Read the article - it affects NON-bus mastering devices. This means:
1. Most block I/O devices ARE BUS-mastering (all IDE controllers on the market today are, and most SCSI controllers are - unless you are trying to use your cheapo scanner card to run a drive, which isn't recommended anyway).
2. Some PCI video cards made post-1996 are bus-mastering.
3. Some NICs are busmastering.
On older mobos (of the original pentium era), there were usually 4 PCI slots, 2 of which were shared ISA, and the first one was the only one that was busmastering, and this caused conflicts when you had 2 devices that required busmastering (such as a NIC and a Videocard). NICs and Vidcards made during the 486-586 era were busmastering because the PCI bus couldn't be shared efficiently. Nowadays, with a bus speed higher than 66mhz and more than 2 independant PCI slots, it isn't necessary to restrict data transfer to one high bandwidth device at a time.
It is no longer necessary to stick the bus-mastering card in the first PCI slot anymore.
Having read all the comments so far, it seems that the issue is INsignificant, has already been fixed, and warrants no further reporting. It doesn't even merit a headline anymore, as existing P4 users and linux hackers alike have agreed that most things don't check CPUID for functionality anyway, only for optimization.
How do you explain the capabilities of the Echelon system which logs all transatlantic electronic transmission. From what I know, it uses a queueing system to process each transmission stream for keywords which tag those parties for further electronic surveillance.
There have been real reports of civilians who were arrested for communicating minimally classified government information, and credit was given to Echelon.
This sound just like Echelon is is going to be purchasing some several hundread Terabyte RAID disks as well as probably half a million spools of magnetic tape (the CIA still uses magnetic tape you know!) so they can keep semi-permanent logs.
The human body is only a shell. Humanity exists because of what you may call a ghost or a soul. It is being sentient that makes us human. No other animal is sentient on this planet.
By being sentient, we have created medical technology that allows humans who might otherwise die, to continue their influence on others.
Suppose that Stephen Hawking were about to die from ALS, would you not want a medical treatment that could restore his ability to exist, so that he may continue unraveling the secrets of the universe?
As it takes several petabytes of information to hold the contents of the brain in a computer, in addition to the fact that we have not been able to figure out how the brain stores that information so that we may be able to retrieve it, we can't preserve a person who's about to die but has not completed 'their life's work'.
Immortality certainly has philsophical implications (although i wouldn't say ethical) - If you have seen Robin William's 'Bicentennial Man', that movie addresses such philosophical issues.
I don't see how using xenografts degrade humanity - they actually are better than allografts because instead of using another human's organs, youre using an animal. If you had 1 month to live due to congestive heart failure would you rather receive a heart transplant from your dead sibling/parent or would rather receive a pig heart?
The funeral should be the traditional jazz funeral.
Basically it is one huge party. Death's significance is as large as one of a wedding. So people should party. Life is celebrated, but shouldn't we celebrate his passing on to Afterlife?.
May he find all the answers to his unanswerable questions that he has asked on earth. Maybe he can now really come back and fiddle around with those electrons flowing through these pesky networks here. But he has no reason to... in the Great unfailing Internet in the sky, he probably won't even need 'ping'.
Muuss is dead! Long live Muuss!
LANCE ISA cards still not working (lnc0)
on
FreeBSD 4.2 Is Out
·
· Score: 1
ISA NICs using LANCE AMD chipsets are still borked in 4.2's default kernel (the one used to boot the setup interfaces). I was informed by someone on -questions that prior to 4.2, the lance driver (lnc0) conflicted with the pnp driver called 'pcn0'. However my card (a 79C960 chip) still won't initialise properly under 4.2 (and all "trivial" problems have been ruled out extensively).
I have tried and tried again to get someone to troubleshoot the card - AT1500BT - **Which works perfectly in Linux** including emailing the author of the LANCE driver (lnc0), with no avail!
This is considered critical as the system has no cdrom drive, and thus needs to use FTP or NFS based install which means that it needs the nic to work.
