Because something system-level with access to run the file (presumably) has to apply regexp's to almost it's entire contents to correctly determine the type of it (e.g. is it a ZIP or is it a JAR with the same compression?).
It'd be trivial to generate a DOS of some kind by just feeding it a malformed file.
Which involves some system-level process opening every zip-compressed file, peeking into its contents and performing the equivalent of "file" regexp searches throughout its content to tell you that it is, in fact, an OOXML or JAR file rather than just a ZIP of the same.
The processing overhead and security requirements of the basic OS file-managing processes just went up an order of magnitude.
Tell them to use a half-decent browser that won't let you run executable-named files from the browser anyway. Or at least put up horrible click-through warnings when you do.
If the extension determines something that the user should be aware of (what program will open it), then it needs to be shown, and modifiable to correct mis-identified files.
If the extension is merely a nice name, then the user need not care about seeing it or knowing it and should be able to change it at will.
Sadly, in Windows, the extension determines what program is associated and therefore opens a file (and, in part, whether it's executable!). It's encoding a file attribute into the name itself, which is a "layering" violation as far as I'm concerned.
Ideally, the mimetype would be associated with the file metadata and be as easily changeable as the name (e.g. drop-down box to "change" the file type it's interpreted as whatever the extension), but that's a nightmare that no sensible OS seems to have tackled. It wouldn't be that hard to "default" the mimetype to whatever the "file" utility detects it at on introduction of a new file, but nobody seems to want to do that.
Similarly,the "dot-hidden" files of Linux suffer the same fate. Just the name determines a kind of file attribute, which is silly and potentially dangerous or confusing ("I renamed it.jpg and it disappeared!").
While we're still using OS that trust and encode the filetype into the filename itself, we need the facility to change the detected filetype, and the facility to view the detected filetype. Hiding extensions only achieves the later (with a separate file type column, "Adobe Reader Document", etc.). Any cleverness with showing the extension on rename where it wasn't showing before will mean the user will strip out the file attribute when they rename over the top, and not showing the extension makes it more difficult to change what it opens in.
We created this problem back in the 8.3 filename days and NEVER properly solved it. Encoding metadata into the filename is the cause. Removing it is the solution. But without the infrastructure for that in place in every app, it's pointless to attempt it on a commercial world-wide OS.
I usually compare my computer desktop to my physical desktop.
In this instance, I think it's quite close to compare my smartphone to my pocket.
Sure, if I was walking along and someone handed me a free album, I'd take it. I'd probably put it in my pocket if there was room. But I would also reserve the right to say "No thanks".
What Apple/U2 did is tantamount to chasing you down the street trying to shove their product into your pocket. Sorry, you have no right to do that and it's damn rude, free or not.
People don't have free downloads everywhere, and those that do may not want this crap - whether through playlist OCD of just not wanting to see something they have no interest in or, more likely, BECAUSE it was shoved into their pockets without any option.
If free stuff's so good, I'll ship you a bunch of "free" cardboard box to your house and fill it up (removing it is another issue entirely if we follow the Apple/U2 thing too closely!). Hey, it's free! Why don't you want it? You might like it. You might be a cardboard fanatic. If I offered you a bunch of free cardboard at some points in the year you might tear my arms off to get at it (e.g. moving house). If you'd asked for it, it'd be great. But instead some guy showed up, shoved a bunch of cardboard through your door and said "Hey, it's free, don't complain!".
In a zombie apocalypse, all movies and games depict their being some kind of "safe-house" where all the uninfected will gather together.
This, it seems to me, is the most stupid idea ever. Heading towards that is certain death. The zombies only need to be able to read, or get lucky, for it to be game over - they know where the highest concentration of juicy victims will be. And that concentration will increase as everyone piles to go there.
Sure, that's where the firepower might be concentrated too, but that's not good enough when one breaking through is enough to create a game-over scenario.
It seems to me that when everyone is piling towards the safehouse, a prudent course of action would be to stay still / run the other way, unless there's something that stops you doing so.
If I was a zombie, I know exactly where I'd head first.
Similarly, when everyone tries to flee a city, the first thing they do is jam the motorways (freeways). It seems quite stupid to even try if there are really that many people headed that way. All you'll do is trap your vehicle in a queue that you can't escape and then have to get out on foot.
Go the other way, go down the backstreets and side-alleys, stay off the main paths is surely the intelligent thing to do (besides using a bike or other fast transport in the first place).
Claiming, or falling for, any argument that "open-source is secure" is a complete failure to understand. Security is relative, not absolute. To get this ass-backwards just makes you look like an idiot. Believing anyone who says ANYTHING "is secure" is utter stupidity (rather than "is more secure", for instance)
It's like saying "metal's secure". No it's not. I can walk around a sheet of metal just as easily as a pane of glass. However, a metal lock built to the same design as, say, a glass one is likely to be MORE secure.
