Nonsense, the election was decided long before the polls closed. After Ontario's polls were closed, the winner was decided, the rest of the country be damned. Manitoba, Saskatchewan, Alberta, British Columbia, the Northwest Territories and the Yukon have little influence in the polls when it comes to the federal elections. That's why we have this draconian law prohibiting broadcasting election results until after the polls close in the viewer area, or nationally.
There's an important difference between get rich schemes that fail and lives lost. These days, it seems that if you play with fire against everyone's recommendations, you can still sue someone for giving you the fire that you burnt yourself on. Laywers may be needed for certain things, but do they need to become a way of life? It seems that more and more they are becoming so. We've got people sueing McDonalds because they were stupid enough to pour hot coffee on themselves, people going after the entertainment industry because parents are screwing up raising their kids. It's SICK.
While WMA will often do a better job than straight MP3, it tends to be less consistent than MP3 is. While a 128kbps MP3 may have lost many of the high frequency tones, WMA sometimes seems to have trouble and generates tinny metallic or underwater sounding sections when the tones gets too complex for it to compress well.
Also keep in mind there's a much higher differential between MP3 encoders. Certain encoders, in particular early Xing encoders and those based on the ISO source code, tended to generate very poor quality MP3s, with a great deal of lost information from the MP3. The Microsoft WMA encoder, on the other hand, is only developed by Microsoft, since the standards are not available to anyone who doesn't want to pay the license fees.
Of course, this doesn't stop anyone from intercepting the decrypted data leaving the hard disk -- but that is akin to recording your favorite DVD onto VHS (or mpeg-4, or whatever). Plug Intel's new encrypted monitor spec in, and the data won't be decrypted until it gets to the monitor...
Yes, I am afraid you can securely encrypt data. They know how, and they will do it eventually. Until then, we need to educate. Just like dongles of yesteryear, but without the hassle of plugging anything in.
Still easy enough to break -- write a program that acts as a debugger and captures the data from RAM itself. Even if you have to capture a byte at a time, it's virtually impossible to protect the decryption process that you cannot break it. Every hour that one takes in designing a copy protection software technique, delays breaking it by a skilled cracked by 1-5 minutes. I don't believe there's any form of copy protection out there that hasn't been broken, nor do I believe there ever will be. There's enough possible flaws in this that even if most were able to be closed, there'd still be methods for getting at the data. This will only protect against the least determined pirates, the rest will find a work-around. Remember that even DeCSS wasn't the first DVD decrypting software -- most just used the DVD software for decrypting and either captured the decrypted frames from video RAM or via DirectShow, or captured the decrypted data from RAM as the DVD player ran.
You forget that Microsoft is opposed to the entire CPRM idea. Hopefully they'll remain opposed to this and any/all derrivative technology that puts copy protection in hardware. They might end up being one of the biggest allys in the war against this technology.
On the other hand, this technology was designed and licensed by the LMI, LLC, the same company that designed and licensed the CSS algorithm. Anyway, if you look at it, no half-decent OS is going to give apps direct access to your HDD, rather you'll likely have to do this through APIs that delve through all the driver levels to get to the device. For Windows, this could be broken as easily as writing a VxD or kernel-mode.SYS driver that patches into the standard ATA driver and returns bogus (known, repeatable) data for the encryption keys.
People forget that people had found ways to get unencrypted.VOB files long before DeCSS was designed -- there are many ways of doing it, everything from grabbing data in memory after decryption to the old low-tech standbys of hooking video out to video in. Don't underestimate the ability for people to find ways around these technological control measures. Macrovision only stops the least determined of pirates. CD-R audio with it's serialization and high price never caught on (and likely never will). CSS was broken long before DeCSS was even thought of.
The first problem is you don't rent word every time you use it. That's just an outright lie.
I believe the point he was making that the.NET version of Office 10 is going to be subscription based -- you want it, you pay the per-month/year/whatever fee for it. Microsoft's trying to avoid the situation they're in now with many companies still running Win95 and Office 95 with no intention of upgrading now, or perhaps ever. This way they get to extort money out of any company that adopts this on a subscription basis. It's the same basic idea as some of Microsoft's more recent licensing schemes, that involves 'renting' of the software instead of an outright purchase.
