This just means all the people who wanted to demand something of Dell - more RAM, shinier cases, please go AMD - demanded Linux too half the time.
The people who signed up on the site to make their little ideastorms aren't representative of Dell's entire customer base even in the slightest.
You can bet there is a post on every Linux distro forum linking Dell's site telling people to go and suggest preinstalled Linux of some flavor. That skews the results too much. Internet polls just don't work, it's a fact. For statistics to be useful you first have to know what you are going to do with them once you've got the data (i.e. have a goal that needs statistics, don't collect statistics for shits and giggles) and when you do collect the data you need to be suitably impartial. There are good ways of collecting data about customer needs the same way there are good ways to interview employees for a job (psychometric testing ftw)
I'm in the UK at the moment, aiming to get back to my apartment in Texas. When I see a store has one I want to get one.. but what's the point out here? I would just have to get another.
The problem when I get back is despite the dearth of new games for the system since launch, there are enough that'll make it a painful choice of first purchases. As time goes on more are released and.. argh.. maybe I will have to pass up Zelda, or some other game, so I can get a good selection.
So, not excited since I can't have one, and when I do get one I won't be playing ALL the games I wanted to or could have bought up until now. It's like.. being 6 months behind, for a geek, that's bad..
MAC addresses can be changed but despite the "currently in use" MAC on your board being different to the factory default, the original hardcoded MAC address is always visible to the OS somehow. Just changing the setting does not lose that information.
You could always uniquely identify an ethernet adapter, and barring reflashing the chip eeprom that stores this information, it's not user changable.
Processor serial numbers are about as innocuous as a privacy concern as if you used your grocery store loyalty card. To say that someone is going to target you because you have a certain loyalty to the grocery store is ludicrous.
Uniquely identifying systems is ESSENTIAL to the current internet and DRM problems.
Just think, if a processor serial number had become a standard, they may not have decided so fast that they needed TPM and per-machine iTunes authorizing so hackneyed, and so on. Of course you can be uniquely identified on the internet. How much crazy hashing crap like this would it have made totally unecessary?
How do you call code in a "linear fashion" on Linux but supposedly not in Windows? It follows a line, in both cases, from a first function call to the end of the function call. None of the diagrams show what data is going through (even as an example) or what the intent of the function call is apart from the backtrace. It doesn't show anything but that every call trace goes back through a certain set of functions. Windows will have more for a lot of reasons - the fact that it is fundamentally a microkernel, it has a much more engaged message passing subsystem and set of HALs and interaction with "kernel-mode" drivers (especially the IIS acceleration features which are NOT implemented the same way as Linux's kernel http acceleration in the same way that neither are FreeBSD's HTTP accept filter).
That doesn't mean it's fundamentally worse, it just means it is architectured a little differently. As I said for each call there is no trace of how much code is executed nor it's data. For all you know from the diagram a lot of it is passing through DLL interfaces to other DLLs, kernel interfaces and driver subsystems. You can't just count the code back and say "well, it passed through more wrappers so it's harder to secure". Which of those functions are simply passing the data back onto the stack and calling the same principle function in another DLL?
The VisualC++ libc runtime has a crapload of "security checks" wrestled into it if you use the right compiler flags, which pass it through many more layers than in glibc.. at the expense of slowing it down. On Linux, it's not in the C runtime, it's in stuff like SELinux and PAX. Stack checking and canaries and suchlike. You think that makes the code LESS secure that it is in the C runtime?
Starting with Windows Server 2003 (and XP SP2 and obviously a shitload more in Vista), even more is done on the OS side of things, built into NX bit handling and a ton of other checks. All of these run through all kinds of little calls. Is it less secure if you have 1 bug in 10,000 massively interlinked calls on Windows than if you have the same bug in only 1000 calls in Linux?
This is also just the STARTUP of Apache and IIS. Does IIS create more buffers, caches, spawn processes earlier, do any configuration details, doesn't registry access take a lot longer to do, isn't pulling configuration data out of Active Directory a very heavy process compared to parsing an Apache config file? How do these relatively innocuous things affect the security? It is just as easy for Apache to have a maliciously encoded line in a config file which is NOT going to affect a fixed, due to a parser bug or somesuch, which simply cannot be gained from accessing a fixed, structured LDAP database with significantly more strict rules (a lot of the data you'd want to exploit by mangling directives and arguments, you couldn't put in the Active Directory anyway - other parts of the OS that IIS doesn't even consider, is at work here).
The diagrams mean absolutely nothing. However of one thing I am sure; this guy earned ZDnet a lot of banner impressions today.