As such, open-source is not "secure". It's considered to be "more" secure if everything else is equal.
Tell me, how many Windows-based Blu-Ray players can you buy in your local supermarket? Zero? Shocking. Why? Because Windows is not properly designed for embedded use at all really, certainly not until very recently. There were ATM's that run on XP Embedded, and there were a few cars that had Windows-based control systems (I remember a story of some ambassador being locked in a car by it).
As such, a complete install of Windows would be vast overkill and a huge attack surface for such things because it's just not ready for that. So it's not a fair comparison at all, as Windows is inherently "less secure" in such circumstances as it's just not designed to do that. However, Linux / open-source Blu-Ray players are "more" secure than many of the alternatives still, everything else being equal.
And why - because not only can you see if your player would be affected by the bugs but, by the same licence that grants you the ability to see the bug, you can CHANGE THE CODE if you so want. No waiting for vendor updates. One person in the world who knows how to code can look, see the problem, fix it, publicise it and therefore purge the world's devices of it.
Tell me, how many "security patches" do you think other commercial Blu-Ray players ever get given? Even the ones connected to the net 24/7 for their "extra media" functionlaity. There's even a facility to update Blu-Ray firmwares via buying new disks (not unlike the Wii games that bundle an update of the underlying OS before they'll let you play them). It's rarely used.
If you're going to pick arguments, have some vague understanding of what the real argument is, not some child on the Internet's poor re-stating of it.
Open-source is potentially MORE secure than closed-source, everything else being equal.
Those conditions and caveats make a HUGE difference to the intent, meaning, and truth of the statement.
Guess what, AES-128 isn't "secure" either. But it may be "more" secure than other algorithms, for example.
Well, when they're only getting 20-30 comments on most stories now, of course they have to remove quite how few they get on each article from the front page listing too.
Someone had to tell me - It's there if you hover over the bar, but for some reason the "Change" button (which I've never used) is as grey and obvious as it ever was.
Damn, that's stupidly impossible to see on a whole range of monitors that I have here. I've complained to support, but I doubt they'll do anything.
It's like the Metro Start hover all over again - you have to play some kind of pixel-hunting adventure game the first time you do anything to work out where to go next.
I was LOOKING for it and couldn't find it. And why would you ever want the button to be the same colour as the bar it's in? It's there for a reason - to be pressed. Don't hide it from me.
No, I have Javascript enabled, have latest Chrome. It's just borked, and it only happened today (no updates to software between yesterday and today - same browser session still running in fact!) but the site now doesn't render at all nicely and it LOOKS deliberate, but I'm missing any kind of Post button at all.
I only use Windows because it was "for free" because of my employer buying me a laptop.
But for five years, I managed and supported a 90% Windows network with hundreds of devices primarily using a laptop which had LibreOffice, etc. installed.
OS - sorted.
Office suite - sorted (sorry, but it is. I used to get people envy my LibreOffice setup, as I could do everything they could do, and manage their same files they managed, and also do things like open ancient foreign formats that people emailled us still).
General apps - sorted.
Games - 1/3rd of my Steam account "just works" on Linux.
For years, I didn't have Windows or Office, as an IT professional supporting users on Windows and Office. Sure, it would have been nice to have a native tool occasionally, but for the odd things I needed (e.g. AD admin tools) it was always safer to just remote-desktop into a Windows machine, or use VM's (Samba tools just aren't there yet).
For everyday use, personal and business, I used Linux as the base OS and for the vast majority of tasks. Only when I was doing something very Windows-specific did I have to load up a Windows tool and always did it from a Linux machine.
If something had to be doing the maximum possible for its entire existence to get that far, chances are that's not the maximum possible, or we're measuring something wrong.
For that cost, for the price of a couple of drones you could put another couple of officers, stationed permanently to do just their job. And thereby free up whatever officers would also, presumably, need to be present to enable the original drones to operate too.
Simplify the choice - one drone, or two officers (maybe an officer and a half) on the ground doing the same job and NOTHING else - and the value motive really comes to the fore.
Law enforcement isn't about what it costs. Hence why the UK police are still sitting outside the Ecuadorian embassy at the moment waiting for Assange to move out of his personal prison to go to an official one. But it's about value. One high-profile "celebrity" openly-flouting the law is enough to encourage a whole spate of lawlessness in following suit, and you'll have every shoplifter and petty criminal claiming asylum in embassies to evade the law within a month.