I'm sure all the universities are happy now
on
Scour is Dead
·
· Score: 1
All the universities must ecstatic now that scour exchange is dead. We just spent a crapload of money on a traffic shaper that does layer 7 (content) analysis and packet caching to prevent all the kids from leeching and using up all the bandwidth (scour/napster traffic contributed 85% of the total bandwidth usage). Primarily because many college kids are clueless and don't know that if napster/scour ends up in the system tray and is not manually disconnected, it is still uploading stuff.
I guess it was pretty ironic that right after this happened our traffic on the T3 dropped 50% because no one was using scour anymore.
Solution: if you use pirated software and don't have time to comply-
write a worm that will seek and delete all copies of the offending software, activated by a specific combination of commands/keystrokes.
So while you appear to be hitting the send button on your email client, it's essentially propagating itself through the network, searching and destroying.
This approach works even better when you have 'ghosted' machines, (when you clone installations to make all machines have the same set of software).
Of course, you would have to set up a miltary style of authentication and security etc. that can be executed in dire situations and minimises false alarms.
sample scenario:
When the troops come in the front door, you get advanced warning (i'm sure most companies have receptionists).
Exchange immediate authentication for the worm activation.
"step away from the keyboard!"
"*keystroke keystroke keystroke, click.*"
boom.
Repairing the damage may take resources, but they are nothing compared to what you need to pay or what you will face.
Americans have a hard time understanding this, because the UK taxes televisions, and they periodically have vans that come out looking for RF emissions typical of a TV in suspected tv-tax evaders.
In the Molecular and Cell biology department, BSA stands for a commonly used biochemical, namely Bovine Serum Albumin. Albumin is a protein found in egg whites and the blood (serum). BSA has an 'average' molecular weight for a protein, stores well, is easily detectable using photospectrometry. It is slightly viscous when you first make it so...
Imagine some giant monster goop-like blob that's BSA coming to inspect...
says that it has a built in LiIon battery which will last 2 weeks or 6 continous hours.
The thing i find more of a disadvantage is that it still only comes with 8M of ram but is only $50 less than the iPaq. I would personally ditch the Visor Prism and get an iPaq because it comes with double the RAM and is only $50USD more (at MSRP $500USD), and you can never have 'enough' ram (no one needs more than 640K? yah right!).
There is the older GCC available (2.7) in RH7 which lets you compile the kernel. I think it comes under the label 'kgcc' for 'kernel gcc'.
I don't remember the exact details would someone elaborate?
Honda has a developmental bi-pedal robot that looks like a human in a spacesuit. It has 4 independant limbs and balance control. It can negotiate stairs and will use it's arms/hands to stabilize itself. I think they want to put basic neural net functionality in it too (so it can do complex tasks like hoola-hooping).
They are developing this at a 'secret test facility' (according to the press). But, if you really want, and have the credentials (e.g. Ph.D in cognitive psychology) and would like to preview i don't know if they wouldn't let you. Of course, you'd probably have to play with it in Japan.
In Singapore, bubble gum is banned in public areas including malls and restaurants because it is "disruptive, rude, and litter-producing". This comes from a country which is considered one of the most digital ("the intelligent island") in asia, second to Japan. If Singapore can ban bubble gum, it's not surprising to find other things banned in other asian countries such as video arcades..
The legislation as it stands has good ideas in it except for that it can be misinterpreted too easily.
I would have rather that someone addressed these points:
Point 1: The forcing of giving credit to where it's due. This is done in the scientific community where you must cite all of the people you derive your work from. This is what a copyright is supposed to be for.
Point 2: disallow profiting in the absence of compensation.
One interesting thing is that the government isn't supposed to profit from anything.
The academic sector is unique in that it's a mix of the public and private sector: the professor can claim rights to his work, but he can also turn them over to the university. Furthermore:...
Phil's argument that "your own lectures are an example of commercially profiting through the distribution of a derivative work", isn't entirely true. The DoD or NIH or any educational institution and even commercial businesses can fund a faculty member for research. This technically doesn't fall under "making a profit", since the beneficiary needs an operating budget so they can continue research, this includes budgeting for living needs.
As a faculty member, one doesn't make a profit. Making a profit from research only occurs if you work in the private sector, but then the company makes a profit too.