Looking at them both I notice the Linux one has a lot of empty space and draws the graph longer, and has a wider pixel size.
The Windows one is shorter and thinner.
Those simple differences make the Linux one look less messy, even if they ARE that simple.
I'm sure you could easily refactor the Windows one to look less messy. The real details though are in what those system calls are; and the two images provided are too small to SEE the names and routes of the system calls where relevant. Just looking at the tangle and counting lines is NOT a security audit.
Just because it doesn't generate heat doesn't mean it isn't losing power. The energy wasted in relation to the power on the cable is probably quite high (DC doesn't travel well, that's why wall power is AC remember) compared to just wiring up wall sockets and using warts or switching PSUs.
You're just transferring the wall-wart to another room though, and making the loss over the cable add to the power inefficiency. Imagine the extra airconditioning provision the room with the new site-wide AC-DC converter will need:D
PoE is a clever way to power devices that are in hard to power places (where you can wire a network using a thin cable but far away from a power socket) and keeps devices cheap (no need to do anything but DC-DC conversion from PoE to components) but it's not any better energy-efficiency-wise.
Can't this IEEE stuff they're talking about simply be built into drivers? I know my laptop ethernet (Intel) has the ability to scale down the ethernet speed when the battery is in use, or during standby and so on. Would it cause too much trouble to have the driver anticipate and schedule a renegotiation on a power source change or based on activity? Why would ethernet vendors need to be involved if it was simply a driver 'problem' - apart from having to write drivers that do it for their hardware (which most of them DO already).
Can't we have a sysctl or a sysfs tweak in Linux/BSD/whatever to demonstrate it and see if it even helps? Does networking hardware at the other end (for instance a 32-port Cisco switch) actually use less power if half it's ports are at 10mbit rather than 100mbit?
Can't we do this with wireless? 802.11b etc. already has power calibration built in but could it pull it back when the bandwidth requirement isn't so high, saving battery life and not polluting the airwaves with high powered chatter? My card uses the same transmit power whatever the state of the laptop is..
I used to use Crystal as well as doing real code; neither job paid any better than the other.
There's no precedent for businesses who use technology to pay more than businesses that are IN technology either. It depends on the workload, and how useful the stuff is. The chef at Google cafeteria probably gets paid more than I do; however I am fairly sure I am being paid more still than the guys I used to work with who used Crystal or Delphi RAD crap.
A lot of people who invented and trailblazed all the stuff we're using now even on this site, are in very cushy very high paid, maybe research grant-assisted jobs. Far more secure than some guy clicking buttons to create a quick app to help the secretaries.
I don't see how you took how Linux and Windows make money from the programmers who made it.
Plenty of companies market Linux perfectly well. It has a larger server share than Windows does. RedHat hasn't gone bankrupt yet:)
There's no correlation whatsoever between business development programmers, and systems level programmers, and the sales of their product. You're talking about marketing.
Your salary doesn't start again at the lowest grade just because you get a degree!? Maybe when you are 24 and just get out of Uni with your little bits of paper, you would be on $40k less than he is now, but in your example he would STILL have 15 years experience. That counts for a lot. Not as much as the degree AND experience, but a lot anyway.
If he can get decent jobs already without the CS degree I'd wonder if it was worth getting, although I've been thinking the same thing - it doesn't make the job you're in any easier (most CS after so much experience is stuff you've done before. I did all the project management and general "Computing" stuff during my GCSE/A-Levels and learnt to code off my own back. It's not failed me yet and all the jobs I've been in have been on experience and general talent.
At least at the University I worked, there were two pay grades; academic and technical. If you didn't have a degree you were confined to technical. It paid less and you were automatically given less responsibilities. By the time I got out of the job I was earning as much as the incoming academic grade employees (and no student loan debt!). Now that I think I could run a CS degree in my spare time in quick time, I might. If anything, his pay should go up for his next job, just because they will be more sure of him because of the degree - when there is a range of starting salaries, you start from the degree and ramp up based on experience. Instead of them offering you the lowest amount for that position (either as a graduate student or as uncertainty and insurance for them if they can't quantify past experience to the new job) they will start higher because of the added experiene.
It won't make any job you're in more complicated - the current employers won't think "he has a degree now let's move him on to harder stuff" - they already know what you can do, you won't be asked to do anything more than you know.
