Who would have paid for the privilege to run Windows 3.1 in Windows 95?
You're joking, right? I mean, it's not quite the same since you didn't actually run the Program Manager and File Manager (although the 16-bit applicaitons were still there if you wanted to!) but Windows 95 could run just about precisely like Windows 3.x in terms of applications - and application compatibility is why they're doing this thing with XP now. The number of crazy hacks Microsoft has historically put in to support, in some cases, a *single application* that did something improper which an older version allowed, is incredible. Some of them were also security risks or otherwise serious problems, and Vista was largely about security, so yeah, some of the broken stuff doesn't work anymore.
As for your dig at "innovation" you should play around with Win7's window manager a little some time. It's got some very cool tricks up its sleeve - small, but the kind of thing that gets used a lot until you find it annoying when you try it on an older system and it doesn't work (that was my experience, at least). There are lots of other changes - heck, just the inclusion of a virtualized complete OS (which you'll notice doesn't actually require Ultimate edition; it's available with Professional) is pretty new and interesting (no, MacOS "Classic" doesn't count; that was an application compatibility layer, not an actual, full operating system.
I should be surprised this got modded up, but it *is* on/.
Linux, as a kernel, does not AFAIK run significantly faster on equivalent hardware vs. NT. Some of the userspace certainly does, but some is also a lot slower - searches always take longer even though there's a lot less installed on my Linux partition (I keep it pretty clean), and without superfetch it feels that applications like WarCraft 3 (in Wine) or even Firefox take ages to start.
Viruses are a wild goose chase - they have existed since before Windows, and they will probably exist long after unless there's a drastic change in the fundamental capabilities of computers (i.e. mor ethan just an apprximate Turing machine). Security holes do still exist for *nix applicaitons and even kernels - for better or worse, I get more security patches per month on Linux than I ever do on Windows, although only occasionally are they at kernel or base library level - but even if malware authors can't xploit those, they'll fall back to the standard approach that has worked so well against Windows (itself a rather hard target these days) for the past few years: the user. There is absolutely nothing in *nix security that can protect against the dancing bunnies problem, especially if that user can get root access (although lots of damage can be done even without).
As for things you can do on Windows that you can't on Wine: well, try Exchange for starters. No other groupware solution has yet come close to the integration, feature set, and market deployment levels. Office 2007 is another; OO.o is an impressive project but they're still far behind in a number of areas (although Office 2008 does run on Mac, so that might not count). Then there are the games (wine is doing wonders here, but new stuff that doesn't work right is coming out all the time too), the Windows-only drivers (my modem *still* doesn't work in Linux, nor does the WiFi on one of my older laptops), and all the thousands of custom-written programs, only ever tested on their target machines, that businesses and other organizations have been creating for the last decade or so to run on Windows. Oh, you might also want to look at power management; with the proprietary nVidia driver (since the FOSS one is nowhere near ready yet), suspend-to-RAM in Linux quite simply does not work (on my current system, or the last two before it). This is, to put it mildly, a problem on a laptop.
Well, the 32-bit version of Win7 should still have the NTVDM (16-bit machine layer, including virtualized sound card, video, etc.) that XP (and Vista, and AFAIK all other versions of NT) have used to run DOS or Win16 applications since NT was created as 32-bit from the ground up.
Of course, x64 Windows lacks the NTVDM - I forget if it was deemed too legacy to be worth supporting or if there's a technical reason you can't switch straight froma 64-bit OS to a 16-bit application - and by the time Win7 is released, even laptops will routinely ship with 4GB or more of RAM (they're pushing 3GB already, and available with up to 8GB). Running a virtualied 32-bit Windows on x64 has been the standard soultion to this problem historically, and since Win7 is likely to be the first consumer version of Windows that will be installed primarily on x64, providing a drop-in 32-bit virtual system makes sense.
Shortly after Vista came out, I saw the same functionality appearing in the K menu on my openSuse system. It's definitely not as polished though; the search is slower, you must select the result before clicking Enter (on Vista/Win7, Enter just selects the top option), and it doesn't seem to index the equivalent of some of the Windows defaults (browser history, for example) although I'm sure I can tell it to do so.