As you can see in the commercial world:
Profit-earning only works if everyone in the chain gets a share of the benefits. If the professors could charge royalties for redistribution of notes for profit, that would be ok, but then everyone on slashdot would complain about 'free speech'. This isn't about 'free speech' it's about "should companies and students make money off professors who aren't directly compensated" in addition to "companies and students selling information at the expense of professors".
Technically, you can't measure the rate of devel speed for linux per se because...
Let's see, Linux has no revision control system? As said by someone else, "The whole, 'let's distribute a tarball every other period of time is getting old'".
Just because you get yet another release of kernel 2.2.x (what is it at now? 17?) more often then FreeBSD 4.x, that doesn't mean the development is faster, it just means the kernel writers abitrarily increment the subversion number everytime they feel like it. You don't see what's going on with FreeBSD dev because from your trolling it appears that you don't use CVS, or if you do, you don't know how to use it correctly.
Let's see when 2.2.18 and 19 come out with just 2 minor things patched.
One word. FreeBSD.
Perfect for both server and workstation.
Easy installation.
Easy upgrade and low maintenance. Also, free uptodate security advisories on freebsd.org itself.
CVS support = global upgradeability. You cvsup to freebsd's cvs tree, pull the new stuff, then 'make world' to synchronize kernelspace with userspace and voila. Upgrade done.
Tons and tons of software - on the/usr/ports collection.
No native library conflicts like linux. BSD binaries, are...BSD binaries. That means they are cross compatible with BSD based kernels. Furthermore, you can install the linux binary compatiblity port which lets you run ELF binaries (I haven't used this so I don't know how well it works).
No linux kernel issues. Since everything is based on the revision control system called CVS, you don't need to diff patch your kernel, just cvsup.
And if you write new things for BSD, the BSD licence makes more sense than GPL *wink* *wink* since you can actually resell your code instead of being forced to give it away!
My roommate happens to run ipv6, and i can *assure* you the issue over the MAC address inclusion is far from the truth. Parts of the MAC are used in the address, as well as probably the last 2-3 fields. You can't possiblly track a NIC down with that information. A MAC has 2 principle parts: a manuafacturer ID (the first 2-3 fields) and the variable part (the rest of the MAC).
This is the general case, since depending on the OS, your MAC could be read differently.
You can even kernel hack it in Linux to create a different MAC! (someone I know did that and changed their OS to report '00DEADC0FF33' as the new MAC).
All of this variability and recombination makes it very difficult (if impossible) to deduce the MAC from the ipV6 address. The other hex parts of the address are generated on the fly anyway, when you get assigned your address after boot.
If they do NOT print the EULA out, it is unenforceable. IIRC, all legal documents must be either on hardcopy, or immediately be printable when the user opens the package (e.g. the 'print' button on the EULA when you install stuff).
I think there is CodeWarrior for Linux out (I was offered a copy over the summer by a colleague). If the version that came with the book is windows only and if you can not get access to a windows machine to install it and use, you could probably contact CodeWarrior and ask if you could swap the windows version for the linux port. Most likely they should let you, for either a low or no price at all. I still remember the age where you could trade in your set of 3.5" floppies for a 5.25" set for free (or s+h). This 'de-facto standard' may still apply today.
Of course, CodeWarrior for Linux is a little messy - since they try to integrate it with all the gnu tools - so the same compiler differences might still crop up.
At least you can still say "I did compile it on CodeWarrior"
Some more suggestions: In high school, we did not have a CS course, but rather MIS. Here are some useful skills in the industry to build and some ideas:
1. Get in touch with the local college or community college and see if you can bus the HS kids over there for some 'real' CS courses. I know that 3Com has set up many Networking Technician courses in community colleges around the country that will get you certified in 3Com equipment maintence in addition to teaching you more than you want to know about routers, switches, networks, etc. If there is an IBM office nearby, check to see if they offer interships in web-design or advanced r&d testing. If you can get in touch with JPL or NASA, they routinely fly missions where high school students can submit proposals for microgravity experiments - find a n accepted proposal and have your students volunteer to write the control program for the experiment. Similarly, see if you can enter any programming competitions - or robotics for that matter.