This shouldn't be too hard since vs.net is fantastic. It has all the nice things from Delphi from a company powerful enough to get it done. Delphi always had a "not quite finished" feel to it. Its version of intelisense made me cringe.
Delphi was too late to the Windows RAD party. It never managed to produce a critical mass or enough income for Borland/Imprise to ever make it a completely finished product. Part of the reason it continually failed to catch on was the fact it still to this day, carries legacy from the Borland products of olde, such as Turbo C/Pascal, particularily in the methods it requires to get things done. Few will debate that in the days of DOS, Borland's compilers were for the most part, king. Borland's Turbo C generally beat the pants off Microsoft's Quick C, however, Borland had two strikes against them for Windows compilers... One was the fact that Microsoft continually extends the Win32 API at their sole discretion, and guess what, Visual C++ is always the first to get the new capabilities. Second was that Borland got off to a slow start with their Windows products.
Did it ever occour to you that some people understand the limits of their driving skill and drive at an apropriate speed? Not everyone has the stupid obsession of driving at the fastest possible speed they think they can get away with.
heh. I can think of one good example of that. Every year in Winnipeg during the first few snowfalls, the roads end up EXTREMELY slippery, and everyone goes around 30KM/h instead of the 60-80KM/h that's normal, but there's always the moron in the 4x4 that thinks they can drive and stop on the slush and ice going full tilt down the road, and they end up either in the ditch (most common), or plowing into someone's back end. I always laugh when I drive down the highway, and see 4x4 trucks, jeeps and SUVs all in the ditches. I know my RWD truck has little or no traction on the ice and snow, and usually match my speed to my rather conservative estimation of the road condition. If you want to pass me, go right ahead, just don't pass me, pull in front of me and then slow down -- I may not be able to stop in time...
Calling someone a communist seems to almost be an instinct to some Americans. They so fear and therefore hate any economic system that is not their own that these words have become a generic insult. In that world, the world revolves around the almighty buck, and it becomes hard to believe that anyone would willingly donate to charity if it didn't involve some kind of reimbursement by the government in the form of a tax break. From that perspective, anyone who would give away hours of work to everyone without demanding any kind of immediate reimbursement for their efforts seems abnormal. I'm sure there are people out there who truly believe that Linux and communism must be connected in some way. Just as I'm sure there are people who are convinced that corporations are just looking out for the public's best interest when they lobby government to pass stupid laws.
Many people tend to cover their eyes, afraid they might see something they don't like, afraid they might know something they'd rather not know, afraid they might change and cease being the same person they were before...
I'm a firm believer that patent (and copyright) lifetimes should be linked directly to the innovation happening in that field. Hopefully promoting lengths that allow companies to recoup their costs and make some profit, but still make it possible for the works to enter the public domain within a reasonable timeframe. Unfortunately, as long as governments are firmly within the grasp of corporate owned special interest groups, the age of insane patents and eternal copyright may not end anytime soon.
From a business perspective, he must be smart. However, most people overstate his programming and technical skills, since for the most part he didn't create the products his company sold, even in the early days, but rather just supervised their design and helped create the market for them. He had enough contacts and enough business sense to make lucrative deals in the early days that pretty much assured him the position he has now.
Not necessarily. Some 'liberal' philosophers justify affirmative action on the basis that at a collective level, it goes some way towards redressing injustices - even though - and they actually admit this - it is unfair to some/many individuals. This sounds bogus to me.
Perhaps in an ideal world, 'affirmative action' would remedy such problems. However, the world is far from ideal. The reality is that affirmative action programs and racial quotas, especially when instigated by a lawsuit do not breed a workplace that anyone hired under such a program would find beneficial. It also doesn't tackle the core reason for this, but rather a symptom. Personally, if I were a member of a minority and knew I was hired solely based on such a policy, I would be insulted.