Still, for finding a particular program in the whole application menu, it's handy. On the other hand, on Vista, the search is faster even if you know exactly where in the menu the thing you want is. Not so on Linux in my experience.
That certainly looks nice if you're always operating in a CLI environment (well, until you run any of those). It does seem that actually running a command script and selecting an option is more awkward than just tapping the WinKey and typing the first word or so of what you want. On the other hand, if you're not quite sure what you want, the list is pretty nice.
You can certainly select the locations you want indexed, and the file types. By default it indexes your Start menu, your Documents/Downloads/Music/Pictures/etc. directories, your favorites and browsing history (find a page you closed just by typing a couple words from it), your email, and programs in your PATH. I may be missing a few, but that pretty much covers it. For the Start search, Start menu entries and program names typicaly come up first in the results, while things like email are lower-ranked.
If it finds no results, or you don't see the one you're looking for, the fallback options "Search Everywhere" will search the entire heard drive (like the old Windows search) and "Search the Internet" will open your default browser and search with the default engine.
For extra customization, install Start++, an applet that adds extensions and user-extensibility to the Start search. For example, it could automatically generate a playlist from music in your collection based on keywords and start playing it, or it could display (in the Start menu) results from Dictionary.com or Weather.com or similar. This is a slightly more esoteric tool, however - convenience at the cost of slightly less usability.
Konqueror 4.2 works just fine in File Manager mode on my system (openSuse, not Ubuntu, but it shouldn't matter too much). Yes, Dolphin is the new *default* file manager, but it's quite easy to change that. Aside from minor Look and Feel updates, I can't find anything terribly different about Konq 4.2; it still has the same (customizable) view modes, still opens files happily in the embedded viewer, still browses filesystems and the web happily.
In fact, the biggest change that comes to mind is the option to have it treat middle-click upon a tab as "Close this tab" (like all other tabbed browsers) rather than "navigate to whatever is in the clipboard as a URL" (I'm sure somebody out there uses this functionality, but to me it's nothing but a pain). Otherwise, it's very much like the browser/filemanager I've been using for years, with a slightly slicker look to it.
The ones you're referring to destroyed themselves on use, yes? I'm not sure there was anything in those systems that could be termed reusable. By comparison, the LCLS is possibly the first example of what most people think of when they say "laser" - not the beam of coherent light itself, but the device which generates said beam (without vaporizing itself).
The idea has been used in Sci-Fi for a long time - I seem to recall Asimov making references to it, and probably others, in fairly golden-age stuff. It does actually work - I think it's actually been tested as a potential anti-ICBM measure - but aiming seems like it would be tricky.
Yes, for values of "glow" equal to "burst into flames." If sufficiently concentrated, it really doesn't take much energy to ignite something assuming it has a relatively low flash temperature (like wood, paper, even plastic or paint).
If you want rsync, or other *nix tools of that flavor, Interix (a UNIX-like operating environment that runs on the NT kernel using their POSIX subsystem) is great. It's only available on the higher editions of Windows, which ticks me off (not because I don't have them - being a CS student is great, you get free software - but because a lot of people can't use code I write for it) but it looks and feels pretty much like any CLI *nix environment, and has a package manager which includes rsync. Curses and ncurses programs work fine, manpages are included for everything, X client libraries are included (and there's a free X server available), and it comes with a working build toolchain (gcc based). http://suacommunity.com/
Alternatively, there's the Windows Backup tool. In Win7, you can select locations where files you want backed up are stored, file types that you want to back up, how often it runs, where the backups are stored, etc. Very handy tool. It's also integrated with the Previous Versions feature (basically just a built-in tool for automatically backing up files using free space on the disc) so you can restore specific files to specific versions from backup, or you can a full restore (such as in the case of a corrupted system or hard drive failure).