2. Have your kids set up a server of some sort, either NT or UNIX, because systems administration is also a hot job market, especially NT nowadays (my university is half run on IBM s/390 and the other half on NT). Then see if the school board needs new servers or applications. By giving them a 'real' project to do with real deadlines and a real goal (such as porting over the school's 10 year old student database stuff) you prepare them to go to college as well as industry where everything is done in committees and round-table discussion groups. Also, send your kids to do tech support every where in the school system and see if they can either get extra credit or get paid. Customer support might be a dirty job but it requires extensive knowledge of whatever system theyre working on. To make a broken system, especially Microsoft-based, usable, lots of tricks can resurrect a machine that most people would want to reformat. The 'real' world is 99% people interaction and workplace-politics, even if your kids are planning to sit at a desk all their lives and write code, they still need to interact with their superiors and peers. Get them started on learning professional relationship building, it could mean the difference between being the ten thousand dollar raise or promotion to chief software architect.
3. Here's a plug for distributed.net: If any high school kids know C/C++ or other languages, have them join distributed.net as code contributors. It's still 'real' world stuff, you'll see other people actually use the code. Plus, getting involved in distributed computing ventures is putting them ahead for the future.
4. Get lots and lots of O'reilly books. Learning Perl and Python and VB and other scripting languages is very useful - I was denied many a promotion opportunity because I am not a versed database programmer.
In conclusion you want to have your kids do something productive and worthwhile and that they might see other people outside of school using some script or application they built.
Ok i am starting to sound like the parent topic i replied to. so i'll stop.
Also, a web browser isn't like some sort of black box. If you put code in it that doesn't work, then find the modified tags that will and then you have essentially discovered all the new 'features' and incompatibilities. Thus you can release draft specifications. AFAIK, browser technologies can't be classified as trade secrets because every one with HTML4 know-how can figure out what the quirks are.
NS sucks. And when the rest of your organization uses stuff like Exchange Server with M$ authentication, you can't even check your email with Nutscrape! NS for Linux and Mac lags behind IE so much. I just wish there was an IE port for Linux (which if I were Mr. Gates would do - increase the market dominance by another 10% or so) Same goes for linux/bsd versions of Office. Hey Corel did that!
This guy sounds like he works at Apple, as he praises and lauds the MacOS interface as 'the gold standard for GUIs' which I think should really die.
The only reason why I look forward to OS X is that it contains a bsd-based kernel and finally finally has a command line! I actually run a beta of OS X and the GUI is still pretty shameful. The only other redeeming quality is that it supports both bsd and mac applications, which is required in my workplace for backwards compatility.
Otherwise MacOS is NOT intuitive, actually lets people screw the system up more easily, and makes writing Mac applications a pain in the a$$!
1. I for one use almost exclusively TrueType fonts on XFree86 so I HNFC what you are saying with the lack of fonts/anti-aliasing on X.
2. VNC might be light and dandy, but it's horribly slow and I only use it when I don't have a local X server. Yeah, X might be a bloated piece of antiquity, but at least it gets the job done better than VNC or PCAnywhere. If you want to talk more about font replication, I can't stand looking at Corel Wordperfect over VNC. And X also means I can start a specific X app instead of having to refresh the whole desktop every time I move the mouse.
3. X does start up faster than MacOS or Windows at any rate. (it takes 3 min. for Win98SE to start up vs. 50s Slackware 7 with XDM on the same dual-booting system).
4. Maybe everyone who's cracking on X should port BeOS to run on top of the unix kernel eh? Since that GUI supports the newest multimedia hardware. (and BTW, Be also starts up in 20s and never ever requires a e2fsck or some other filesys integrity util)
Re:DNA testing nearly impossible with this inventi
on
Walk-By DNA Testing
·
· Score: 1
good question...hmmmm:)))
Legally, IIRC, DNA evidence in court can only be used to prove that someone is *not* the father or some did *not* commit a crime. It can't be used to prosecute someone, which is why other evidence is needed to convict. I don't know why this is the case, because mathematically the chances that you and someone else has the same DNA fingerprints is 1 in 6 billion. Maybe the judges all got scared when the movie Judge Dredd came out (starring sylvester stallone).