I have 10 years experience at least on my part, I used to work in the CS department in a University, ironically after I left school and couldn't be bothered to get a degree. I determined I would rather have the money and not be bogged down with loans and homework, I valued my social life more than any student could manage without flunking, and the work they did on a CS degree then was.. really a bit much. Now, I look at the work some of my friends and colleagues are doing for CS degrees, and I end up helping with their homework and explaining past exam questions to them. It's SO easy to get one, especially if you've got the experience and been doing that for years.
It will do nothing but make the job hunting process a little easier; you can't evaluate experience past a certain point, but the degree will make prospective employers at least consider you a baseline of knowledge based on the degree or certification. Oh, and if you go the whole hog you can have letters after your name in 3 or 4 years, without even flinching.
Microsoft will simply refuse to activate a product where the key is a well-known pirated one. Most of them turn up on Google and ostensibly on the MSN/Live search engine so I suspect they are not above checking for them and blacklisting them regularly. The moment it goes onto the internet and someone uses it more than once without good reason (think of bittorrent sites where you see copies of XP with 1000 seeders..).
Most people even with heavy hardware-upgrade needs only ever have to call Microsoft once or twice a year, if that. I've called them once a year when Windows has gone down on SOME machine but it's never been the same copy of Windows or the same machine at the time. If you have to run 3 or 4 machines and repair or make updates, you're going to have to call:)
Buying another copy of Windows after a repair, shouldn't really be necessary, but this guy's problem seems to be quite unique, but they'll understand anyway if it's only the first time he's had to activate.
Whenever I've done it, it's taken far, far less time than 15 minutes.
Anyone who complains that they need to reactivate their system and is worried about having to "justify" it to Microsoft is talking shit. 3 minutes with barely 3 bars of hold music, you get to an operator - "I had to replace my graphics card because it exploded", or "my laptop died and Sony repaired it" - both real events here, and they shrug and say "thanks".
It's just so they can enter it into their logs on how things are used. I think it's more market research than antipiracy. At the end of the day a pirate will NEVER call the 1-800 number, so if you ARE calling up with a legit key, they are happy to assist. And, like I said, it's barely 3 minutes.
Neither of those are any good for this hypothetical guy who installed Ubuntu using a one-click installer are they? No repartitioning, that was the point.
I would say that having a common user folder/document folder structure across operating systems is something important to interoperability. At some point you did care about booting Windows and sharing the files, wouldn't it be easier if you didn't have to resort to a 3rd-party tool to get things done for you, or pick a non-standard location to save a file (everything on Windows defaults to Desktop or My Documents, everything on Linux will try and save it to your ~/Documents folder or ~/Desktop folder) - sharing user accounts (at least by association) would ease people into Linux by letting them have all their stuff in one place.
Even if it's having the Linux installer symlink the directories to the appropriate Linux locations via a little clicky installer.
I thought the FUSE ntfs-3g was pretty stable now? First release it sucked and trashed a couple of simple test disks I had but they fixed that stuff pretty damn quick.
That kind of system didn't work so well on Windows 2000 though. I am really surprised it took so long for the distros to work something up for 2K/XP/Vista, I guess the stumbling block is loop-mounting a file as disk space requires write access to the underlying filesystem and they were waiting on something like ntfs-3g - although ntfs write support in the kernel was always just fine and dandy.. I guess it is a matter of quality assurance and safety of data and not technical ability..
I'd be impressed to see what else comes out of this. Perhaps a Linux distro which will install under the Program Files folder in Vista, and reuse the Windows documents folder, all running off the native NTFS filesystem (one thing users will always hate about dual boot is all their documents being spread around, and if the filesystem for Ubuntu or Debian's installers is in a disk image, and people default to/home/myuseraccount, it will do nothing but annoy..)
It's a loop-mounted ext3 "hard file" like you get in an emulator.
The process is, basically - GRUB loads a kernel+initrd from the Windows filesystem. Kernel loads, mounts / from the initrd, mounts the NTFS or FAT filesystem from the Windows box, and finds the hardfile and initrd - then it swivels root to use the image via the loopback filesystem (so you can mount files as disks).
Not sure how this bodes for expandability of the disk image though. I guess the idea is the Ubuntu install just works, and you can put the data back onto your Windows disk..?
Meh. What about BSD, the embedded systems like eCos/RTMS? GNU in general runs everywhere. MacOS X is based on Open Source.
Why only focus on Linux? OSDL used to be a bit Linux-biased but now this is just ridiculously narrow in scope, Linux just isn't suitable in every environment for every task.
TCPA, HDMI and all the other DRM technologies do one simple thing to data; attempt to ensure that a simple man-in-the-middle attack to intercept that data cannot happen between a trusted source and an approved target.