Powershell is "like bash" in many ways, but it is not bash and I'm not sure if there's a curses library for it yet (it's certainly possible - even DOS had plenty of programs with curses-style interface, for example). Aside from finally including things like aliases, its main strengths are extensibility (cmdlets, effectively shell builtins, are written using.NET languages and can be added at will - the really cool feature is that.NET objects can be piped between cmdlets, rather than just text strings like most shells) and scriptability (essentially, a better way to automate things than the old Windows Script Host). It can also be used as a CLI shell, of course, with nice features like aliases, command completion, and a help system somewhat like manpages. The ability to treat the registry as a filesystem is quite nice too. Its scripting language isn't compatible with cmd.exe, so cmd is still included for batch files and the like.
Oh, please. Do you have *ANY* idea what he's talking about? The search box at the bottom of the Vista and Win7 Start menu is not a command prompt has no syntax you need to remember allows searching by simple names/descriptions, not esoteric executable filenames filters options as you type - there's no reason to type all of "network connections" makes it easy to find the tool you want doesn't actually run anything unless you choose it from the results requires no special knowledge or complex explanation remembers what you run most often and places it first in the list of results
Aside from the fact that you use a keyboard, how is this in ANY way like a command line? Don't get me wrong, I use bash, and grep, and sed, and ssh all the time. I prefer CLI subversion to TortoiseSVN, for an example of graphical vs. CLI preference. However, for sheer convenience and dead-easy usability, the Start search is fantastic.
Somebody might actually think you know what you're talking about. It is *NOT* an automatic update. It is available through Microsoft Update, but it is an optional update which is unchecked by default. Unless you go and actually look for it, Microsoft Update will never install it. Leaving the computer on automatic updates, like most people do, or even just accepting the default options for manual update, will leave this one out.
As for your driver, it's vaguely possible that it is set to update automatically - a few driver updates are - but usually only if you don't have any driver installed for that device at all. (For example, Windows would happily give me a Intel Pro Wireless update, but - like the Windows Live Essentials option right below it - the box is unchecked and will stay that way until I personally check it. In any case, if you're computer-savvy enough to actually know the meaning of what you just posted, you're also probably sufficiently competent to click the "Hide update" option in the context menu of every listed update, which will prevent that update from installing.
Actually, tons of cash vs. tons of experience (but not tons of cash) will likely result in the experienced player winning. I once killed a faction-fitted Vagabond in a fairly standard PvP Hurricane... the guy must have either bought ISK or just bought the character, because he had NO idea how to fly such a ship.
For those who don't know the game, a fully fitted and insured Hurricane might cost about 60M (of which 45M is hull and insurance). The hull of the Vagabond I killed was worth over 100M, and it had roughly half a billion in fittings (the hull itself does nothing except fly and provide basic defense; all weapons and such are extra). In other words, I killed a very fancy PvP ship worth at least 10 times the value of my ship... just because I knew how to use my ship (I wasn't that experienced either, just much more than the other guy).
There's actually a logical solution to the issue of those with lots of time but little money, and those with lots of money but little time, in the same game. In EVE Online, players can purchase game time (PLEX - Pilot License EXtensions) for real-world money, for about the cost of the standard subscription. They can then sell these in game on an open market - monitored by CCP to prevent abuse, but with pricing based on supply and demand - to other pilots (i.e. the ones with the time to make in game cash). Everybody wins - if you like playing a lot, you never have to pay for your subscription. If you want to be able to sign on at any time, buy a battleship, get into a huge fight, log off at the end, and do it again whenever you get the chance... well, you can do that too, and the game sanctions it.
The problem isn't the blocking of bi-di characters (or other wacky Unicode that breaks stuff). The problem is the blocking of ALL non-ASCII, even perfectly valid things like currency symbols, accented letters, and similar helpful little characters.
To be fair, a dual-core 400MHz smartphone is nowhere close to an "average phone today" like the GP mentioned. That said, I'm pretty sure the CPU in my basic flip phone (capable of some voice recognition and a really crappy web browser) still beats the pants off the 386SX16 I had in '92, so your point stands.