So the cost motive would mean we'd leave him in there and forget about him because "he's too expensive to care about". And also that we wouldn't bother to deport illegals.
The value motive says we stay there to dissuade this kind of activity in future and make sure it doesn't cost more in the long run. And that illegals are deported at huge costs to prevent being seen as "weak".
Funny how changing one word (cost -> value) can change the whole intention of your post, isn't it?
We have gone back to a desktop akin to Program Manager (large square blocks of programs on a plain background).
We have gone passed through Active Desktop, only to have it ripped to shreds and reincarnate in the form of the (still loathed) Metro.
We have gone back pre-Start Menus and removed taskbars (in Metro at least), albeit with added features. Then we've put them back in.
The people who work in IT know that these fads come round in cycles and eventually you'll go back to how you were because some things work and some don't. At one point, Windows 95 was going to be "the last Windows OS with a CLI", now we've brought in PowerShell, etc.
And all because, at some point, Microsoft have this ego trip about never being wrong and knowing better. And all the IT guys want is an option. Do I want shiny new stuff, or boring old stuff? Let me choose. It's really that simple. You can put EVERYTHING you want into an OS. Just give me a switch (and ideally a Group Policy setting) to turn it back off if it's not what I want.
I don't want Metro to be thrown in the bin, I want my users to be able to choose whether or not they want to use it. And that means giving them years of both that they can choose between at a click. It takes THIRD-PARTY FREEWARE and faffing about to do that, which isn't the best solution when I've paid for the OS for all my users on an annual subscription.
Nobody cares whether you want new-and-fancy stuff or not. But inevitably, it'll go out of fashion, then come back, then go out. All I want is the option. It's not hard. If a freeware project can give you the option, why can't Windows?
(P.S. Why can't I make Group Policy use AD stored user photos as the Windows 8 login? It's almost impossible with lots of login scripts and downloading an image via third-party software into a particular location on EVERY machine that they might even log onto. Why can't I set a myriad of basic options that are there on the screen for my users, and why can't I *stop* them changing them just as easily? Why can't I just have "skip Metro screen" as an option in Windows to go straight to an old fashioned desktop? Really... it's not that hard).
Scaling is a one-off for the particular size of screen.
There's almost zero overhead in keeping the icons as SVG or similar, and rendering to a bitmap in the device dpi that you require, and then using that bitmap until the screen resolution changes.
Cache enough of them and after the first few resolution changes, you'll never have to render the SVG on that machine again.
So the "scalable has a cost" rules go out of the window, really. And even back in the 90's, did you ever see the stuff you could do with vector formats? Some of them were amazing even if they took 10 or so seconds to render on the machines of the day. Nowadays? Pah. Make it render in the background the first time you change resolution with a "Please Wait" screen showing while you do so. If it takes more than 20-30 seconds for extremely complex icons for the entire Windows control set on a modern machine, I'll be amazed.
There are issues, much like those that plague font sizing, where you have to "hint" at lower resolutions how you want the final bitmap to show (because you might want what looks like an insignificant single pixel to render as a single pixel no matter what the size to make it look right, e.g. cursors, etc.). But those are all solved problems.
Have you seen the speed and zooming size you can get on the MacOS bottom bar? It's fabulous (and I hate Mac with a vengeance). And I've seen it do the same speed on a VM with no 3D acceleration at all.
There's no excuse. This is just designer-pandering where "design" means "look", not functionality, usability, human interface guidelines, intuitiveness, etc.
I've often backed a return to the Windows 3.1 interface for even modern Windows because it was simple, plain, boring and worked, and the icons you could tell roughly what they did. But this is a STEP BACK from Windows 3.1 even. Urk.
And, yes, I have deployed the previews in VM's for on-site testing and, apart from the look-change, there's little of note there. I'd have been infinitely more happy with a 7 or even 8 clone with newer features.
Hell, I've used 8 on a tablet and on touchscreen PC's and each time it's easier to turn that Metro junk off and use it, and even then it's just NOT designed for proper touch usage.
1) Robots cannot replace all jobs. We've yet to make a self-fixing robot of any note. AI is NOWHERE NEAR capable of doing the simplest of paperwork or administration. Hell, we've barely automated anything of the IT departments, let alone anywhere else. All they can automate are mindless, repetitive, labour-intensive (and sometimes dangerous) jobs. Though that puts a LOT of people out of work, that's by far not the majority.
2) If robots do replace all jobs, the "money" comes from sale of goods just the same. Half the workforce are working and doing the work of the other half - the robots produce the goods / services, and the humans lounge at home. In fact, if anything, we'll have more money because we could produce more for less maintenance costs and have no union troubles - once the machine is in, it'll carry on working until the power goes out, effectively, and cost pence to run in the meantime as an "hourly wage".