The threat that "DRM will make your computer useless" is not really relevant. Nothing in TCPA really stipulates that all content must be filtered through DRM or that all content must be encrypted or obfuscated. All pipes between components may be marked trusted or not, and when content marked "only trusted paths" comes in, only those trusted components can be used. For data for which you have no trusted components, this is NOT down to you "being crippled by the industry" so much as "I didn't buy that component yet". Be it a DirectShow module, Apple's iTunes application, DVD player app. Whatever.
However, Vista does not make any impact on.. high definition movies you encoded yourself without DRM, or high definition movies that have in general not been set to enforce DRM. Nor can it reasonably stop you from playing an unencrypted MP3 file. If you don't like your DRM media being crippled by the fact that you do not have these trusted components available you have two choices - either buy the trusted component (DVD player plugin, new monitor with HDCP), or simply don't buy DRM media. In either case, all your chosen media will keep working, encrypted or not, just maybe at a lower resolution, more limited functionality, but certainly not USELESS (no manufacturer on the planet is proposing to flat-out deny access to media based on an insecure path, and the DRM consequence of trying to play someone else's tune on your player is a matter of encryption key availability)
Neither system is truly any different to people insisting on PGP Signing and Encrypting their mail.
Since it IS consumers who drive the market on this, and significant backpeddling has been done to ensure that systems carry on working even through violations, things may get better (relaxing restrictions). Having a non-HDCP monitor means, your content gets displayed at higher-than-SD but lower-than-HD resolution. Those who give a shit about this will have already known about it, and bought the correct monitor. They are not affected by the DRM, therefore.
Apple own and hold copyright on Darwin, they released their own code under APSL for YOU to use and YOU to give modifications back (i.e. this is their license protection for you making a commercial OS X clone)
They can do whatever THEY like to it and never release the source, just like any GPL code author is free (under the terms of the GPL, even) to relicense their code for any party they see fit (BSD, APSL, whatever). It is up to the author and the copyright holder, if they are even in fact different people. Apple are both!
So OS X doesn't run on ARM? Why not? Because OpenDarwin doesn't? This whole article is horseshit speculation and a completely random nonsense of misunderstanding how software licensing works, who wrote and owns Darwin (Apple!) and the technical aspects involved (they've been working on the iPhone for the better part of a year and a half.. that's plenty of time to do a port to a new processor, especially given how abstracted the Darwin kernel is, XNU Apple additions and so on)
Remember this is not just a phone, but an iPod too; with a widescreen display. It fits somewhere inbetween the feature capability of the iPod Nano (storage space) and the full Video iPod (capability) with extra stuff involved.
As a successor to the iPod, considering the prices, it hits a fairly good niche. Do you need an MP3 player *and* a capable PDA-phone? Then why not have them both in one device? You may spend $300 on an iPod and $200 on a phone, now you can get them both in the same device. You also get a fairly reasonable digital camera, and it plays all your iTunes media probably a ton better than a real iPod did (nicer screen anyway).
I simply fear for fingerprints, and earprints (it turns the screen off when you hold it to your ear, but all that surface back and front..). Apple must have their shiny and glassy look. And there's no impetus on using a stylus, it's all multitouch and pinky fingers.. a lot of people will just get frustrated at how grimey it ends up looking during real use.
There's definitely a convergence market for it. MP3 phones these days usually suck as MP3 players. MP3 players as phones.. this is the first. It just so happens to be super-capable at both, without springing for a full PDA.
I never said it should be the main direction, I said it would be nice to see.
You could get KDE into a lot more places (maybe EmbeddedKDE to replace QTopia?:) if it had a more embedded slant, and also reduce the system requirements needs of a lot of devices earmarked to run Linux for operation.
Okay so it wouldn't be any good for you as a desktop user who wants to use major chip design tools, but for the vast majority of people who buy eMachines and Dell boxes not for games, not for huge amounts of work, just for email, maybe VoIP phoning and webcam, or just browsing YouTube..
Tailoring a system for embedded use is a good way of identifying those parts which are "non-essential" or can be optimised for memory or performance on low powered devices. If it runs better on low-power devices, it will run better on high-powered ones too if the components are being worked on to that end.
Slashdot Mirror
This just means all the people who wanted to demand something of Dell - more RAM, shinier cases, please go AMD - demanded Linux too half the time.
The people who signed up on the site to make their little ideastorms aren't representative of Dell's entire customer base even in the slightest.