You may have heard of this little family of software products called Microsoft Windows. Up until version 6, this proprietary software used substantial amounts of BSD-derived code - specifically in the networking stack.
You might also have heard of Apple's OS X. It's a bit rarer than Windows but still reasonably common. It's entire kernel (much of its core, in fact) was derived from BSD as well (they chose to release the source for some portions of some of their platforms, but it's still proprietary code).
While I agree that copyleft licenses aren't needed for everything, and while I certainly prefer LGPL over GPL (although I can understand the motivation behind GPL quite well), permissive licenses like BSD and MS-PL can and do get abused. Copyleft exists for a reason, and it's not obsolete yet. In fact, I doubt it ever really will be.
Outside the tech world, nobody cares in the least (it's not even like they open-sourced a end-user program - this is a development framework). Inside the tech world, everybody already knows about the GPL and Linux.
Inherently, the fact that something is F/OSS really only matters to developers. End users may like its features, or security, or the fact that it's free of charge. However, none of those are exclusive to the F/OSS world. Even things like the ability to extend the product are available in proprietary software too; Visual Studio and Internet Explorer both have tons of available add-ons, as an example.
Umm, why not? Seriously, you just drove a Hummer* through your own argument. It is not something that can only be used in conjunction with a commercial product. It can run on another OS. It doesn't require the.NET framework.
Is your anti-MS bias REALLY that strong that you could KNOW you're wrong and post anyhow?**
* In keeping with/. analogy guidelines, if anybody has a bigger car to suggest let me know. ** Really, I'm not new here, I promise.
Not really. They can create a new license (call it MS-PL v2 for convenience). However, they can't change the MS-PL itself without needing to re-certify it; it would automatically be a new version.
Additionally, they can't do any kind of "retroactive" change to the licensing. They can state that all of their MS-PL code is now MS-PL v2 code, and not available to anybody with brown eyes if they want. However, the old code is already distributed; they can't revoke that.
More specifically and importantly, they can't change the license on a derivative work done by something else. If that person acquired the code under MS-PL as it now stands, and releases a derivative work under MS-PL (as the license requires), there's not a thing Microsoft can legally do about it.
Close, although in this case MS-PL does NOT "go in"; it is incompatible with GPL due to the explicit requirement that redistributed source must be licensed under the MS-PL.
While the JRE did include JIT capability roughly a decade ago, it is only in the last few years that its performance has approached that of native code. A few more optimizations were added with each release, but java 1.5.0 was still slower on many benchmarks than.NET 1.1.
Furthermore, the GP's point was that.NET is slow, as Java used to be (and is still generally perceived to be). (More specifically, the implication was that.NET combines the worst of native code and Java.) While the number of runtime checks performed means that any substantial program written in.NET will probably be slightly slower than the equivalent native program in which such checks are omitted,.NET was never slow in the way that most programmers still associate with Java.
That said, I was being unfair to modern versions of Java. A more correct response might have been to point out that modern versions of both environments are quite fast, with a good JIT offering some advantages that native binaries typically lack.
First, they ARE providing something to you: a world-wide, non-exclusive, royalty-free patent license. They can't sue you over patents in their code base; they already gave you a license to them.
Second, if you bring a patent claim against a contributor over code covered by the MS-PL (not just any code they wrote, as you implied) then you don't lose all rights, you only lose the royalty-free patent license from that specific contributor.
Example: Microsoft releases some code (call it code-base A) under MS-PL. It contains patented algorithm X. You take A and extend it. Your extension (code-base B) contains an improvement on X, which you have patented. Call this improved version Y. If Microsoft sues you over Y (which is basically a better X) then they lose the right to use Y, meaning that if Y is upheld they would have to license it from you. Furthermore, even if they win the case and the patent on Y is invalidated, X can still be used free of charge; they can't revoke your license to use it.
This seems a fair way to handle software patents in open-source software; a sort of copyleft scheme applied to patent right rather than copyright.
Mind you, IANAL, but the terminology seems pretty clear.
Slashdot Mirror
Who would have paid for the privilege to run Windows 3.1 in Windows 95?