The problem is e2fsck ALWAYS runs when a bad shutdown occured. Win2K and NT sometimes seem to have this delusion that the journalling worked and they don't need to run autochk, but you usually end up rebooting again so that chkdsk/f can run. The time they take to check the partition also directly varies depending on the number of files and how fragmented those files are. Heavily fragmented partitions tend to take a lot longer to check for errors than unfragmented partitions. And despite what Microsoft's press releases might say, journalling (or rather metadata journalling, which is the most common variety) is a rather poor defense against data corruption since it only protects the structure of the file system, not the contents of it, and it's easy to forget that in the case of a bad shutdown, especially ones that involve the crash of a system driver, even the metadata journal can't always be trusted.
Even migrating from Windows NT 4.0 to Windows 2000 is far from an easy task. Incompatibilities ranging from incompatible ACPI BIOS revisions, to hardware incompatibilities to software incompatibilties. Migrating servers is even worse, because upgrading domain controllers means upgrading to Active Directory. Windows 2000 servers cannot act as BDCs, only member servers or Active Directory domain controllers, which means if you want to upgrade a BDC server to W2K, you first have to upgrade the PDC, and do all the planning that entails. To make this worse, if you want to say upgrade a Netware/NT mixed network and migrate the data from the Netware server, things get worse. Forcing you to purchase a copy of the Netware Services for Windows 5.0 package if migrating file permissions is even a little important to you, create or fix a pre-generated account mapping file. Then you get to actually migrate the data and find that it made a mess of the file permissions by preventing file permissions from being inherited where it would otherwise be possible. And after that's all done, you get to try and rewrite the login scripts using the rather stupid cmd.exe/command.com batch scripting language. I suffered through this, and had various bad things happen, including the deletion of all the Windows NT accounts (including the Administrator and Guest accounts).
Don't make me laugh. Windows 2000 Pro and Direct3D is a lot more solid and stable than Linux and accelerated OpenGL...
I wasn't talking about Linux in particular. OpenGL on most UNIX workstations (Sun, SGI, etc) is generally very stable. Linux's implementation of glx is very recent and hasn't matured to the point where it's stable. Then there's also issues with the fact that DirectX acts very different at times on NT/2000 than it does on Windows 9x, and often requires more than just a simple debug session.
Remember what people were saying about Unreal when it was released? "Oh! What a beautiful slideshow!" May I mention that they migrated to DirectX in newer versions of the engine and that the OpenGL implementation is half as fast on my NVIDIA machine even though NVIDIA's implementation of OpenGL is considered one of the faster ones?
And do you remember how bad the DirectX versions of Unreal were? Aside from the graphical glitches, it was crashy, often slow or worse. Unreal was slow on many computers due to the fact that some OpenGL drivers were uninteligent enough to drop down to software rendering for the most useless features. Well-designed drivers are often the difference between an enjoyable game and disappointment. My point is not that OpenGL is better than DirectX, but rather it's not inferrior to it either. And more to the point, OpenGL and DirectX (D3D) are more alike than they are different, aside from Microsoft's insistance on COM objects. Nearly every OpenGL function call has an equiv D3D COM method, and vice versa. Only a very few aren't in the other.
The problem with having both is we have a lot of blind advocacy of whatever product they prefer. Unable or unwilling to admit the other group has a better idea or grasp on how to do something, they let pride dictate the nature of the technology. This unfortunately happens between just about any competeing 'groups'. No matter if it's KDE and Gnome, the BSDs, BSD and Linux, or Perl versus the rest of the programming language universe.
The best example of this is back from Feburary when there was an announcement on the KDE pages that said they were going to collaborate on the object sharing framework. A few months later, all mention of this vanished and KParts appeared as KDE's object model and Bonobo on Gnome's end. And guess what? They don't work together, and in fact the work that would be required to get them to work together would be tremendous.
It's not a big challenge to code under crossplatform newbie APIs like OpenGL... Unless you want to make it run at a decent speed.
Now there is an API that is fairly usable without being Windows-only and complex. Glide. Oh, what do you say? Glide doesn't work on your card? Tough luck, that WAS the standard for years.
DirectX is the best API available, it really makes a difference in what quality of content can be provided. Have a guess of why Linux OpenGL versions of games are usually pretty late compared to DirectX? It's so that top-notch computers can run them fast enough that they can't be called slideshows...
