Evidence, please. I would like a citation. I don't like the current administration, but you can't expect to be taken seriously making a charge like that without evidence.
This complaint about MS not securing its own boxen is FUD.
I enjoy MS bashing as much as any Linux and *BSD (and Solaris) user, but this is wrong. The box this document was on was not cracked -- it was fulfilling its intended purpose as a public file server, allowing uploads to MS support and downloads by customers of files left there by the support staff. The security failing was in policy, not a technical problem.
The server did what it was supposed to do -- staff members at MS who put a confidential document on a public server as a way of subverting policies related to internal filesharing.
The boxes that MS puts on the 'net are generally fairly well secured. Most problems are due to people, not systems. (Yes, even though I am a UN*X bigot, I'm saying that it is possible to secure a Windows box on the big bad Interweb. Possible doesn't mean easy.)
There are many, many fine options that are very lightweight by comparison to either of those bloated pigs (I prefer Gnome, generally, but I'm just as happy in WindowMaker or a slim FVWM2 config).
Agreed. The University has been trying to set up a local peering point in Minneapolis, but not all ISPs (esp. Time Warner) seem to be interested in participating. I need to send off another letter to them...:)
Not bloody likely. I'm in Minneapolis, MN. Here's my traceroute to the University of Minnesota:
traceroute to 128.101.101.101 (128.101.101.101), 30 hops max, 38 byte packets
1 xxxxxxxxxxx.rr.com (24.xxx.xxx.xxx) 1.480 ms 1.212 ms 3.600 ms
2 10.y.y.y (10.y.y.y) 12.314 ms 19.837 ms 8.476 ms
3 mplsmn01-rtr2-srp-2-0.mn.rr.com (24.26.162.2) 19.682 ms 9.169 ms 8.995 ms
4 mplsmn01-rtr1-srp-2-0.mn.rr.com (24.26.162.1) 20.112 ms 12.612 ms 12.008 ms
5 pop1-chi-P3-1.atdn.net (66.185.141.89) 28.199 ms 26.546 ms 23.704 ms
6 bb1-chi-P0-0.atdn.net (66.185.141.84) 24.655 ms 25.107 ms 36.789 ms
7 bb1-kcy-P7-0.atdn.net (66.185.152.125) 40.153 ms 38.884 ms 36.182 ms
8 bb2-kcy-P1-0.atdn.net (66.185.152.127) 38.371 ms 71.896 ms 48.152 ms
9 bb2-den-P7-0.atdn.net (66.185.152.188) 48.200 ms 48.099 ms 50.597 ms 10 bb1-den-P1-0.atdn.net (66.185.152.136) 48.182 ms 48.030 ms 56.077 ms 11 bb1-sun-P5-0.atdn.net (66.185.152.253) 74.332 ms 72.269 ms 73.656 ms 12 bb2-sun-P1-0.atdn.net (66.185.152.1) 104.375 ms 73.225 ms 73.054 ms 13 bb2-las-P7-0.atdn.net (66.185.152.22) 79.735 ms 81.554 ms 80.461 ms 14 pop2-las-P1-0.atdn.net (66.185.137.163) 91.439 ms 78.519 ms 92.356 ms 15 aol-gw.la2ca.ip.att.net (192.205.32.101) 98.355 ms 79.452 ms 81.495 ms 16 gbr3-p50.la2ca.ip.att.net (12.123.28.130) 83.982 ms 99.443 ms 93.248 ms 17 gbr4-p20.sffca.ip.att.net (12.122.2.69) 92.254 ms 90.989 ms 112.171 ms 18 gbr3-p50.dvmco.ip.att.net (12.122.2.66) 111.926 ms 110.579 ms 110.642 ms 19 gbr1-p100.dvmco.ip.att.net (12.122.5.18) 115.916 ms 111.989 ms 111.105 ms 20 gar2-p360.dvmco.ip.att.net (12.123.36.137) 111.924 ms 111.556 ms 112.587 ms 21 12.124.158.46 (12.124.158.46) 115.931 ms 120.008 ms 118.364 ms 22 den-core-02.tamerica.net (205.171.16.17) 116.331 ms 117.854 ms 115.497 ms 23 min-core-02.tamerica.net (205.171.8.98) 151.716 ms 141.178 ms 144.119 ms 24 min-edge-01.inet.qwest.net (205.171.128.10) 156.578 ms 141.673 ms 152.590 ms 25 65.121.10.62 (65.121.10.62) 151.691 ms 141.701 ms 242.474 ms 26 tc2-qtr.northernlights.gigapop.net (192.42.152.129) 145.372 ms 144.367 ms 141.991 ms 27 tc3x.router.umn.edu (160.94.26.97) 144.602 ms 143.957 ms 147.239 ms 28 ntc-1-rsmx.rswitch.umn.edu (160.94.26.1) 144.811 ms 148.737 ms 144.713 ms 29 ns.nts.umn.edu (128.101.101.101) 145.145 ms 161.426 ms 144.250 ms
Note: the private network (10.0.0.0/8) is not mine -- it's Time Warner's.