You're joking, right? I mean, it's not quite the same since you didn't actually run the Program Manager and File Manager (although the 16-bit applicaitons were still there if you wanted to!) but Windows 95 could run just about precisely like Windows 3.x in terms of applications - and application compatibility is why they're doing this thing with XP now. The number of crazy hacks Microsoft has historically put in to support, in some cases, a *single application* that did something improper which an older version allowed, is incredible. Some of them were also security risks or otherwise serious problems, and Vista was largely about security, so yeah, some of the broken stuff doesn't work anymore.
As for your dig at "innovation" you should play around with Win7's window manager a little some time. It's got some very cool tricks up its sleeve - small, but the kind of thing that gets used a lot until you find it annoying when you try it on an older system and it doesn't work (that was my experience, at least). There are lots of other changes - heck, just the inclusion of a virtualized complete OS (which you'll notice doesn't actually require Ultimate edition; it's available with Professional) is pretty new and interesting (no, MacOS "Classic" doesn't count; that was an application compatibility layer, not an actual, full operating system.
I should be surprised this got modded up, but it *is* on /.
Linux, as a kernel, does not AFAIK run significantly faster on equivalent hardware vs. NT. Some of the userspace certainly does, but some is also a lot slower - searches always take longer even though there's a lot less installed on my Linux partition (I keep it pretty clean), and without superfetch it feels that applications like WarCraft 3 (in Wine) or even Firefox take ages to start.
Viruses are a wild goose chase - they have existed since before Windows, and they will probably exist long after unless there's a drastic change in the fundamental capabilities of computers (i.e. mor ethan just an apprximate Turing machine). Security holes do still exist for *nix applicaitons and even kernels - for better or worse, I get more security patches per month on Linux than I ever do on Windows, although only occasionally are they at kernel or base library level - but even if malware authors can't xploit those, they'll fall back to the standard approach that has worked so well against Windows (itself a rather hard target these days) for the past few years: the user. There is absolutely nothing in *nix security that can protect against the dancing bunnies problem, especially if that user can get root access (although lots of damage can be done even without).
As for things you can do on Windows that you can't on Wine: well, try Exchange for starters. No other groupware solution has yet come close to the integration, feature set, and market deployment levels. Office 2007 is another; OO.o is an impressive project but they're still far behind in a number of areas (although Office 2008 does run on Mac, so that might not count). Then there are the games (wine is doing wonders here, but new stuff that doesn't work right is coming out all the time too), the Windows-only drivers (my modem *still* doesn't work in Linux, nor does the WiFi on one of my older laptops), and all the thousands of custom-written programs, only ever tested on their target machines, that businesses and other organizations have been creating for the last decade or so to run on Windows. Oh, you might also want to look at power management; with the proprietary nVidia driver (since the FOSS one is nowhere near ready yet), suspend-to-RAM in Linux quite simply does not work (on my current system, or the last two before it). This is, to put it mildly, a problem on a laptop.
Well, the 32-bit version of Win7 should still have the NTVDM (16-bit machine layer, including virtualized sound card, video, etc.) that XP (and Vista, and AFAIK all other versions of NT) have used to run DOS or Win16 applications since NT was created as 32-bit from the ground up.
Of course, x64 Windows lacks the NTVDM - I forget if it was deemed too legacy to be worth supporting or if there's a technical reason you can't switch straight froma 64-bit OS to a 16-bit application - and by the time Win7 is released, even laptops will routinely ship with 4GB or more of RAM (they're pushing 3GB already, and available with up to 8GB). Running a virtualied 32-bit Windows on x64 has been the standard soultion to this problem historically, and since Win7 is likely to be the first consumer version of Windows that will be installed primarily on x64, providing a drop-in 32-bit virtual system makes sense.
Shortly after Vista came out, I saw the same functionality appearing in the K menu on my openSuse system. It's definitely not as polished though; the search is slower, you must select the result before clicking Enter (on Vista/Win7, Enter just selects the top option), and it doesn't seem to index the equivalent of some of the Windows defaults (browser history, for example) although I'm sure I can tell it to do so.