You can bet there is a post on every Linux distro forum linking Dell's site telling people to go and suggest preinstalled Linux of some flavor. That skews the results too much. Internet polls just don't work, it's a fact. For statistics to be useful you first have to know what you are going to do with them once you've got the data (i.e. have a goal that needs statistics, don't collect statistics for shits and giggles) and when you do collect the data you need to be suitably impartial. There are good ways of collecting data about customer needs the same way there are good ways to interview employees for a job (psychometric testing ftw)
I'm in the UK at the moment, aiming to get back to my apartment in Texas. When I see a store has one I want to get one.. but what's the point out here? I would just have to get another.
The problem when I get back is despite the dearth of new games for the system since launch, there are enough that'll make it a painful choice of first purchases. As time goes on more are released and.. argh.. maybe I will have to pass up Zelda, or some other game, so I can get a good selection.
So, not excited since I can't have one, and when I do get one I won't be playing ALL the games I wanted to or could have bought up until now. It's like.. being 6 months behind, for a geek, that's bad..
Ask the law firm attached to the big American company you work for.
That's what I do?
MAC addresses can be changed but despite the "currently in use" MAC on your board being different to the factory default, the original hardcoded MAC address is always visible to the OS somehow. Just changing the setting does not lose that information.
You could always uniquely identify an ethernet adapter, and barring reflashing the chip eeprom that stores this information, it's not user changable.
Processor serial numbers are about as innocuous as a privacy concern as if you used your grocery store loyalty card. To say that someone is going to target you because you have a certain loyalty to the grocery store is ludicrous.
Uniquely identifying systems is ESSENTIAL to the current internet and DRM problems.
Just think, if a processor serial number had become a standard, they may not have decided so fast that they needed TPM and per-machine iTunes authorizing so hackneyed, and so on. Of course you can be uniquely identified on the internet. How much crazy hashing crap like this would it have made totally unecessary?
Absolute bullshit.
2 5/takingBufferOverrunsSeriously.html
How do you call code in a "linear fashion" on Linux but supposedly not in Windows? It follows a line, in both cases, from a first function call to the end of the function call. None of the diagrams show what data is going through (even as an example) or what the intent of the function call is apart from the backtrace. It doesn't show anything but that every call trace goes back through a certain set of functions. Windows will have more for a lot of reasons - the fact that it is fundamentally a microkernel, it has a much more engaged message passing subsystem and set of HALs and interaction with "kernel-mode" drivers (especially the IIS acceleration features which are NOT implemented the same way as Linux's kernel http acceleration in the same way that neither are FreeBSD's HTTP accept filter).
That doesn't mean it's fundamentally worse, it just means it is architectured a little differently. As I said for each call there is no trace of how much code is executed nor it's data. For all you know from the diagram a lot of it is passing through DLL interfaces to other DLLs, kernel interfaces and driver subsystems. You can't just count the code back and say "well, it passed through more wrappers so it's harder to secure". Which of those functions are simply passing the data back onto the stack and calling the same principle function in another DLL?
http://radio.weblogs.com/0105852/stories/2003/01/
The VisualC++ libc runtime has a crapload of "security checks" wrestled into it if you use the right compiler flags, which pass it through many more layers than in glibc.. at the expense of slowing it down. On Linux, it's not in the C runtime, it's in stuff like SELinux and PAX. Stack checking and canaries and suchlike. You think that makes the code LESS secure that it is in the C runtime?
Starting with Windows Server 2003 (and XP SP2 and obviously a shitload more in Vista), even more is done on the OS side of things, built into NX bit handling and a ton of other checks. All of these run through all kinds of little calls. Is it less secure if you have 1 bug in 10,000 massively interlinked calls on Windows than if you have the same bug in only 1000 calls in Linux?
This is also just the STARTUP of Apache and IIS. Does IIS create more buffers, caches, spawn processes earlier, do any configuration details, doesn't registry access take a lot longer to do, isn't pulling configuration data out of Active Directory a very heavy process compared to parsing an Apache config file? How do these relatively innocuous things affect the security? It is just as easy for Apache to have a maliciously encoded line in a config file which is NOT going to affect a fixed, due to a parser bug or somesuch, which simply cannot be gained from accessing a fixed, structured LDAP database with significantly more strict rules (a lot of the data you'd want to exploit by mangling directives and arguments, you couldn't put in the Active Directory anyway - other parts of the OS that IIS doesn't even consider, is at work here).
The diagrams mean absolutely nothing. However of one thing I am sure; this guy earned ZDnet a lot of banner impressions today.
