I wonder if these fiber optic roofs will allow people (spy satellites?) to see inside a room when the luminosity inside the room is higher than that outside. Think of it like peering into a house's front windows at night -- as long as the living room lights are on, you can see in, but they can't see out.
What many people fail to realize is that companies like Cisco and Juniper create an OS that interoperates with specialized ASICs that allow a network device to switch/route packets much faster than a traditional time-shared system. While I think the XORP initiative is a good step forward in the way of low-end routers, will XORP give full support to users, for both the hardware and software? If you call Cisco's TAC for support, they will look at your entire system, not just the software and not just the hardware. Companies pay more but they get more in return.
You beat me to the posting, John.:-) We're actually looking at it more from a "Georgia Tech doesn't have a right to regulate the airwaves" standpoint than anything. Direct links to the relevant information are here:
Given that the Linux is the kernel and a distro is just a series of programs combined with that kernel, it seems ridiculous that they tried 9 distros with approximately the same kernel to get the card working. The problem doesn't exist in Linux, either. If the companies who make the cards don't write the drivers, nothing will get done.
He also discovered in the sales-support database... the names of more than 30 [identity-classified] employees of the United States National Security Agency...
While I agree for national security purposes the list of employees should probably be held for review, I do hope a list exists somewhere and is being looked at. I would hate to think this whistleblower is acting like Joseph McCarthy 50-some-odd years ago.
Last I heard, the entire tty layer needed to be rewritten and the scsi interface was broken. This was about mid-May. Given those two facts alone, I would be impressed if 2.6 was done in 9 months, to say the least.
Like another thread said, Linus has always stuck with a "when it's done" policy that has helped to make the kernel so stable. Why start releasing stable-versioned kernels with potentially bad/highly experimental code now that Linux is a serious competitor in the server market and beginning to encroach on the desktop market?
I personally like the professors who pile on about 25 hours of work per week, including all those mundane multiple page homeworks and lab write-ups that you know they (or the TAs who can't speak quarter-way decent english) never really read. You know that if you go to the professor to talk about it, nothing will get done and you would have simply wasted another precious hour (or two) of sleep and/or other work.
Perhaps it is because that professor too is tired of dealing with non-english speaking TAs/grad students, or perhaps it is because he actually has to teach students rather than simply letting the herd bask in his infinitely superior knowledge -- a classic case of transference where all the anger, hatred, and frustration is directed to the poor unfortunate souls who inadvertently signed up for his section.
Of course, in retaliation, sites like http://www.engrish.com are written instead of doing those thousands of projects, leaving the ingenious creators with a mere C for the semester.:-)
Would the janitor also count as an official representative if he posted about how to properly clean around computers sitting on the floor? Human Resources and managers are the ones who are really the "Official" representatives, little pale faced techies aren't.
One of the greatest parts of the open source creedo is the ability to help others...call centers may suffered a bit because of online forums, but that means less people the corporation has to hire, people who really don't make the company any money. Wouldn't corporations like that better? I think the article's original point sums it up well: "Companies naturally fear litigation, and dislike not being able to sanitize and spin information before it reaches the ears and eyes of their customers."
Only three problems to mention:
(1) "Advanced Linux Sound Architecture" menuconfig option crashes 'make menuconfig'.
(2) Netfilter's 'Owner Match' does not compile as a module.
(3) Possible problem after reboot with 'modprobe char-major-10-135'.
This was all done on a Redhat 7.2 system w/ Create Live! card, 3Com 3c905b card, P3 850 Mhz w/ 256 MB RAM, and a 40 GB ATA5 HDD. The system seemed to speed up a good deal after rebooting into 2.5.40. These errors have been reported to linux-kernel@vger, so hopefully they'll be looked at soon.
It seems 'make menuconfig' has changed in architecture slightly...it's a little bit more logical, and the help sections have been worked out well. I think all in all the Linux kernel should be bumped up to 3.0 just because of all the memory management upgrades and preemptiveness. I know 2.0 came about with ELF, but this is the next big step, IMHO.
Slashdot Mirror
I wonder if these fiber optic roofs will allow people (spy satellites?) to see inside a room when the luminosity inside the room is higher than that outside. Think of it like peering into a house's front windows at night -- as long as the living room lights are on, you can see in, but they can't see out.