Still, for finding a particular program in the whole application menu, it's handy. On the other hand, on Vista, the search is faster even if you know exactly where in the menu the thing you want is. Not so on Linux in my experience.
That certainly looks nice if you're always operating in a CLI environment (well, until you run any of those). It does seem that actually running a command script and selecting an option is more awkward than just tapping the WinKey and typing the first word or so of what you want. On the other hand, if you're not quite sure what you want, the list is pretty nice.
You can certainly select the locations you want indexed, and the file types. By default it indexes your Start menu, your Documents/Downloads/Music/Pictures/etc. directories, your favorites and browsing history (find a page you closed just by typing a couple words from it), your email, and programs in your PATH. I may be missing a few, but that pretty much covers it. For the Start search, Start menu entries and program names typicaly come up first in the results, while things like email are lower-ranked.
If it finds no results, or you don't see the one you're looking for, the fallback options "Search Everywhere" will search the entire heard drive (like the old Windows search) and "Search the Internet" will open your default browser and search with the default engine.
For extra customization, install Start++, an applet that adds extensions and user-extensibility to the Start search. For example, it could automatically generate a playlist from music in your collection based on keywords and start playing it, or it could display (in the Start menu) results from Dictionary.com or Weather.com or similar. This is a slightly more esoteric tool, however - convenience at the cost of slightly less usability.
Konqueror 4.2 works just fine in File Manager mode on my system (openSuse, not Ubuntu, but it shouldn't matter too much). Yes, Dolphin is the new *default* file manager, but it's quite easy to change that. Aside from minor Look and Feel updates, I can't find anything terribly different about Konq 4.2; it still has the same (customizable) view modes, still opens files happily in the embedded viewer, still browses filesystems and the web happily.
In fact, the biggest change that comes to mind is the option to have it treat middle-click upon a tab as "Close this tab" (like all other tabbed browsers) rather than "navigate to whatever is in the clipboard as a URL" (I'm sure somebody out there uses this functionality, but to me it's nothing but a pain). Otherwise, it's very much like the browser/filemanager I've been using for years, with a slightly slicker look to it.
The ones you're referring to destroyed themselves on use, yes? I'm not sure there was anything in those systems that could be termed reusable. By comparison, the LCLS is possibly the first example of what most people think of when they say "laser" - not the beam of coherent light itself, but the device which generates said beam (without vaporizing itself).
The idea has been used in Sci-Fi for a long time - I seem to recall Asimov making references to it, and probably others, in fairly golden-age stuff. It does actually work - I think it's actually been tested as a potential anti-ICBM measure - but aiming seems like it would be tricky.
Yes, for values of "glow" equal to "burst into flames." If sufficiently concentrated, it really doesn't take much energy to ignite something assuming it has a relatively low flash temperature (like wood, paper, even plastic or paint).
If you want rsync, or other *nix tools of that flavor, Interix (a UNIX-like operating environment that runs on the NT kernel using their POSIX subsystem) is great. It's only available on the higher editions of Windows, which ticks me off (not because I don't have them - being a CS student is great, you get free software - but because a lot of people can't use code I write for it) but it looks and feels pretty much like any CLI *nix environment, and has a package manager which includes rsync. Curses and ncurses programs work fine, manpages are included for everything, X client libraries are included (and there's a free X server available), and it comes with a working build toolchain (gcc based). http://suacommunity.com/
Alternatively, there's the Windows Backup tool. In Win7, you can select locations where files you want backed up are stored, file types that you want to back up, how often it runs, where the backups are stored, etc. Very handy tool. It's also integrated with the Previous Versions feature (basically just a built-in tool for automatically backing up files using free space on the disc) so you can restore specific files to specific versions from backup, or you can a full restore (such as in the case of a corrupted system or hard drive failure).