Looking at them both I notice the Linux one has a lot of empty space and draws the graph longer, and has a wider pixel size.
The Windows one is shorter and thinner.
Those simple differences make the Linux one look less messy, even if they ARE that simple.
I'm sure you could easily refactor the Windows one to look less messy. The real details though are in what those system calls are; and the two images provided are too small to SEE the names and routes of the system calls where relevant. Just looking at the tangle and counting lines is NOT a security audit.
Just because it doesn't generate heat doesn't mean it isn't losing power. The energy wasted in relation to the power on the cable is probably quite high (DC doesn't travel well, that's why wall power is AC remember) compared to just wiring up wall sockets and using warts or switching PSUs.
:D
You're just transferring the wall-wart to another room though, and making the loss over the cable add to the power inefficiency. Imagine the extra airconditioning provision the room with the new site-wide AC-DC converter will need
PoE is a clever way to power devices that are in hard to power places (where you can wire a network using a thin cable but far away from a power socket) and keeps devices cheap (no need to do anything but DC-DC conversion from PoE to components) but it's not any better energy-efficiency-wise.
Can't this IEEE stuff they're talking about simply be built into drivers? I know my laptop ethernet (Intel) has the ability to scale down the ethernet speed when the battery is in use, or during standby and so on. Would it cause too much trouble to have the driver anticipate and schedule a renegotiation on a power source change or based on activity? Why would ethernet vendors need to be involved if it was simply a driver 'problem' - apart from having to write drivers that do it for their hardware (which most of them DO already).
Can't we have a sysctl or a sysfs tweak in Linux/BSD/whatever to demonstrate it and see if it even helps? Does networking hardware at the other end (for instance a 32-port Cisco switch) actually use less power if half it's ports are at 10mbit rather than 100mbit?
Can't we do this with wireless? 802.11b etc. already has power calibration built in but could it pull it back when the bandwidth requirement isn't so high, saving battery life and not polluting the airwaves with high powered chatter? My card uses the same transmit power whatever the state of the laptop is..
I used to use Crystal as well as doing real code; neither job paid any better than the other.
There's no precedent for businesses who use technology to pay more than businesses that are IN technology either. It depends on the workload, and how useful the stuff is. The chef at Google cafeteria probably gets paid more than I do; however I am fairly sure I am being paid more still than the guys I used to work with who used Crystal or Delphi RAD crap.
A lot of people who invented and trailblazed all the stuff we're using now even on this site, are in very cushy very high paid, maybe research grant-assisted jobs. Far more secure than some guy clicking buttons to create a quick app to help the secretaries.
I don't see how you took how Linux and Windows make money from the programmers who made it.
:)
Plenty of companies market Linux perfectly well. It has a larger server share than Windows does. RedHat hasn't gone bankrupt yet
There's no correlation whatsoever between business development programmers, and systems level programmers, and the sales of their product. You're talking about marketing.
I don't understand why you'd take a pay cut.
Your salary doesn't start again at the lowest grade just because you get a degree!? Maybe when you are 24 and just get out of Uni with your little bits of paper, you would be on $40k less than he is now, but in your example he would STILL have 15 years experience. That counts for a lot. Not as much as the degree AND experience, but a lot anyway.
If he can get decent jobs already without the CS degree I'd wonder if it was worth getting, although I've been thinking the same thing - it doesn't make the job you're in any easier (most CS after so much experience is stuff you've done before. I did all the project management and general "Computing" stuff during my GCSE/A-Levels and learnt to code off my own back. It's not failed me yet and all the jobs I've been in have been on experience and general talent.
At least at the University I worked, there were two pay grades; academic and technical. If you didn't have a degree you were confined to technical. It paid less and you were automatically given less responsibilities. By the time I got out of the job I was earning as much as the incoming academic grade employees (and no student loan debt!). Now that I think I could run a CS degree in my spare time in quick time, I might. If anything, his pay should go up for his next job, just because they will be more sure of him because of the degree - when there is a range of starting salaries, you start from the degree and ramp up based on experience. Instead of them offering you the lowest amount for that position (either as a graduate student or as uncertainty and insurance for them if they can't quantify past experience to the new job) they will start higher because of the added experiene.
It won't make any job you're in more complicated - the current employers won't think "he has a degree now let's move him on to harder stuff" - they already know what you can do, you won't be asked to do anything more than you know.