What many people fail to realize is that companies like Cisco and Juniper create an OS that interoperates with specialized ASICs that allow a network device to switch/route packets much faster than a traditional time-shared system. While I think the XORP initiative is a good step forward in the way of low-end routers, will XORP give full support to users, for both the hardware and software? If you call Cisco's TAC for support, they will look at your entire system, not just the software and not just the hardware. Companies pay more but they get more in return.
You beat me to the posting, John. :-) We're actually looking at it more from a "Georgia Tech doesn't have a right to regulate the airwaves" standpoint than anything. Direct links to the relevant information are here:
p df
http://headnut.org/files/wnup_fcc_letter.pdf
http://headnut.org/files/wnup_memo_feb19.pdf
http://headnut.org/files/writing_wireless_policy.
If you have $0.02 to share or want to help in this, shoot me an e-mail.
Given that the Linux is the kernel and a distro is just a series of programs combined with that kernel, it seems ridiculous that they tried 9 distros with approximately the same kernel to get the card working. The problem doesn't exist in Linux, either. If the companies who make the cards don't write the drivers, nothing will get done.
He also discovered in the sales-support database... the names of more than 30 [identity-classified] employees of the United States National Security Agency...
While I agree for national security purposes the list of employees should probably be held for review, I do hope a list exists somewhere and is being looked at. I would hate to think this whistleblower is acting like Joseph McCarthy 50-some-odd years ago.
Last I heard, the entire tty layer needed to be rewritten and the scsi interface was broken. This was about mid-May. Given those two facts alone, I would be impressed if 2.6 was done in 9 months, to say the least. Like another thread said, Linus has always stuck with a "when it's done" policy that has helped to make the kernel so stable. Why start releasing stable-versioned kernels with potentially bad/highly experimental code now that Linux is a serious competitor in the server market and beginning to encroach on the desktop market?
I personally like the professors who pile on about 25 hours of work per week, including all those mundane multiple page homeworks and lab write-ups that you know they (or the TAs who can't speak quarter-way decent english) never really read. You know that if you go to the professor to talk about it, nothing will get done and you would have simply wasted another precious hour (or two) of sleep and/or other work.
Perhaps it is because that professor too is tired of dealing with non-english speaking TAs/grad students, or perhaps it is because he actually has to teach students rather than simply letting the herd bask in his infinitely superior knowledge -- a classic case of transference where all the anger, hatred, and frustration is directed to the poor unfortunate souls who inadvertently signed up for his section.
Of course, in retaliation, sites like http://www.engrish.com are written instead of doing those thousands of projects, leaving the ingenious creators with a mere C for the semester. :-)
Would the janitor also count as an official representative if he posted about how to properly clean around computers sitting on the floor? Human Resources and managers are the ones who are really the "Official" representatives, little pale faced techies aren't. One of the greatest parts of the open source creedo is the ability to help others...call centers may suffered a bit because of online forums, but that means less people the corporation has to hire, people who really don't make the company any money. Wouldn't corporations like that better? I think the article's original point sums it up well: "Companies naturally fear litigation, and dislike not being able to sanitize and spin information before it reaches the ears and eyes of their customers."
Only three problems to mention: (1) "Advanced Linux Sound Architecture" menuconfig option crashes 'make menuconfig'. (2) Netfilter's 'Owner Match' does not compile as a module. (3) Possible problem after reboot with 'modprobe char-major-10-135'. This was all done on a Redhat 7.2 system w/ Create Live! card, 3Com 3c905b card, P3 850 Mhz w/ 256 MB RAM, and a 40 GB ATA5 HDD. The system seemed to speed up a good deal after rebooting into 2.5.40. These errors have been reported to linux-kernel@vger, so hopefully they'll be looked at soon. It seems 'make menuconfig' has changed in architecture slightly...it's a little bit more logical, and the help sections have been worked out well. I think all in all the Linux kernel should be bumped up to 3.0 just because of all the memory management upgrades and preemptiveness. I know 2.0 came about with ELF, but this is the next big step, IMHO.