Powershell is "like bash" in many ways, but it is not bash and I'm not sure if there's a curses library for it yet (it's certainly possible - even DOS had plenty of programs with curses-style interface, for example). Aside from finally including things like aliases, its main strengths are extensibility (cmdlets, effectively shell builtins, are written using .NET languages and can be added at will - the really cool feature is that .NET objects can be piped between cmdlets, rather than just text strings like most shells) and scriptability (essentially, a better way to automate things than the old Windows Script Host). It can also be used as a CLI shell, of course, with nice features like aliases, command completion, and a help system somewhat like manpages. The ability to treat the registry as a filesystem is quite nice too. Its scripting language isn't compatible with cmd.exe, so cmd is still included for batch files and the like.
Oh, please. Do you have *ANY* idea what he's talking about? The search box at the bottom of the Vista and Win7 Start menu
is not a command prompt
has no syntax you need to remember
allows searching by simple names/descriptions, not esoteric executable filenames
filters options as you type - there's no reason to type all of "network connections"
makes it easy to find the tool you want
doesn't actually run anything unless you choose it from the results
requires no special knowledge or complex explanation
remembers what you run most often and places it first in the list of results
Aside from the fact that you use a keyboard, how is this in ANY way like a command line? Don't get me wrong, I use bash, and grep, and sed, and ssh all the time. I prefer CLI subversion to TortoiseSVN, for an example of graphical vs. CLI preference. However, for sheer convenience and dead-easy usability, the Start search is fantastic.
Somebody might actually think you know what you're talking about.
It is *NOT* an automatic update. It is available through Microsoft Update, but it is an optional update which is unchecked by default. Unless you go and actually look for it, Microsoft Update will never install it. Leaving the computer on automatic updates, like most people do, or even just accepting the default options for manual update, will leave this one out.
As for your driver, it's vaguely possible that it is set to update automatically - a few driver updates are - but usually only if you don't have any driver installed for that device at all. (For example, Windows would happily give me a Intel Pro Wireless update, but - like the Windows Live Essentials option right below it - the box is unchecked and will stay that way until I personally check it. In any case, if you're computer-savvy enough to actually know the meaning of what you just posted, you're also probably sufficiently competent to click the "Hide update" option in the context menu of every listed update, which will prevent that update from installing.
Forget your brains, all they want to do is eat your wallet...
</joke>
Actually, tons of cash vs. tons of experience (but not tons of cash) will likely result in the experienced player winning. I once killed a faction-fitted Vagabond in a fairly standard PvP Hurricane... the guy must have either bought ISK or just bought the character, because he had NO idea how to fly such a ship.
For those who don't know the game, a fully fitted and insured Hurricane might cost about 60M (of which 45M is hull and insurance). The hull of the Vagabond I killed was worth over 100M, and it had roughly half a billion in fittings (the hull itself does nothing except fly and provide basic defense; all weapons and such are extra). In other words, I killed a very fancy PvP ship worth at least 10 times the value of my ship... just because I knew how to use my ship (I wasn't that experienced either, just much more than the other guy).
There's actually a logical solution to the issue of those with lots of time but little money, and those with lots of money but little time, in the same game. In EVE Online, players can purchase game time (PLEX - Pilot License EXtensions) for real-world money, for about the cost of the standard subscription. They can then sell these in game on an open market - monitored by CCP to prevent abuse, but with pricing based on supply and demand - to other pilots (i.e. the ones with the time to make in game cash). Everybody wins - if you like playing a lot, you never have to pay for your subscription. If you want to be able to sign on at any time, buy a battleship, get into a huge fight, log off at the end, and do it again whenever you get the chance... well, you can do that too, and the game sanctions it.
The problem isn't the blocking of bi-di characters (or other wacky Unicode that breaks stuff). The problem is the blocking of ALL non-ASCII, even perfectly valid things like currency symbols, accented letters, and similar helpful little characters.
To be fair, a dual-core 400MHz smartphone is nowhere close to an "average phone today" like the GP mentioned. That said, I'm pretty sure the CPU in my basic flip phone (capable of some voice recognition and a really crappy web browser) still beats the pants off the 386SX16 I had in '92, so your point stands.
Your sig is amusing here.