Hate to break this to you, but OpenGL is not a newbie API nor is it any slower (or faster) than DirectX. You forget that in many of the early ages of DirectX, the Direct3D component sucked so bad that almost no one WANTED to use it. Today things are somewhat better since the APIs have evolved to the point where they're usable, and since Microsoft promotes DirectX, part of the 'logo'ed driver qualifications is passing the conformance tests for Direct3D, etc. However, OpenGL has been available on Windows for nearly as long as DirectX has -- Windows NT 4.0 shipped with an OpenGL DLL, although few vendors shipped a corresponding ICD with their driver sets.
OpenGL has a lot going for it. One is that it's a mature API and was designed by a company that is responsible for a great deal of the graphics technology that we've seen on TV and now are happening on our computers -- SGI. DirectX, or rather Direct3D in particular, was a late comer to the party and has had to play the catch-up game for the past couple years. Another advantage is the cross-platform capabilities, which also means you may be able to do portions of the development work on platforms a little more stable than the regular Windows 9x systems the game will likely be run on. Also don't forget that there have been large numbers of games using OpenGL... Quake II is a good early example, some later ones include Unreal and Half-Life also used OpenGL engines, and there wasn't any slowdown. I believe Quake 3 Arena also uses OpenGL, but I don't really play it.
But most of this is all beside the point. It's fairly simple to abstract the game and engine logic from the graphics back-end. For the most part these days, OpenGL and Direct3D are more alike than they are different, and most of the reason we have unportable games is developer laziness and development schedules that run far beyond what they should be forcing the developers to do a half-assed job and deliver an unfinished product.
Linux has other problems regarding games, aside from this silly Direct3D/OpenGL nonsense, the least of which is the rather poor driver support (we've got what? nvidia (through binary-only X servers/drivers), ATI, Matrox, and maybe one or two other manufacturers?), XFree86 4 just finally getting usable for regular use, and still rather poor support for advanced features of current sound cards (Yes, my SoundBlaster Live works in Linux, but still no EAX or 3D audio).
I can think of only one reason why a developer would want to always choose DirectX over OpenGL -- driver support for DirectX is almost always better than driver support for OpenGL on Microsoft Windows.
I think the original poster is a little confused. Word 97 and 2000 both support tracking changes to a document, however it's an optional feature that must be turned on to be useful. Despite this, however, it's known that Word documents can carry all sorts of data that has been deleted, on the clipboard, or from apps running on the same machine. Running a 'strings' command on a word document will often reveal data that was never in the document, was removed from the document at some point and other tidbits of information.
Actually, on most Windows systems, you'll hit the 255 character filename/path limitation, and it'll die promptly, stuck in a loop where it can no longer change directories or make new ones under the current directory. XCOPY on windows/dos used to be stupid enough to do a cyclic copy which would eventually fill all available directory entries and clusters.
I think some UNIX-style operating systems have integrated protection from this sort of simplistic DoS attack, as well, aside from disk space/inode quotas.
The only thing they could do is license the diffs as public domain. They can't relicense the kernel as public domain since it's already licensed under the GNU GPL and they have no authorization to relicense it under different terms. Government agencies have to play by the same rules as everyone else -- the GPL grants no special permissions to military agencies.
It's easy to forget that a large portion of the world doesn't have cheap unlimited access to the internet, and in fact in many places in the world, you end up paying per-minute for any phone calls. And even within the US and Canada, there are many regions where you have no choice but to pay per-hour for your internet service.
E-mail spam in many ways mimics the so-called junk faxes that have been outlawed in most places. Despite this fact, companies still send these junk faxes. I believe that companies and individuals that send these junk e-mails should be treated in the exact same way people who commit the same crime with facsimile machines are. Both are unauthorized use of my resources, which I pay for. I have much less of a problem with junk mail, since I don't have to pay to receive it. The sender pays, and I trash it. I also have a problem with telemarketers because they waste a considerable amount of my time, tie up my telephone and call at inappropriate times.