3) The result of the above is that food and goods become so cheap and plentiful that the concept of "buying" them will seem old hat. If a government could pay for itself by selling goods to other countries still, then it doesn't really matter what you do - you could quite literally be paid to stay at home, if all you have to do to "do" the same amount of work is power up a robot and oil him once a week Mass amounts of automated robots also make self-sufficiency much more possible - imagine that you don't have to farm or buy goods, just let the same robot that works for Kelloggs loose on your land and it'll feed you for minimal cost.
4) This is all a pipe-dream. By the time you have that sort of automation, the only jobs left would be bureaucratic - and they'd realise they're the only ones working. Things would flip on their head.
5) Who cares? If you have no job and no money but food is so cheap that going an oiling a robot once a month pays for everything - wow... perfect life.
To protect the metadata of the recipient is daft. How are intermediary servers ever supposed to know? And if you and the other end of the connection both set up a connection and know who it's for, that blows out the "fact that a message was sent" before you start.
Message length is also stupid to try to hide. Sure, it may not be exact but if I send a 200Mb email and you send 20 characters, how are you supposed to encrypt those to be indistinguishable without literally padding to the nearest 200Mb? And padding might be able to be done on a smaller block basis, but it's exactly that kind of padding that broke open gzipped-and-encrypted data in SSL sessions for a while.
Timestamps? Who needs them - if someone's listening, they know what time you contacted the server and it'll be +/- a couple of minutes of when you wanted the email sent (i.e. immediately, in almost all cases).
Sender? Obscuring that while connecting FROM an email server that's trusted to be honest about only delivering mail for known local users (untrusted ones are spam sources and will get blocked no matter what encryption they use)... there's your metadata.
Metadata is, in and of itself, almost impossible to remove. The fact that you've connected to a target mail server is metadata.that you're probably sending an email to an email account at that server. Anything else is either between you and that first server (so forwarding is a pain, re-enveloping is a pain, bouncebacks are a pain, there's a stupidly high computing cost associated with receiving spam, etc.) or has to be announced somehow.
And that would all come back to entirely encrypted and obfuscated and peer-to-peer networking globally. You're talking about a Tor problem, not an email problem.
If you don't want people to read a message, encrypt it. If you don't want people to ever know a message existed, you have to be able to send it in such a way that nobody could ever know. And that's all-but impossible at the moment. The best we have is hoping that sending it in bits to a random selection of unknown strangers will protect you enough that they could never collaborate against you (Tor).
The problem with it is that I could never be bothered to use it, not because of privacy (it would be incredibly convenient to send, say, a password required in an emergency via a verifiably-encrypted email) - but because it's such a faff. And it interferes with everything (searching, archiving, re-enveloping etc.). And to do so is all bolt-on-and-bodge-job methods. None of the major email clients offered anything like proper encryption by default.
And as soon as you get into using plugins, most people just won't bother. There are plugins for PFS for all your instant messenger programs, etc. - I had one installed for about 5 years, the only other person I know with one installed has a different, incompatible one. Now I don't use IM much any more anyway, so it's dead in the water.
And all email encryption is a ton of messing about with publishing keys in the right places, and having to verify against those places, etc. It's ugly.
The only place I've seen anything like GPG working is in package signing for third-party software. And there you have to download the package, download the key (either from the same website as the package - WOOP WOOP - or independently crowd-source a verified key), and then check it works. I've only ever bothered for Slackware, for which I believe the ISO images are signed with the official Slackware key.
GPG is just a pain in the butt and not automated at all. It's easier to compose and encrypt the message ENTIRELY OFFLINE and then send the encrypted text, and that shows you what kind of automation is missing, and what kind of trust system is actually in place.
Sure, there are plugins, helpers, hacks, extensions and all sorts. But none of them ever progressed to being "in" the software. Not even software designed to do nothing else but send email.
Slashdot Mirror
Because something system-level with access to run the file (presumably) has to apply regexp's to almost it's entire contents to correctly determine the type of it (e.g. is it a ZIP or is it a JAR with the same compression?).
It'd be trivial to generate a DOS of some kind by just feeding it a malformed file.
Which involves some system-level process opening every zip-compressed file, peeking into its contents and performing the equivalent of "file" regexp searches throughout its content to tell you that it is, in fact, an OOXML or JAR file rather than just a ZIP of the same.
The processing overhead and security requirements of the basic OS file-managing processes just went up an order of magnitude.
Tell them to use a half-decent browser that won't let you run executable-named files from the browser anyway. Or at least put up horrible click-through warnings when you do.
Because it's a stupid idea.