Actually where I'm at, the Univ of Connecticut, some dude used skin cells to clone a cow. So DNA evidence might someday have to be thrown out of the window when every other person is a clone (slightly Huxley-esque eh?) Bioethics is getting to be a very interesting field
Slashdot Mirror
Read the article - it affects NON-bus mastering devices. This means:
1. Most block I/O devices ARE BUS-mastering (all IDE controllers on the market today are, and most SCSI controllers are - unless you are trying to use your cheapo scanner card to run a drive, which isn't recommended anyway).
2. Some PCI video cards made post-1996 are bus-mastering.
3. Some NICs are busmastering.
On older mobos (of the original pentium era), there were usually 4 PCI slots, 2 of which were shared ISA, and the first one was the only one that was busmastering, and this caused conflicts when you had 2 devices that required busmastering (such as a NIC and a Videocard). NICs and Vidcards made during the 486-586 era were busmastering because the PCI bus couldn't be shared efficiently. Nowadays, with a bus speed higher than 66mhz and more than 2 independant PCI slots, it isn't necessary to restrict data transfer to one high bandwidth device at a time.
It is no longer necessary to stick the bus-mastering card in the first PCI slot anymore.
Having read all the comments so far, it seems that the issue is INsignificant, has already been fixed, and warrants no further reporting. It doesn't even merit a headline anymore, as existing P4 users and linux hackers alike have agreed that most things don't check CPUID for functionality anyway, only for optimization.
How do you explain the capabilities of the Echelon system which logs all transatlantic electronic transmission. From what I know, it uses a queueing system to process each transmission stream for keywords which tag those parties for further electronic surveillance.
There have been real reports of civilians who were arrested for communicating minimally classified government information, and credit was given to Echelon.
This sound just like Echelon is is going to be purchasing some several hundread Terabyte RAID disks as well as probably half a million spools of magnetic tape (the CIA still uses magnetic tape you know!) so they can keep semi-permanent logs.
The human body is only a shell. Humanity exists because of what you may call a ghost or a soul. It is being sentient that makes us human. No other animal is sentient on this planet.
By being sentient, we have created medical technology that allows humans who might otherwise die, to continue their influence on others.
Suppose that Stephen Hawking were about to die from ALS, would you not want a medical treatment that could restore his ability to exist, so that he may continue unraveling the secrets of the universe?
As it takes several petabytes of information to hold the contents of the brain in a computer, in addition to the fact that we have not been able to figure out how the brain stores that information so that we may be able to retrieve it, we can't preserve a person who's about to die but has not completed 'their life's work'.
Immortality certainly has philsophical implications (although i wouldn't say ethical) - If you have seen Robin William's 'Bicentennial Man', that movie addresses such philosophical issues.
I don't see how using xenografts degrade humanity - they actually are better than allografts because instead of using another human's organs, youre using an animal. If you had 1 month to live due to congestive heart failure would you rather receive a heart transplant from your dead sibling/parent or would rather receive a pig heart?
The funeral should be the traditional jazz funeral.
Basically it is one huge party. Death's significance is as large as one of a wedding. So people should party. Life is celebrated, but shouldn't we celebrate his passing on to Afterlife?.
May he find all the answers to his unanswerable questions that he has asked on earth. Maybe he can now really come back and fiddle around with those electrons flowing through these pesky networks here. But he has no reason to... in the Great unfailing Internet in the sky, he probably won't even need 'ping'.
Muuss is dead! Long live Muuss!
ISA NICs using LANCE AMD chipsets are still borked in 4.2's default kernel (the one used to boot the setup interfaces). I was informed by someone on -questions that prior to 4.2, the lance driver (lnc0) conflicted with the pnp driver called 'pcn0'. However my card (a 79C960 chip) still won't initialise properly under 4.2 (and all "trivial" problems have been ruled out extensively).
I have tried and tried again to get someone to troubleshoot the card - AT1500BT - **Which works perfectly in Linux** including emailing the author of the LANCE driver (lnc0), with no avail!
This is considered critical as the system has no cdrom drive, and thus needs to use FTP or NFS based install which means that it needs the nic to work.
All the universities must ecstatic now that scour exchange is dead. We just spent a crapload of money on a traffic shaper that does layer 7 (content) analysis and packet caching to prevent all the kids from leeching and using up all the bandwidth (scour/napster traffic contributed 85% of the total bandwidth usage). Primarily because many college kids are clueless and don't know that if napster/scour ends up in the system tray and is not manually disconnected, it is still uploading stuff.