You may have heard of this little family of software products called Microsoft Windows. Up until version 6, this proprietary software used substantial amounts of BSD-derived code - specifically in the networking stack.
You might also have heard of Apple's OS X. It's a bit rarer than Windows but still reasonably common. It's entire kernel (much of its core, in fact) was derived from BSD as well (they chose to release the source for some portions of some of their platforms, but it's still proprietary code).
While I agree that copyleft licenses aren't needed for everything, and while I certainly prefer LGPL over GPL (although I can understand the motivation behind GPL quite well), permissive licenses like BSD and MS-PL can and do get abused. Copyleft exists for a reason, and it's not obsolete yet. In fact, I doubt it ever really will be.
Outside the tech world, nobody cares in the least (it's not even like they open-sourced a end-user program - this is a development framework). Inside the tech world, everybody already knows about the GPL and Linux.
Inherently, the fact that something is F/OSS really only matters to developers. End users may like its features, or security, or the fact that it's free of charge. However, none of those are exclusive to the F/OSS world. Even things like the ability to extend the product are available in proprietary software too; Visual Studio and Internet Explorer both have tons of available add-ons, as an example.
And yes I know it could work with Mono, but why?
Umm, why not? Seriously, you just drove a Hummer* through your own argument. It is not something that can only be used in conjunction with a commercial product. It can run on another OS. It doesn't require the .NET framework.
Is your anti-MS bias REALLY that strong that you could KNOW you're wrong and post anyhow?**
* In keeping with /. analogy guidelines, if anybody has a bigger car to suggest let me know.
** Really, I'm not new here, I promise.
Not really. They can create a new license (call it MS-PL v2 for convenience). However, they can't change the MS-PL itself without needing to re-certify it; it would automatically be a new version.
Additionally, they can't do any kind of "retroactive" change to the licensing. They can state that all of their MS-PL code is now MS-PL v2 code, and not available to anybody with brown eyes if they want. However, the old code is already distributed; they can't revoke that.
More specifically and importantly, they can't change the license on a derivative work done by something else. If that person acquired the code under MS-PL as it now stands, and releases a derivative work under MS-PL (as the license requires), there's not a thing Microsoft can legally do about it.
Close, although in this case MS-PL does NOT "go in"; it is incompatible with GPL due to the explicit requirement that redistributed source must be licensed under the MS-PL.
While the JRE did include JIT capability roughly a decade ago, it is only in the last few years that its performance has approached that of native code. A few more optimizations were added with each release, but java 1.5.0 was still slower on many benchmarks than .NET 1.1.
Furthermore, the GP's point was that .NET is slow, as Java used to be (and is still generally perceived to be). (More specifically, the implication was that .NET combines the worst of native code and Java.) While the number of runtime checks performed means that any substantial program written in .NET will probably be slightly slower than the equivalent native program in which such checks are omitted, .NET was never slow in the way that most programmers still associate with Java.
That said, I was being unfair to modern versions of Java. A more correct response might have been to point out that modern versions of both environments are quite fast, with a good JIT offering some advantages that native binaries typically lack.
This is untrue.
First, they ARE providing something to you: a world-wide, non-exclusive, royalty-free patent license. They can't sue you over patents in their code base; they already gave you a license to them.
Second, if you bring a patent claim against a contributor over code covered by the MS-PL (not just any code they wrote, as you implied) then you don't lose all rights, you only lose the royalty-free patent license from that specific contributor.
Example: Microsoft releases some code (call it code-base A) under MS-PL. It contains patented algorithm X.
You take A and extend it. Your extension (code-base B) contains an improvement on X, which you have patented. Call this improved version Y.
If Microsoft sues you over Y (which is basically a better X) then they lose the right to use Y, meaning that if Y is upheld they would have to license it from you. Furthermore, even if they win the case and the patent on Y is invalidated, X can still be used free of charge; they can't revoke your license to use it.
This seems a fair way to handle software patents in open-source software; a sort of copyleft scheme applied to patent right rather than copyright.
Mind you, IANAL, but the terminology seems pretty clear.