I do, however, agree that the MAPS RBL is a rather ineffective measure, and certainly no better than any of the censorware filtering software that we've all grown to hate. The MAPS RBL blocks as much or more legitimate traffic than it stops, and the real solution is to either prevent anyone from sending unsolicited commercial e-mail or to require that all unsolicited spam be marked in a header or on the subject line in a standardized fashion so that the recipient can always filter if he or she chooses to do so. I would certainly prefer the first measure come into being for the simple reason that I don't see why trash e-mails need to be sent -- there are better, less intrusive ways of advertising, the easiest being putting up a website and letting a few search engines spider it.
And to all the spammers out there: For the millionth time, no I don't want a non-accredited diploma or degree, to make money fast, mass-mailer software, e-mail list CDs, or viagra.
On the other hand, this is an artifact of how human sight works. If you look at RGB(255,0,0) versus RGB(0,0,255) and RGB(0,255,0), the blue box tends to be the darkest, the red is next, and green is the brightest, therefore it stands to reason you'd be able to distinguish more shades of green than blue, if for no other reason than the percieved intensity difference. Intensity itself isn't even percieved linearly by the eye to make things worse. Even if we were to go to 16-bits per colour, or 48bpp, it's still conceivable that there would be some colours that the human could noticibly percieve but the monitor could not generate. However, for the vast majority of people, 24bpp is 'good enough' and achieves everything they'd want out of a video card. I could, however, see 36bpp becoming an option on some video cards for people who require very accurate colour matching for things like publishing. And 3D accelerators could also use the extra bits even if they don't display them directly by preventing lighting from polluting the textures.
Much more entertaining was Microsoft BASIC for the Commodore line of computers... Not only were the programs stored in RAM in bytecode, but they were also stored on diskette/cassette the same way. Not only that, but due to the fact you could only enter 80 characters per line, you were allowed to remove all spaces, and use shortcuts consisting of the first two or three letters of the command, with the last character shifted. (reT would expand into RETURN, goS into GOSUB, etc). Microsoft even used this bytecode in their Visual Basic product until recently, although they felt the need to rename it P-code (pseudo-code).
Java could be considered a uncompiled binary as well... The program isn't natively executable and either requires compiling for whatever platform it's running on, or be run interpretively.
It's not the point because I was responding to the assumption that X takes an enormous amount of RAM. I was merely pointing out that it doesn't take nearly as much RAM as everyone thinks, and in that same RAM, has more native features than the comparative BeOS 'app_server'. I won't deny that X isn't without it's fair share of problems, not the least of which is the same client/server architecture that makes it so useful. My point was that comparing memory consumption between BeOS and XF86 is a pointless an inaccurate exercise -- they have very different capabilities and functions.
Slashdot Mirror
Nonsense, the election was decided long before the polls closed. After Ontario's polls were closed, the winner was decided, the rest of the country be damned. Manitoba, Saskatchewan, Alberta, British Columbia, the Northwest Territories and the Yukon have little influence in the polls when it comes to the federal elections. That's why we have this draconian law prohibiting broadcasting election results until after the polls close in the viewer area, or nationally.
There's an important difference between get rich schemes that fail and lives lost. These days, it seems that if you play with fire against everyone's recommendations, you can still sue someone for giving you the fire that you burnt yourself on. Laywers may be needed for certain things, but do they need to become a way of life? It seems that more and more they are becoming so. We've got people sueing McDonalds because they were stupid enough to pour hot coffee on themselves, people going after the entertainment industry because parents are screwing up raising their kids. It's SICK.
Also keep in mind there's a much higher differential between MP3 encoders. Certain encoders, in particular early Xing encoders and those based on the ISO source code, tended to generate very poor quality MP3s, with a great deal of lost information from the MP3. The Microsoft WMA encoder, on the other hand, is only developed by Microsoft, since the standards are not available to anyone who doesn't want to pay the license fees.
You forget that Microsoft is opposed to the entire CPRM idea. Hopefully they'll remain opposed to this and any/all derrivative technology that puts copy protection in hardware. They might end up being one of the biggest allys in the war against this technology.
People forget that people had found ways to get unencrypted .VOB files long before DeCSS was designed -- there are many ways of doing it, everything from grabbing data in memory after decryption to the old low-tech standbys of hooking video out to video in. Don't underestimate the ability for people to find ways around these technological control measures. Macrovision only stops the least determined of pirates. CD-R audio with it's serialization and high price never caught on (and likely never will). CSS was broken long before DeCSS was even thought of.