I guess it was pretty ironic that right after this happened our traffic on the T3 dropped 50% because no one was using scour anymore.
Solution: if you use pirated software and don't have time to comply-
write a worm that will seek and delete all copies of the offending software, activated by a specific combination of commands/keystrokes.
So while you appear to be hitting the send button on your email client, it's essentially propagating itself through the network, searching and destroying.
This approach works even better when you have 'ghosted' machines, (when you clone installations to make all machines have the same set of software).
Of course, you would have to set up a miltary style of authentication and security etc. that can be executed in dire situations and minimises false alarms.
sample scenario:
When the troops come in the front door, you get advanced warning (i'm sure most companies have receptionists).
Exchange immediate authentication for the worm activation.
"step away from the keyboard!"
"*keystroke keystroke keystroke, click.*"
boom.
Repairing the damage may take resources, but they are nothing compared to what you need to pay or what you will face.
Americans have a hard time understanding this, because the UK taxes televisions, and they periodically have vans that come out looking for RF emissions typical of a TV in suspected tv-tax evaders.
In the Molecular and Cell biology department, BSA stands for a commonly used biochemical, namely Bovine Serum Albumin. Albumin is a protein found in egg whites and the blood (serum). BSA has an 'average' molecular weight for a protein, stores well, is easily detectable using photospectrometry. It is slightly viscous when you first make it so... Imagine some giant monster goop-like blob that's BSA coming to inspect...
The spec sheet
http://www.handsprin g.c om/products/visorprism/compare.jhtml
http://www.handsprin g.c om/products/visorprism/details.jhtml
says that it has a built in LiIon battery which will last 2 weeks or 6 continous hours.
The thing i find more of a disadvantage is that it still only comes with 8M of ram but is only $50 less than the iPaq. I would personally ditch the Visor Prism and get an iPaq because it comes with double the RAM and is only $50USD more (at MSRP $500USD), and you can never have 'enough' ram (no one needs more than 640K? yah right!).
There is the older GCC available (2.7) in RH7 which lets you compile the kernel. I think it comes under the label 'kgcc' for 'kernel gcc'. I don't remember the exact details would someone elaborate?
Honda has a developmental bi-pedal robot that looks like a human in a spacesuit. It has 4 independant limbs and balance control. It can negotiate stairs and will use it's arms/hands to stabilize itself. I think they want to put basic neural net functionality in it too (so it can do complex tasks like hoola-hooping).
They are developing this at a 'secret test facility' (according to the press). But, if you really want, and have the credentials (e.g. Ph.D in cognitive psychology) and would like to preview i don't know if they wouldn't let you. Of course, you'd probably have to play with it in Japan.
In Singapore, bubble gum is banned in public areas including malls and restaurants because it is "disruptive, rude, and litter-producing". This comes from a country which is considered one of the most digital ("the intelligent island") in asia, second to Japan. If Singapore can ban bubble gum, it's not surprising to find other things banned in other asian countries such as video arcades..
The legislation as it stands has good ideas in it except for that it can be misinterpreted too easily. I would have rather that someone addressed these points: Point 1: The forcing of giving credit to where it's due. This is done in the scientific community where you must cite all of the people you derive your work from. This is what a copyright is supposed to be for. Point 2: disallow profiting in the absence of compensation. One interesting thing is that the government isn't supposed to profit from anything. The academic sector is unique in that it's a mix of the public and private sector: the professor can claim rights to his work, but he can also turn them over to the university. Furthermore:... Phil's argument that "your own lectures are an example of commercially profiting through the distribution of a derivative work", isn't entirely true. The DoD or NIH or any educational institution and even commercial businesses can fund a faculty member for research. This technically doesn't fall under "making a profit", since the beneficiary needs an operating budget so they can continue research, this includes budgeting for living needs. As a faculty member, one doesn't make a profit. Making a profit from research only occurs if you work in the private sector, but then the company makes a profit too. As you can see in the commercial world: Profit-earning only works if everyone in the chain gets a share of the benefits. If the professors could charge royalties for redistribution of notes for profit, that would be ok, but then everyone on slashdot would complain about 'free speech'. This isn't about 'free speech' it's about "should companies and students make money off professors who aren't directly compensated" in addition to "companies and students selling information at the expense of professors".
