1. FreeBSD is not a commercial OS, BSDi is the commercial version.
2. Market share? Qualify this please! Is this commercial market share or is market share being judged be Netcraft surveys? Or worst yet, MicroSoft press?
3. FreeBSD DID NOT OWN walnut Creek. Walnut Creek is a major distributer of FreeBSD. Walnut Creek is also a major distributor of Linux distributions. By your conclusion, Slackware would have also owned Walnut Creek. This is just plain incorrect. Incidentally, Walnut creek = ftp.cdrom.com (the largest public FTP site on the Internet, supporting 6000 simultaneous users, on FreeBSD OS).
4. Given your conclusion that "hobbyists" use *BSD, please explain the use of FreeBSD by:
Hotmail.com
yahoo.com
cdrom.com
Then factor in press articles such as:
http://www.webtechniques.com/archives/2001/01/in fr revu/
http://www.byte.com/column/BYT20010130S0010
And the numerous others at
http://www.FreeBSD.org/news/press.html
A while back there was a message thread on the FreeBSD Stable mailing list dealing with a comparison of heat output (Linux vs FreeBSD) on the same SMP enabled box. Discussion got down into the internals of when to sleep a processor.
Snatched from thread:
"Under SMP, there is more than one cpu. If one cpu is asleep in a HLT, and the other one changes the run queues, there is nothing to wake up the sleeping CPU even though there is now a job in the queue. So.. we let the cpu spin looking at the "queue not empty" bits so it never gets caught sleeping on the job." Peter Wemm peter@netplex.com.au
Thread is a little dated but quite relevant to why power management is not a good idea on server class machines.
FreeBSD -Stable messsage
Good coverage but you missed two big points that might sway the vote. (Election time. Cheesy, I know.)
This issue originally evolved around the Respondent having registered Guinness.com (can the class say "cybersquatting"?) and losing that domain to Guinness. Respondent then registers GuinnessSucks.com and variations in retaliation. Guinness learns of the GuinnessSucks.com and variations and files a complaint with the WIPO. August 25, 2000 the Respondent is sent (e-mail, fax and snail-mail) a "Notification of Complaint and Commencement of Administrative Proceedings" and that "the Respondent was required to submit a Response to the Center on or before September 13, 2000." No response was received and on September 25, 2000 the Respondent was sent Notification of Respondent Default.
To summerize: If you don't even bother to field a team on game day, you are assured only one outcome. Bemoan the fact that the "Respondent" couldn't be troubled to not even contest the complaint, not that Guinness won.
To summerize further: Two wrongs don't make a right, but three lefts do.
Slashdot Mirror
Several problems with your conclusions:
n fr revu/
1. FreeBSD is not a commercial OS, BSDi is the commercial version.
2. Market share? Qualify this please! Is this commercial market share or is market share being judged be Netcraft surveys? Or worst yet, MicroSoft press?
3. FreeBSD DID NOT OWN walnut Creek. Walnut Creek is a major distributer of FreeBSD. Walnut Creek is also a major distributor of Linux distributions. By your conclusion, Slackware would have also owned Walnut Creek. This is just plain incorrect. Incidentally, Walnut creek = ftp.cdrom.com (the largest public FTP site on the Internet, supporting 6000 simultaneous users, on FreeBSD OS).
4. Given your conclusion that "hobbyists" use *BSD, please explain the use of FreeBSD by:
Hotmail.com
yahoo.com
cdrom.com
Then factor in press articles such as:
http://www.webtechniques.com/archives/2001/01/i
http://www.byte.com/column/BYT20010130S0010
And the numerous others at
http://www.FreeBSD.org/news/press.html
Thanks for all the fish,
Anonymous
A while back there was a message thread on the FreeBSD Stable mailing list dealing with a comparison of heat output (Linux vs FreeBSD) on the same SMP enabled box. Discussion got down into the internals of when to sleep a processor.
Snatched from thread: "Under SMP, there is more than one cpu. If one cpu is asleep in a HLT, and the other one changes the run queues, there is nothing to wake up the sleeping CPU even though there is now a job in the queue. So.. we let the cpu spin looking at the "queue not empty" bits so it never gets caught sleeping on the job." Peter Wemm peter@netplex.com.au
Thread is a little dated but quite relevant to why power management is not a good idea on server class machines.
FreeBSD -Stable messsage
This issue originally evolved around the Respondent having registered Guinness.com (can the class say "cybersquatting"?) and losing that domain to Guinness. Respondent then registers GuinnessSucks.com and variations in retaliation. Guinness learns of the GuinnessSucks.com and variations and files a complaint with the WIPO. August 25, 2000 the Respondent is sent (e-mail, fax and snail-mail) a "Notification of Complaint and Commencement of Administrative Proceedings" and that "the Respondent was required to submit a Response to the Center on or before September 13, 2000." No response was received and on September 25, 2000 the Respondent was sent Notification of Respondent Default.
To summerize: If you don't even bother to field a team on game day, you are assured only one outcome. Bemoan the fact that the "Respondent" couldn't be troubled to not even contest the complaint, not that Guinness won.
To summerize further: Two wrongs don't make a right, but three lefts do.