Many people tend to cover their eyes, afraid they might see something they don't like, afraid they might know something they'd rather not know, afraid they might change and cease being the same person they were before...
I'm a firm believer that patent (and copyright) lifetimes should be linked directly to the innovation happening in that field. Hopefully promoting lengths that allow companies to recoup their costs and make some profit, but still make it possible for the works to enter the public domain within a reasonable timeframe. Unfortunately, as long as governments are firmly within the grasp of corporate owned special interest groups, the age of insane patents and eternal copyright may not end anytime soon.
From a business perspective, he must be smart. However, most people overstate his programming and technical skills, since for the most part he didn't create the products his company sold, even in the early days, but rather just supervised their design and helped create the market for them. He had enough contacts and enough business sense to make lucrative deals in the early days that pretty much assured him the position he has now.
The problem is e2fsck ALWAYS runs when a bad shutdown occured. Win2K and NT sometimes seem to have this delusion that the journalling worked and they don't need to run autochk, but you usually end up rebooting again so that chkdsk /f can run. The time they take to check the partition also directly varies depending on the number of files and how fragmented those files are. Heavily fragmented partitions tend to take a lot longer to check for errors than unfragmented partitions. And despite what Microsoft's press releases might say, journalling (or rather metadata journalling, which is the most common variety) is a rather poor defense against data corruption since it only protects the structure of the file system, not the contents of it, and it's easy to forget that in the case of a bad shutdown, especially ones that involve the crash of a system driver, even the metadata journal can't always be trusted.
Even migrating from Windows NT 4.0 to Windows 2000 is far from an easy task. Incompatibilities ranging from incompatible ACPI BIOS revisions, to hardware incompatibilities to software incompatibilties. Migrating servers is even worse, because upgrading domain controllers means upgrading to Active Directory. Windows 2000 servers cannot act as BDCs, only member servers or Active Directory domain controllers, which means if you want to upgrade a BDC server to W2K, you first have to upgrade the PDC, and do all the planning that entails. To make this worse, if you want to say upgrade a Netware/NT mixed network and migrate the data from the Netware server, things get worse. Forcing you to purchase a copy of the Netware Services for Windows 5.0 package if migrating file permissions is even a little important to you, create or fix a pre-generated account mapping file. Then you get to actually migrate the data and find that it made a mess of the file permissions by preventing file permissions from being inherited where it would otherwise be possible. And after that's all done, you get to try and rewrite the login scripts using the rather stupid cmd.exe/command.com batch scripting language. I suffered through this, and had various bad things happen, including the deletion of all the Windows NT accounts (including the Administrator and Guest accounts).
The best example of this is back from Feburary when there was an announcement on the KDE pages that said they were going to collaborate on the object sharing framework. A few months later, all mention of this vanished and KParts appeared as KDE's object model and Bonobo on Gnome's end. And guess what? They don't work together, and in fact the work that would be required to get them to work together would be tremendous.
I see their PHP3 skills are top notch, allowing people to add scripting to the page, and even disrupt the PHP3 file itself.
OpenGL has a lot going for it. One is that it's a mature API and was designed by a company that is responsible for a great deal of the graphics technology that we've seen on TV and now are happening on our computers -- SGI. DirectX, or rather Direct3D in particular, was a late comer to the party and has had to play the catch-up game for the past couple years. Another advantage is the cross-platform capabilities, which also means you may be able to do portions of the development work on platforms a little more stable than the regular Windows 9x systems the game will likely be run on. Also don't forget that there have been large numbers of games using OpenGL... Quake II is a good early example, some later ones include Unreal and Half-Life also used OpenGL engines, and there wasn't any slowdown. I believe Quake 3 Arena also uses OpenGL, but I don't really play it.
But most of this is all beside the point. It's fairly simple to abstract the game and engine logic from the graphics back-end. For the most part these days, OpenGL and Direct3D are more alike than they are different, and most of the reason we have unportable games is developer laziness and development schedules that run far beyond what they should be forcing the developers to do a half-assed job and deliver an unfinished product.