Technically, you can't measure the rate of devel speed for linux per se because...
Let's see, Linux has no revision control system? As said by someone else, "The whole, 'let's distribute a tarball every other period of time is getting old'".
Just because you get yet another release of kernel 2.2.x (what is it at now? 17?) more often then FreeBSD 4.x, that doesn't mean the development is faster, it just means the kernel writers abitrarily increment the subversion number everytime they feel like it. You don't see what's going on with FreeBSD dev because from your trolling it appears that you don't use CVS, or if you do, you don't know how to use it correctly.
Let's see when 2.2.18 and 19 come out with just 2 minor things patched.
One word. FreeBSD. /usr/ports collection.
Perfect for both server and workstation.
Easy installation.
Easy upgrade and low maintenance. Also, free uptodate security advisories on freebsd.org itself.
CVS support = global upgradeability. You cvsup to freebsd's cvs tree, pull the new stuff, then 'make world' to synchronize kernelspace with userspace and voila. Upgrade done.
Tons and tons of software - on the
No native library conflicts like linux. BSD binaries, are...BSD binaries. That means they are cross compatible with BSD based kernels. Furthermore, you can install the linux binary compatiblity port which lets you run ELF binaries (I haven't used this so I don't know how well it works).
No linux kernel issues. Since everything is based on the revision control system called CVS, you don't need to diff patch your kernel, just cvsup.
And if you write new things for BSD, the BSD licence makes more sense than GPL *wink* *wink* since you can actually resell your code instead of being forced to give it away!
My roommate happens to run ipv6, and i can *assure* you the issue over the MAC address inclusion is far from the truth. Parts of the MAC are used in the address, as well as probably the last 2-3 fields. You can't possiblly track a NIC down with that information. A MAC has 2 principle parts: a manuafacturer ID (the first 2-3 fields) and the variable part (the rest of the MAC).
This is the general case, since depending on the OS, your MAC could be read differently.
You can even kernel hack it in Linux to create a different MAC! (someone I know did that and changed their OS to report '00DEADC0FF33' as the new MAC).
All of this variability and recombination makes it very difficult (if impossible) to deduce the MAC from the ipV6 address. The other hex parts of the address are generated on the fly anyway, when you get assigned your address after boot.
If they do NOT print the EULA out, it is unenforceable. IIRC, all legal documents must be either on hardcopy, or immediately be printable when the user opens the package (e.g. the 'print' button on the EULA when you install stuff).
I think there is CodeWarrior for Linux out (I was offered a copy over the summer by a colleague). If the version that came with the book is windows only and if you can not get access to a windows machine to install it and use, you could probably contact CodeWarrior and ask if you could swap the windows version for the linux port. Most likely they should let you, for either a low or no price at all. I still remember the age where you could trade in your set of 3.5" floppies for a 5.25" set for free (or s+h). This 'de-facto standard' may still apply today. Of course, CodeWarrior for Linux is a little messy - since they try to integrate it with all the gnu tools - so the same compiler differences might still crop up. At least you can still say "I did compile it on CodeWarrior"
Some more suggestions:
In high school, we did not have a CS course, but rather MIS. Here are some useful skills in the industry to build and some ideas:
1. Get in touch with the local college or community college and see if you can bus the HS kids over there for some 'real' CS courses.
I know that 3Com has set up many Networking Technician courses in community colleges around the country that will get you certified in 3Com equipment maintence in addition to teaching you more than you want to know about routers, switches, networks, etc.
If there is an IBM office nearby, check to see if they offer interships in web-design or advanced r&d testing. If you can get in touch with JPL or NASA, they routinely fly missions where high school students can submit proposals for microgravity experiments - find a n accepted proposal and have your students volunteer to write the control program for the experiment. Similarly, see if you can enter any programming competitions - or robotics for that matter.