If the extension determines something that the user should be aware of (what program will open it), then it needs to be shown, and modifiable to correct mis-identified files.
If the extension is merely a nice name, then the user need not care about seeing it or knowing it and should be able to change it at will.
Sadly, in Windows, the extension determines what program is associated and therefore opens a file (and, in part, whether it's executable!). It's encoding a file attribute into the name itself, which is a "layering" violation as far as I'm concerned.
Ideally, the mimetype would be associated with the file metadata and be as easily changeable as the name (e.g. drop-down box to "change" the file type it's interpreted as whatever the extension), but that's a nightmare that no sensible OS seems to have tackled. It wouldn't be that hard to "default" the mimetype to whatever the "file" utility detects it at on introduction of a new file, but nobody seems to want to do that.
Similarly,the "dot-hidden" files of Linux suffer the same fate. Just the name determines a kind of file attribute, which is silly and potentially dangerous or confusing ("I renamed it .jpg and it disappeared!").
While we're still using OS that trust and encode the filetype into the filename itself, we need the facility to change the detected filetype, and the facility to view the detected filetype. Hiding extensions only achieves the later (with a separate file type column, "Adobe Reader Document", etc.). Any cleverness with showing the extension on rename where it wasn't showing before will mean the user will strip out the file attribute when they rename over the top, and not showing the extension makes it more difficult to change what it opens in.
We created this problem back in the 8.3 filename days and NEVER properly solved it. Encoding metadata into the filename is the cause. Removing it is the solution. But without the infrastructure for that in place in every app, it's pointless to attempt it on a commercial world-wide OS.
I usually compare my computer desktop to my physical desktop.
In this instance, I think it's quite close to compare my smartphone to my pocket.
Sure, if I was walking along and someone handed me a free album, I'd take it. I'd probably put it in my pocket if there was room. But I would also reserve the right to say "No thanks".
What Apple/U2 did is tantamount to chasing you down the street trying to shove their product into your pocket. Sorry, you have no right to do that and it's damn rude, free or not.
People don't have free downloads everywhere, and those that do may not want this crap - whether through playlist OCD of just not wanting to see something they have no interest in or, more likely, BECAUSE it was shoved into their pockets without any option.
If free stuff's so good, I'll ship you a bunch of "free" cardboard box to your house and fill it up (removing it is another issue entirely if we follow the Apple/U2 thing too closely!). Hey, it's free! Why don't you want it? You might like it. You might be a cardboard fanatic. If I offered you a bunch of free cardboard at some points in the year you might tear my arms off to get at it (e.g. moving house). If you'd asked for it, it'd be great. But instead some guy showed up, shoved a bunch of cardboard through your door and said "Hey, it's free, don't complain!".
It's not WHAT was done, it's HOW it was done.
In a zombie apocalypse, all movies and games depict their being some kind of "safe-house" where all the uninfected will gather together.
This, it seems to me, is the most stupid idea ever. Heading towards that is certain death. The zombies only need to be able to read, or get lucky, for it to be game over - they know where the highest concentration of juicy victims will be. And that concentration will increase as everyone piles to go there.
Sure, that's where the firepower might be concentrated too, but that's not good enough when one breaking through is enough to create a game-over scenario.
It seems to me that when everyone is piling towards the safehouse, a prudent course of action would be to stay still / run the other way, unless there's something that stops you doing so.
If I was a zombie, I know exactly where I'd head first.
Similarly, when everyone tries to flee a city, the first thing they do is jam the motorways (freeways). It seems quite stupid to even try if there are really that many people headed that way. All you'll do is trap your vehicle in a queue that you can't escape and then have to get out on foot.
Go the other way, go down the backstreets and side-alleys, stay off the main paths is surely the intelligent thing to do (besides using a bike or other fast transport in the first place).
Non-user-replaceable battery, and no SD expansion.
Stick it up your backside, Samsung, and stop emulating the WORST features of your competitors.
Claiming, or falling for, any argument that "open-source is secure" is a complete failure to understand. Security is relative, not absolute. To get this ass-backwards just makes you look like an idiot. Believing anyone who says ANYTHING "is secure" is utter stupidity (rather than "is more secure", for instance)
It's like saying "metal's secure". No it's not. I can walk around a sheet of metal just as easily as a pane of glass. However, a metal lock built to the same design as, say, a glass one is likely to be MORE secure.
As such, open-source is not "secure". It's considered to be "more" secure if everything else is equal.
Tell me, how many Windows-based Blu-Ray players can you buy in your local supermarket? Zero? Shocking. Why? Because Windows is not properly designed for embedded use at all really, certainly not until very recently. There were ATM's that run on XP Embedded, and there were a few cars that had Windows-based control systems (I remember a story of some ambassador being locked in a car by it).
