness testimony. But as a general rule, I tell the truth as I see it. Makes life easier, albeit rather less affluent.

I am happy to leave my veracity up to individuals capable of actual reading comprehension. They'll be perfectly able to see that while I may have been wrong about some details of how iPads are current used. (I assumed they were used for slightly more tasks than they currently are, based on how I have seen laptops used in various situations in aircraft before.) I did state my assumptions in various posts, along with the history behind the rationale for those assumptions, but at no point did I lie about anything.

Within the context of those assumptions - that the devices perform tasks which are technically non critical to flying, because backup procedures or skills should theoretically exist to compensate - I stand by my original assessment of why the Surface is not a good choice. Mostly because even if device is filling a technically non-critical role, lazy or overworked pilots absolutely can become dependent on that device anyways. It behooves the IT nerds then to presume criticality and pick the least worst option available.

Unfortunately for me, bullshit artistry isn't something that is easily learned. I don't think I'll be a millionaire on the back of my ability to bamboozle. Fortunately, reading comprehension is something that can be taught fairly easily, so there's hope for one of us.

Re: Using toys as tools... @Pott

@BillG

Where am I evidencing brand tribalism, hmm? "iPad is the least worst amongst a bunch of options that are kind of all crap" is a hell of a long way from "Surface is trustworthy because Microsoft says it is and I trust them despite all the times they've broken their word".

Re: Using toys as tools...

"If the 'electricity is not working' it will be fixed before the plane is allowed to take off."

And if it fails in flight? That's the thing you need to plan for. You don't need to plan for everything going to plan. You need to plan for it going pear shaped in the least convenient way at the least convenient time.

Please, never design a critical system for anything, ever.

Re: Using toys as tools...

"For crying out loud, Trevor, cut out the bullshit. Earlier, you thought pilots were flying planes through their iPads."

No I didn't. You're full of shit, Vic.

"What you are suggesting is nonsensical. Aircraft are not designed to have laptops plugged into them[1]. Even if you can find a diagnostic port, you're not going to get to it without stripping out quite a bit of the interior, and even then you're not going to know how to control anything through it."

Not something like an ODBII port, no. But depending on the plane there absolutely are external sensors and diagnostics that will get plugged into laptops.

A plane's core bits that are required to fly the thing will always be built into the plane itself. Usually with redundancies. I never argued otherwise, and in fact have argued a few times here that this is how things are properly done.

That said, new things are added over time to make the lives of pilots easier, or to add non-critical features. These can - and do - get piped through things like the aforementioned laptop. Let me give you some practical examples.

1) On a smallish aircraft that did runs up to the diamond mines the plane was refitted to carry some fairly nasty chemicals. Sensors were added to detect these chemicals and the computer that detected issues buried alongside the other computers. This computer had no primary display (other than a warning light) and no input directly. Detailed information was accessed and filtered entirely through a laptop. This was the same laptop that held mapping information, handled weather reports (which, incidentally, was actually handled by another computer buried in the plane, and merely displayed and filtered by the laptop) and also was used to receive text alerts from the local mining camps when the plane was in range. (I have no idea what the alerts were for, or how the plane received them.)

2) I remember a laptop being used as the interface to a new (IIRC it was a prototype and still in testing) security system for a plane. There were all sorts of sensors in the cargo hold that were beyond what was originally fitted. The thing also (for reasons I do not understand) was tied into bathroom usage. Again: the computer vacuuming up the data wasn't the laptop itself. It was merely how you got at the data in question.

" Do you really imagine we'd put anything into that to allow remote control from a laptop?"

No. I don't. I never did. In fact, I expect you'd use an analogue computer or a really hardened real time embedded digital computer. I can think of no modern commercial OS or system I'd trust to run avionics.

So get bent, mate. At no point did I say anyone flew a plane using a laptop. I said they were used as the primary interface to systems that were typically added after the fact. By this very definition that can't be critical to flying the plane because the bits required to fly the plane are the things that there will be controls for.