2. Have your kids set up a server of some sort, either NT or UNIX, because systems administration is also a hot job market, especially NT nowadays (my university is half run on IBM s/390 and the other half on NT). Then see if the school board needs new servers or applications. By giving them a 'real' project to do with real deadlines and a real goal (such as porting over the school's 10 year old student database stuff) you prepare them to go to college as well as industry where everything is done in committees and round-table discussion groups.
Also, send your kids to do tech support every where in the school system and see if they can either get extra credit or get paid. Customer support might be a dirty job but it requires extensive knowledge of whatever system theyre working on. To make a broken system, especially Microsoft-based, usable, lots of tricks can resurrect a machine that most people would want to reformat. The 'real' world is 99% people interaction and workplace-politics, even if your kids are planning to sit at a desk all their lives and write code, they still need to interact with their superiors and peers. Get them started on learning professional relationship building, it could mean the difference between being the ten thousand dollar raise or promotion to chief software architect.
3. Here's a plug for distributed.net: If any high school kids know C/C++ or other languages, have them join distributed.net as code contributors. It's still 'real' world stuff, you'll see other people actually use the code. Plus, getting involved in distributed computing ventures is putting them ahead for the future.
4. Get lots and lots of O'reilly books. Learning Perl and Python and VB and other scripting languages is very useful - I was denied many a promotion opportunity because I am not a versed database programmer.
In conclusion you want to have your kids do something productive and worthwhile and that they might see other people outside of school using some script or application they built.
Ok i am starting to sound like the parent topic i replied to. so i'll stop.
Agreed.
Also, a web browser isn't like some sort of black box. If you put code in it that doesn't work, then find the modified tags that will and then you have essentially discovered all the new 'features' and incompatibilities. Thus you can release draft specifications. AFAIK, browser technologies can't be classified as trade secrets because every one with HTML4 know-how can figure out what the quirks are.
NS sucks. And when the rest of your organization uses stuff like Exchange Server with M$ authentication, you can't even check your email with Nutscrape! NS for Linux and Mac lags behind IE so much. I just wish there was an IE port for Linux (which if I were Mr. Gates would do - increase the market dominance by another 10% or so)
Same goes for linux/bsd versions of Office. Hey Corel did that!
This guy sounds like he works at Apple, as he praises and lauds the MacOS interface as 'the gold standard for GUIs' which I think should really die.
The only reason why I look forward to OS X is that it contains a bsd-based kernel and finally finally has a command line! I actually run a beta of OS X and the GUI is still pretty shameful. The only other redeeming quality is that it supports both bsd and mac applications, which is required in my workplace for backwards compatility.
Otherwise MacOS is NOT intuitive, actually lets people screw the system up more easily, and makes writing Mac applications a pain in the a$$!
1. I for one use almost exclusively TrueType fonts on XFree86 so I HNFC what you are saying with the lack of fonts/anti-aliasing on X.
2. VNC might be light and dandy, but it's horribly slow and I only use it when I don't have a local X server. Yeah, X might be a bloated piece of antiquity, but at least it gets the job done better than VNC or PCAnywhere. If you want to talk more about font replication, I can't stand looking at Corel Wordperfect over VNC. And X also means I can start a specific X app instead of having to refresh the whole desktop every time I move the mouse.
3. X does start up faster than MacOS or Windows at any rate. (it takes 3 min. for Win98SE to start up vs. 50s Slackware 7 with XDM on the same dual-booting system).
4. Maybe everyone who's cracking on X should port BeOS to run on top of the unix kernel eh? Since that GUI supports the newest multimedia hardware. (and BTW, Be also starts up in 20s and never ever requires a e2fsck or some other filesys integrity util)
good question...hmmmm :)))
Legally, IIRC, DNA evidence in court can only be used to prove that someone is *not* the father or some did *not* commit a crime. It can't be used to prosecute someone, which is why other evidence is needed to convict.
I don't know why this is the case, because mathematically the chances that you and someone else has the same DNA fingerprints is 1 in 6 billion. Maybe the judges all got scared when the movie Judge Dredd came out (starring sylvester stallone).
Actually where I'm at, the Univ of Connecticut, some dude used skin cells to clone a cow. So DNA evidence might someday have to be thrown out of the window when every other person is a clone (slightly Huxley-esque eh?)
Bioethics is getting to be a very interesting field