As such, a complete install of Windows would be vast overkill and a huge attack surface for such things because it's just not ready for that. So it's not a fair comparison at all, as Windows is inherently "less secure" in such circumstances as it's just not designed to do that. However, Linux / open-source Blu-Ray players are "more" secure than many of the alternatives still, everything else being equal.
And why - because not only can you see if your player would be affected by the bugs but, by the same licence that grants you the ability to see the bug, you can CHANGE THE CODE if you so want. No waiting for vendor updates. One person in the world who knows how to code can look, see the problem, fix it, publicise it and therefore purge the world's devices of it.
Tell me, how many "security patches" do you think other commercial Blu-Ray players ever get given? Even the ones connected to the net 24/7 for their "extra media" functionlaity. There's even a facility to update Blu-Ray firmwares via buying new disks (not unlike the Wii games that bundle an update of the underlying OS before they'll let you play them). It's rarely used.
If you're going to pick arguments, have some vague understanding of what the real argument is, not some child on the Internet's poor re-stating of it.
Open-source is potentially MORE secure than closed-source, everything else being equal.
Those conditions and caveats make a HUGE difference to the intent, meaning, and truth of the statement.
Guess what, AES-128 isn't "secure" either. But it may be "more" secure than other algorithms, for example.
I couldn't do anything but reply (could not create a new post), there were only four of five comments when I did, and that was the one most off-topic.
Well, when they're only getting 20-30 comments on most stories now, of course they have to remove quite how few they get on each article from the front page listing too.
Someone had to tell me - It's there if you hover over the bar, but for some reason the "Change" button (which I've never used) is as grey and obvious as it ever was.
+1.
Damn, that's stupidly impossible to see on a whole range of monitors that I have here. I've complained to support, but I doubt they'll do anything.
It's like the Metro Start hover all over again - you have to play some kind of pixel-hunting adventure game the first time you do anything to work out where to go next.
I was LOOKING for it and couldn't find it. And why would you ever want the button to be the same colour as the bar it's in? It's there for a reason - to be pressed. Don't hide it from me.
Confirmed in the Vivaldi browser too, fresh session.
"Want to see Slashdot on your mobile device?" - No, I want it to fecking work on my beast of a PC, thanks.
No, I have Javascript enabled, have latest Chrome. It's just borked, and it only happened today (no updates to software between yesterday and today - same browser session still running in fact!) but the site now doesn't render at all nicely and it LOOKS deliberate, but I'm missing any kind of Post button at all.
Anybody else's Slashdot break today?
I've gone to this top-menu-bar thing, with no left gap at all, with no comment button at all (only Reply To This, sorry!) without warning.
Also, the content is trapped in the left-hand half of the page and won't stretch across.
Not only that, by on the same screen where I have "Ads Disabled" checked, I see an ad.
Slashdot, seriously, without a comment button, I'm gone for good this time.
Sorry, but I could live tomorrow on Linux.
I only use Windows because it was "for free" because of my employer buying me a laptop.
But for five years, I managed and supported a 90% Windows network with hundreds of devices primarily using a laptop which had LibreOffice, etc. installed.
OS - sorted.
Office suite - sorted (sorry, but it is. I used to get people envy my LibreOffice setup, as I could do everything they could do, and manage their same files they managed, and also do things like open ancient foreign formats that people emailled us still).
General apps - sorted.
Games - 1/3rd of my Steam account "just works" on Linux.
For years, I didn't have Windows or Office, as an IT professional supporting users on Windows and Office. Sure, it would have been nice to have a native tool occasionally, but for the odd things I needed (e.g. AD admin tools) it was always safer to just remote-desktop into a Windows machine, or use VM's (Samba tools just aren't there yet).
For everyday use, personal and business, I used Linux as the base OS and for the vast majority of tasks. Only when I was doing something very Windows-specific did I have to load up a Windows tool and always did it from a Linux machine.
Nearly one third of my 900+ games on Steam not enough for you?
Hell, the thing isn't even out yet and already it's prompted hundreds of developers to release their games on Linux too.
My thoughts.
If something had to be doing the maximum possible for its entire existence to get that far, chances are that's not the maximum possible, or we're measuring something wrong.
Not about cost. It's about value.
For that cost, for the price of a couple of drones you could put another couple of officers, stationed permanently to do just their job. And thereby free up whatever officers would also, presumably, need to be present to enable the original drones to operate too.
Simplify the choice - one drone, or two officers (maybe an officer and a half) on the ground doing the same job and NOTHING else - and the value motive really comes to the fore.