I have 10 years experience at least on my part, I used to work in the CS department in a University, ironically after I left school and couldn't be bothered to get a degree. I determined I would rather have the money and not be bogged down with loans and homework, I valued my social life more than any student could manage without flunking, and the work they did on a CS degree then was.. really a bit much. Now, I look at the work some of my friends and colleagues are doing for CS degrees, and I end up helping with their homework and explaining past exam questions to them. It's SO easy to get one, especially if you've got the experience and been doing that for years.
It will do nothing but make the job hunting process a little easier; you can't evaluate experience past a certain point, but the degree will make prospective employers at least consider you a baseline of knowledge based on the degree or certification. Oh, and if you go the whole hog you can have letters after your name in 3 or 4 years, without even flinching.
:)
:)
Microsoft will simply refuse to activate a product where the key is a well-known pirated one. Most of them turn up on Google and ostensibly on the MSN/Live search engine so I suspect they are not above checking for them and blacklisting them regularly. The moment it goes onto the internet and someone uses it more than once without good reason (think of bittorrent sites where you see copies of XP with 1000 seeders..).
Most people even with heavy hardware-upgrade needs only ever have to call Microsoft once or twice a year, if that. I've called them once a year when Windows has gone down on SOME machine but it's never been the same copy of Windows or the same machine at the time. If you have to run 3 or 4 machines and repair or make updates, you're going to have to call
Buying another copy of Windows after a repair, shouldn't really be necessary, but this guy's problem seems to be quite unique, but they'll understand anyway if it's only the first time he's had to activate.
Whenever I've done it, it's taken far, far less time than 15 minutes.
Anyone who complains that they need to reactivate their system and is worried about having to "justify" it to Microsoft is talking shit. 3 minutes with barely 3 bars of hold music, you get to an operator - "I had to replace my graphics card because it exploded", or "my laptop died and Sony repaired it" - both real events here, and they shrug and say "thanks".
It's just so they can enter it into their logs on how things are used. I think it's more market research than antipiracy. At the end of the day a pirate will NEVER call the 1-800 number, so if you ARE calling up with a legit key, they are happy to assist. And, like I said, it's barely 3 minutes.
Neither of those are any good for this hypothetical guy who installed Ubuntu using a one-click installer are they? No repartitioning, that was the point.
I would say that having a common user folder/document folder structure across operating systems is something important to interoperability. At some point you did care about booting Windows and sharing the files, wouldn't it be easier if you didn't have to resort to a 3rd-party tool to get things done for you, or pick a non-standard location to save a file (everything on Windows defaults to Desktop or My Documents, everything on Linux will try and save it to your ~/Documents folder or ~/Desktop folder) - sharing user accounts (at least by association) would ease people into Linux by letting them have all their stuff in one place.
Even if it's having the Linux installer symlink the directories to the appropriate Linux locations via a little clicky installer.
I thought the FUSE ntfs-3g was pretty stable now? First release it sucked and trashed a couple of simple test disks I had but they fixed that stuff pretty damn quick.
That's SO user friendly! A 30 page document! Much better than a one-click installer!!!
That kind of system didn't work so well on Windows 2000 though. I am really surprised it took so long for the distros to work something up for 2K/XP/Vista, I guess the stumbling block is loop-mounting a file as disk space requires write access to the underlying filesystem and they were waiting on something like ntfs-3g - although ntfs write support in the kernel was always just fine and dandy.. I guess it is a matter of quality assurance and safety of data and not technical ability..
/home/myuseraccount, it will do nothing but annoy..)
I'd be impressed to see what else comes out of this. Perhaps a Linux distro which will install under the Program Files folder in Vista, and reuse the Windows documents folder, all running off the native NTFS filesystem (one thing users will always hate about dual boot is all their documents being spread around, and if the filesystem for Ubuntu or Debian's installers is in a disk image, and people default to
It's a loop-mounted ext3 "hard file" like you get in an emulator.
The process is, basically - GRUB loads a kernel+initrd from the Windows filesystem. Kernel loads, mounts / from the initrd, mounts the NTFS or FAT filesystem from the Windows box, and finds the hardfile and initrd - then it swivels root to use the image via the loopback filesystem (so you can mount files as disks).
Not sure how this bodes for expandability of the disk image though. I guess the idea is the Ubuntu install just works, and you can put the data back onto your Windows disk..?
Surely a judge rules, not a jury? Juries render verdicts.
JUST Linux?
Meh. What about BSD, the embedded systems like eCos/RTMS? GNU in general runs everywhere. MacOS X is based on Open Source.
Why only focus on Linux? OSDL used to be a bit Linux-biased but now this is just ridiculously narrow in scope, Linux just isn't suitable in every environment for every task.