Linux has other problems regarding games, aside from this silly Direct3D/OpenGL nonsense, the least of which is the rather poor driver support (we've got what? nvidia (through binary-only X servers/drivers), ATI, Matrox, and maybe one or two other manufacturers?), XFree86 4 just finally getting usable for regular use, and still rather poor support for advanced features of current sound cards (Yes, my SoundBlaster Live works in Linux, but still no EAX or 3D audio).
I can think of only one reason why a developer would want to always choose DirectX over OpenGL -- driver support for DirectX is almost always better than driver support for OpenGL on Microsoft Windows.
I think the original poster is a little confused. Word 97 and 2000 both support tracking changes to a document, however it's an optional feature that must be turned on to be useful. Despite this, however, it's known that Word documents can carry all sorts of data that has been deleted, on the clipboard, or from apps running on the same machine. Running a 'strings' command on a word document will often reveal data that was never in the document, was removed from the document at some point and other tidbits of information.
I think some UNIX-style operating systems have integrated protection from this sort of simplistic DoS attack, as well, aside from disk space/inode quotas.
The only thing they could do is license the diffs as public domain. They can't relicense the kernel as public domain since it's already licensed under the GNU GPL and they have no authorization to relicense it under different terms. Government agencies have to play by the same rules as everyone else -- the GPL grants no special permissions to military agencies.
E-mail spam in many ways mimics the so-called junk faxes that have been outlawed in most places. Despite this fact, companies still send these junk faxes. I believe that companies and individuals that send these junk e-mails should be treated in the exact same way people who commit the same crime with facsimile machines are. Both are unauthorized use of my resources, which I pay for. I have much less of a problem with junk mail, since I don't have to pay to receive it. The sender pays, and I trash it. I also have a problem with telemarketers because they waste a considerable amount of my time, tie up my telephone and call at inappropriate times.
I do, however, agree that the MAPS RBL is a rather ineffective measure, and certainly no better than any of the censorware filtering software that we've all grown to hate. The MAPS RBL blocks as much or more legitimate traffic than it stops, and the real solution is to either prevent anyone from sending unsolicited commercial e-mail or to require that all unsolicited spam be marked in a header or on the subject line in a standardized fashion so that the recipient can always filter if he or she chooses to do so. I would certainly prefer the first measure come into being for the simple reason that I don't see why trash e-mails need to be sent -- there are better, less intrusive ways of advertising, the easiest being putting up a website and letting a few search engines spider it.
And to all the spammers out there: For the millionth time, no I don't want a non-accredited diploma or degree, to make money fast, mass-mailer software, e-mail list CDs, or viagra.
On the other hand, this is an artifact of how human sight works. If you look at RGB(255,0,0) versus RGB(0,0,255) and RGB(0,255,0), the blue box tends to be the darkest, the red is next, and green is the brightest, therefore it stands to reason you'd be able to distinguish more shades of green than blue, if for no other reason than the percieved intensity difference. Intensity itself isn't even percieved linearly by the eye to make things worse. Even if we were to go to 16-bits per colour, or 48bpp, it's still conceivable that there would be some colours that the human could noticibly percieve but the monitor could not generate. However, for the vast majority of people, 24bpp is 'good enough' and achieves everything they'd want out of a video card. I could, however, see 36bpp becoming an option on some video cards for people who require very accurate colour matching for things like publishing. And 3D accelerators could also use the extra bits even if they don't display them directly by preventing lighting from polluting the textures.
Java could be considered a uncompiled binary as well... The program isn't natively executable and either requires compiling for whatever platform it's running on, or be run interpretively.
It's not the point because I was responding to the assumption that X takes an enormous amount of RAM. I was merely pointing out that it doesn't take nearly as much RAM as everyone thinks, and in that same RAM, has more native features than the comparative BeOS 'app_server'. I won't deny that X isn't without it's fair share of problems, not the least of which is the same client/server architecture that makes it so useful. My point was that comparing memory consumption between BeOS and XF86 is a pointless an inaccurate exercise -- they have very different capabilities and functions.