Law enforcement isn't about what it costs. Hence why the UK police are still sitting outside the Ecuadorian embassy at the moment waiting for Assange to move out of his personal prison to go to an official one. But it's about value. One high-profile "celebrity" openly-flouting the law is enough to encourage a whole spate of lawlessness in following suit, and you'll have every shoplifter and petty criminal claiming asylum in embassies to evade the law within a month.
So the cost motive would mean we'd leave him in there and forget about him because "he's too expensive to care about". And also that we wouldn't bother to deport illegals.
The value motive says we stay there to dissuade this kind of activity in future and make sure it doesn't cost more in the long run. And that illegals are deported at huge costs to prevent being seen as "weak".
Funny how changing one word (cost -> value) can change the whole intention of your post, isn't it?
And what's happened?
We have gone back to large, simple, flat icons.
We have gone back to a desktop akin to Program Manager (large square blocks of programs on a plain background).
We have gone passed through Active Desktop, only to have it ripped to shreds and reincarnate in the form of the (still loathed) Metro.
We have gone back pre-Start Menus and removed taskbars (in Metro at least), albeit with added features. Then we've put them back in.
The people who work in IT know that these fads come round in cycles and eventually you'll go back to how you were because some things work and some don't. At one point, Windows 95 was going to be "the last Windows OS with a CLI", now we've brought in PowerShell, etc.
And all because, at some point, Microsoft have this ego trip about never being wrong and knowing better. And all the IT guys want is an option. Do I want shiny new stuff, or boring old stuff? Let me choose. It's really that simple. You can put EVERYTHING you want into an OS. Just give me a switch (and ideally a Group Policy setting) to turn it back off if it's not what I want.
I don't want Metro to be thrown in the bin, I want my users to be able to choose whether or not they want to use it. And that means giving them years of both that they can choose between at a click. It takes THIRD-PARTY FREEWARE and faffing about to do that, which isn't the best solution when I've paid for the OS for all my users on an annual subscription.
Nobody cares whether you want new-and-fancy stuff or not. But inevitably, it'll go out of fashion, then come back, then go out. All I want is the option. It's not hard. If a freeware project can give you the option, why can't Windows?
(P.S. Why can't I make Group Policy use AD stored user photos as the Windows 8 login? It's almost impossible with lots of login scripts and downloading an image via third-party software into a particular location on EVERY machine that they might even log onto. Why can't I set a myriad of basic options that are there on the screen for my users, and why can't I *stop* them changing them just as easily? Why can't I just have "skip Metro screen" as an option in Windows to go straight to an old fashioned desktop? Really... it's not that hard).
Scaling is a one-off for the particular size of screen.
There's almost zero overhead in keeping the icons as SVG or similar, and rendering to a bitmap in the device dpi that you require, and then using that bitmap until the screen resolution changes.
Cache enough of them and after the first few resolution changes, you'll never have to render the SVG on that machine again.
So the "scalable has a cost" rules go out of the window, really. And even back in the 90's, did you ever see the stuff you could do with vector formats? Some of them were amazing even if they took 10 or so seconds to render on the machines of the day. Nowadays? Pah. Make it render in the background the first time you change resolution with a "Please Wait" screen showing while you do so. If it takes more than 20-30 seconds for extremely complex icons for the entire Windows control set on a modern machine, I'll be amazed.
There are issues, much like those that plague font sizing, where you have to "hint" at lower resolutions how you want the final bitmap to show (because you might want what looks like an insignificant single pixel to render as a single pixel no matter what the size to make it look right, e.g. cursors, etc.). But those are all solved problems.
Have you seen the speed and zooming size you can get on the MacOS bottom bar? It's fabulous (and I hate Mac with a vengeance). And I've seen it do the same speed on a VM with no 3D acceleration at all.
There's no excuse. This is just designer-pandering where "design" means "look", not functionality, usability, human interface guidelines, intuitiveness, etc.
I've often backed a return to the Windows 3.1 interface for even modern Windows because it was simple, plain, boring and worked, and the icons you could tell roughly what they did. But this is a STEP BACK from Windows 3.1 even. Urk.
And, yes, I have deployed the previews in VM's for on-site testing and, apart from the look-change, there's little of note there. I'd have been infinitely more happy with a 7 or even 8 clone with newer features.
Hell, I've used 8 on a tablet and on touchscreen PC's and each time it's easier to turn that Metro junk off and use it, and even then it's just NOT designed for proper touch usage.
1) Robots cannot replace all jobs. We've yet to make a self-fixing robot of any note. AI is NOWHERE NEAR capable of doing the simplest of paperwork or administration. Hell, we've barely automated anything of the IT departments, let alone anywhere else. All they can automate are mindless, repetitive, labour-intensive (and sometimes dangerous) jobs. Though that puts a LOT of people out of work, that's by far not the majority.