TCPA, HDMI and all the other DRM technologies do one simple thing to data; attempt to ensure that a simple man-in-the-middle attack to intercept that data cannot happen between a trusted source and an approved target.
The threat that "DRM will make your computer useless" is not really relevant. Nothing in TCPA really stipulates that all content must be filtered through DRM or that all content must be encrypted or obfuscated. All pipes between components may be marked trusted or not, and when content marked "only trusted paths" comes in, only those trusted components can be used. For data for which you have no trusted components, this is NOT down to you "being crippled by the industry" so much as "I didn't buy that component yet". Be it a DirectShow module, Apple's iTunes application, DVD player app. Whatever.
However, Vista does not make any impact on.. high definition movies you encoded yourself without DRM, or high definition movies that have in general not been set to enforce DRM. Nor can it reasonably stop you from playing an unencrypted MP3 file. If you don't like your DRM media being crippled by the fact that you do not have these trusted components available you have two choices - either buy the trusted component (DVD player plugin, new monitor with HDCP), or simply don't buy DRM media. In either case, all your chosen media will keep working, encrypted or not, just maybe at a lower resolution, more limited functionality, but certainly not USELESS (no manufacturer on the planet is proposing to flat-out deny access to media based on an insecure path, and the DRM consequence of trying to play someone else's tune on your player is a matter of encryption key availability)
Neither system is truly any different to people insisting on PGP Signing and Encrypting their mail.
Since it IS consumers who drive the market on this, and significant backpeddling has been done to ensure that systems carry on working even through violations, things may get better (relaxing restrictions). Having a non-HDCP monitor means, your content gets displayed at higher-than-SD but lower-than-HD resolution. Those who give a shit about this will have already known about it, and bought the correct monitor. They are not affected by the DRM, therefore.
We should loose a pack of wolves on them. That would be loosing, wouldn't it? :)
Apple own and hold copyright on Darwin, they released their own code under APSL for YOU to use and YOU to give modifications back (i.e. this is their license protection for you making a commercial OS X clone)
They can do whatever THEY like to it and never release the source, just like any GPL code author is free (under the terms of the GPL, even) to relicense their code for any party they see fit (BSD, APSL, whatever). It is up to the author and the copyright holder, if they are even in fact different people. Apple are both!
So OS X doesn't run on ARM? Why not? Because OpenDarwin doesn't? This whole article is horseshit speculation and a completely random nonsense of misunderstanding how software licensing works, who wrote and owns Darwin (Apple!) and the technical aspects involved (they've been working on the iPhone for the better part of a year and a half.. that's plenty of time to do a port to a new processor, especially given how abstracted the Darwin kernel is, XNU Apple additions and so on)
Remember this is not just a phone, but an iPod too; with a widescreen display. It fits somewhere inbetween the feature capability of the iPod Nano (storage space) and the full Video iPod (capability) with extra stuff involved.
As a successor to the iPod, considering the prices, it hits a fairly good niche. Do you need an MP3 player *and* a capable PDA-phone? Then why not have them both in one device? You may spend $300 on an iPod and $200 on a phone, now you can get them both in the same device. You also get a fairly reasonable digital camera, and it plays all your iTunes media probably a ton better than a real iPod did (nicer screen anyway).
I simply fear for fingerprints, and earprints (it turns the screen off when you hold it to your ear, but all that surface back and front..). Apple must have their shiny and glassy look. And there's no impetus on using a stylus, it's all multitouch and pinky fingers.. a lot of people will just get frustrated at how grimey it ends up looking during real use.
There's definitely a convergence market for it. MP3 phones these days usually suck as MP3 players. MP3 players as phones.. this is the first. It just so happens to be super-capable at both, without springing for a full PDA.
http://www.devrandom.us/
:D
Oh, we make the board he wants to control it with too
I never said it should be the main direction, I said it would be nice to see.
:) if it had a more embedded slant, and also reduce the system requirements needs of a lot of devices earmarked to run Linux for operation.
You could get KDE into a lot more places (maybe EmbeddedKDE to replace QTopia?
Okay so it wouldn't be any good for you as a desktop user who wants to use major chip design tools, but for the vast majority of people who buy eMachines and Dell boxes not for games, not for huge amounts of work, just for email, maybe VoIP phoning and webcam, or just browsing YouTube..
Tailoring a system for embedded use is a good way of identifying those parts which are "non-essential" or can be optimised for memory or performance on low powered devices. If it runs better on low-power devices, it will run better on high-powered ones too if the components are being worked on to that end.