Adding more computers to a plane doesn't make those computers essential to flying the plane. You don't need a computer to fly most planes. (Well, not a digital computer, in any case. The need for analogue ones in some circumstances has been well proven.) Those planes that do need computers (analogue or digital) absolutely have every control you need to access and interface with those systems available without third party interface required.

That doesn't mean those third party systems which are used only to access "extra data" don't eventually become mission critical, in spite of themselves. Might I also suggest that as someone who flies light aircraft as a hobbyist you're more likely to do things the old fashioned way and take pride in maintaining the pen-and-paper skills?

A commercial pilot - especially ones who are on the "work is a pain in the ass" side of the spectrum might come to rely on the theoretically non-critical third party tools too much making them more mission critical than originally designed. Which is something that anyone doing IT design needs to account for.

Re: Using toys as tools...

"No you haven't.

You have crossed over into fantasy-land. Please return to topics you know something about."

Yes I have. Back in the day the laptops themselves weren't powerful enough to do much, and there were systems on the aircraft which gathered (non critical) data. The laptops were the control interface. Now that interface was mostly "select between information sources" or "filter the data to deliver me this information", but that's what they did.

And why would that even be a bad thing? It's just burying the extra oomph in a wall somewhere. Now, that was long enough ago that it wouldn't apply today. Your average phone has more than enough heft to do all that and more besides. It can listen to a dozen satellites, broadcasts and what-have-you and parse the data for display without breaking a sweat.

Wasn't always the case though. Mind you, back then, it was more of a novelty, and people still used paper books, manuals and charts a lot more.

Re: Using toys as tools...

" It is not like pilots randomly walk into unknown planes with random devices and get surprised by electricity not working."

Sure it is. Faults develop in airplanes just like any other workplace. if you don't anticipate these sorts of issues on an airplane, people can die. That's why you must have contingencies for absolutely everything. Clearly you've never worked in such an environment. Too bad, there's a lot to learn about proper IT architecture by having to design for failure.

Re: Using toys as tools...

"Just one concern, I had two iPads lock up, on the same day. Access to all applications ceased.

I later worked out that it was 365 days since they had been connected to iTunes.

This alone makes them unsuitable for mission critical applications."

Now that is interesting. I haven't heard of that one...though I sadly do believe it is entirely possible. Worth investigating,a t the very least. You're right, it might make iPads pretty bad for any situation where they weren't regularly returned to base for service. Of course, if simply rebooting them gets them back to what they need to be, and they don't lock up again, that might be an acceptable glitch.

A one-time characterized event that is easily recoverable from in a short timeframe isn't the end of the world. It's the unpredictable ones over which you have no control that are the real problems.

Re: Using toys as tools...

@ James Micallef

Well, based on some input from actual pilots for what these devices get used for, I understand better which systems they are augmenting and replacing. The short version is: no, they aren't replacing systems that are critical to flying.

The long version is that despite them technically not being "mission critical", they in reality are mission critical. Many of the things that pilots have for decades done manually are being done on this iPad. Especially things that required a lot of maths.

Every pilot - even the non-commercial ones - is trained to do this stuff by hand. The problem is that if you do this with the assistance of a computer often enough, the "do it by hand" is no longer second nature. And that becomes an issue, especially in stressful situations.

In the past, this sort of thing was done by a (typically not very reliable) laptop awkwardly bolted into whatever space it would go in the cabin. It was never a good fit. And the computers that this laptop served as an interface to were not the sorts of things that controlled the plane. They provided information (maps, positioning, even weather updates, calculated course corrections,) to the pilots. I can absolutely see how an iPad is eleventeen times less awkward.

The iPad isn't used to control the plane directly. Despite this, I can see how it easily becomes a very heavily relied-upon tool.

So that a grey area. Properly trained pilots who maintain their skills despite having more convenient tools available to deal with the miserable stuff won't have problems if the iPad goes kablooie. That said, we live in a world where pilots aren't paid enough and work stupid hours.

I argue neither the iPad nor the Surface are adequate for the task. Given the task to hand, however, I believe the iPad to be the least worst choice and thus the logical choice. The Surface may in fact be the worst choice (though, TBH, Android probably is the worst choice), as it is most likely to go completely fucked from an official mechanism (not an application going bonkers.)