2) If robots do replace all jobs, the "money" comes from sale of goods just the same. Half the workforce are working and doing the work of the other half - the robots produce the goods / services, and the humans lounge at home. In fact, if anything, we'll have more money because we could produce more for less maintenance costs and have no union troubles - once the machine is in, it'll carry on working until the power goes out, effectively, and cost pence to run in the meantime as an "hourly wage".
3) The result of the above is that food and goods become so cheap and plentiful that the concept of "buying" them will seem old hat. If a government could pay for itself by selling goods to other countries still, then it doesn't really matter what you do - you could quite literally be paid to stay at home, if all you have to do to "do" the same amount of work is power up a robot and oil him once a week Mass amounts of automated robots also make self-sufficiency much more possible - imagine that you don't have to farm or buy goods, just let the same robot that works for Kelloggs loose on your land and it'll feed you for minimal cost.
4) This is all a pipe-dream. By the time you have that sort of automation, the only jobs left would be bureaucratic - and they'd realise they're the only ones working. Things would flip on their head.
5) Who cares? If you have no job and no money but food is so cheap that going an oiling a robot once a month pays for everything - wow... perfect life.
To protect the metadata of the recipient is daft. How are intermediary servers ever supposed to know? And if you and the other end of the connection both set up a connection and know who it's for, that blows out the "fact that a message was sent" before you start.
Message length is also stupid to try to hide. Sure, it may not be exact but if I send a 200Mb email and you send 20 characters, how are you supposed to encrypt those to be indistinguishable without literally padding to the nearest 200Mb? And padding might be able to be done on a smaller block basis, but it's exactly that kind of padding that broke open gzipped-and-encrypted data in SSL sessions for a while.
Timestamps? Who needs them - if someone's listening, they know what time you contacted the server and it'll be +/- a couple of minutes of when you wanted the email sent (i.e. immediately, in almost all cases).
Sender? Obscuring that while connecting FROM an email server that's trusted to be honest about only delivering mail for known local users (untrusted ones are spam sources and will get blocked no matter what encryption they use)... there's your metadata.
Metadata is, in and of itself, almost impossible to remove. The fact that you've connected to a target mail server is metadata.that you're probably sending an email to an email account at that server. Anything else is either between you and that first server (so forwarding is a pain, re-enveloping is a pain, bouncebacks are a pain, there's a stupidly high computing cost associated with receiving spam, etc.) or has to be announced somehow.
And that would all come back to entirely encrypted and obfuscated and peer-to-peer networking globally. You're talking about a Tor problem, not an email problem.
If you don't want people to read a message, encrypt it. If you don't want people to ever know a message existed, you have to be able to send it in such a way that nobody could ever know. And that's all-but impossible at the moment. The best we have is hoping that sending it in bits to a random selection of unknown strangers will protect you enough that they could never collaborate against you (Tor).
That's not the problem with GPG.
The problem with it is that I could never be bothered to use it, not because of privacy (it would be incredibly convenient to send, say, a password required in an emergency via a verifiably-encrypted email) - but because it's such a faff. And it interferes with everything (searching, archiving, re-enveloping etc.). And to do so is all bolt-on-and-bodge-job methods. None of the major email clients offered anything like proper encryption by default.
And as soon as you get into using plugins, most people just won't bother. There are plugins for PFS for all your instant messenger programs, etc. - I had one installed for about 5 years, the only other person I know with one installed has a different, incompatible one. Now I don't use IM much any more anyway, so it's dead in the water.
And all email encryption is a ton of messing about with publishing keys in the right places, and having to verify against those places, etc. It's ugly.
The only place I've seen anything like GPG working is in package signing for third-party software. And there you have to download the package, download the key (either from the same website as the package - WOOP WOOP - or independently crowd-source a verified key), and then check it works. I've only ever bothered for Slackware, for which I believe the ISO images are signed with the official Slackware key.
GPG is just a pain in the butt and not automated at all. It's easier to compose and encrypt the message ENTIRELY OFFLINE and then send the encrypted text, and that shows you what kind of automation is missing, and what kind of trust system is actually in place.
Sure, there are plugins, helpers, hacks, extensions and all sorts. But none of them ever progressed to being "in" the software. Not even software designed to do nothing else but send email.
It gets as bright as a quarter full moon on a pinpoint in the sky.
Where the feel of "what we would see (and feel)" comes into it, I have no idea.
Long article, for simple answer, that isn't even that interesting.
Personally, the most interesting bit was the bit about a previous supernova in the 1000's that looks like a cloud of dust now.