Even in the same state, I'm bouncing through 26 hops to reach the U of MN's border. More to the point, if I'm reading this right, the path on atdn.net is MSP-> Chicago-> Kansas City-> Denver-> sun(?)-> Las Vegas-> L.A.-> San Francisco-> Denver (again)-> Finally, back to Minnesota.
Yes, I do. Not in the text of the constitution, but in several SCOTUS rulings that established the doctrine of prior restraint.
The basic idea is that a restriction on speech, such as the FBI actions in the libraries, will affect my speech and ability to speak in the future. Read the linked page; it explains it better than I can.
It sounds like your biggest problem isn't the cross-platform code as much as the Makefiles and the compiler differences.
I would suggest first using gcc on all UN*X platforms, and also trying out something like ant instead of the various forms of make you're dealing with now.
Also, have you considered using a library like Qt to handle most of the porting details? It's not free, but it is good if you can deal with it's oddities (I personally consider preprocessing to be evil).
The Mac version is not overpriced. It costs more because they have to recoup the cost of driver development for the Macintosh. It's a problem of volume. Their costs for developing the BIOS (OpenFirmware FCODE, most likely) for a Mac are probably similar to the cost of the PC BIOS, possibly more, but there are fewer people buying it.
The end result of your solution (just buy the IA16 card and flash it with the OF BIOS) is that there will not be a OF BIOS available at all, since it will no longer be profitable for ATI to develop it. Think about the consequences of your actions.
Solaris is a hell of a lot better than Linux for NFS services, and a MU for 9 will include NFSv4 (ah, finally, support for ACLs over nfs -- that is, if you're using Solaris.)
I can build a fileserver a hell of a lot cheaper with Dell hardware than Sun hardware, and this lets me run the best OS for the job in my environment.
Nope. Let it go to voice mail and leave the fucking theatre to call them back. People can handle a slight delay in response. I generally don't answer my phone while driving, or even in a conversation with another person who's standing next to me or with me at a restaurant. I'll let it go, then say (based on caller-ID) whether I need to check the message then or later.
Dammit, people, it's not that hard to be polite. You don't need to be reachable immediately at the press of a button all the time.
Aren't SSE ops lower precision? I'm guessing that the original poster was more interested in performance for scientific computing of some kind, where precision matters.
DVD-R will work fine... iff you have the time. Using (as an example) a Exabyte M2, you get 12MB/s raw (uncompressed) writing to the tape. That's 720MB/m, or about 40GB/hour. Each tape is 60G, so I need about 5 tapes to back up my 320G disk fully. I can do that in a day with a single drive, at about $1/G for the tape and (admittedly) $3000 or so for the drive.
Based on what I remember and what I've read, a 2X dvd-r(w) drive will take about 30 minutes to write a 4.7G disk. So, we get about 10G per hour, or 32 hours (and 64 disk switches) to back up our 320G of (non-compressable) data. We'll assume that the data is non-compressable video or music, since that's the most likely uses for this size disk in IDE.
Additional issues with backup to DVD-R include longevity of media -- does anybody really know yet whether I could read my DVD-R I write today in 10 years? I know that I can still read tapes that old, if the format is chosen appropriately. Hell, at work I've got a 9-track drive that gets active use with researchers wanting access to data they collected in the early 80's and have left in a file cabinet since.
I'm not saying that the average user is going to need to back it up in a safe enough way (ie, at least two copies offsite, tested, daily incrementals, etc.) but that is how I look at the problem..
Most of the replies to my original comment have focused on the idea that most people don't need a true backup, but I can't let someone call a RAID-1 array a "backup" without disputing it.
Just remember, when reading this comment, that I'm a paranoid fuck and that's how I look at the world...
However, if I drop my fucking tape, it's still readable. Not so for a hard drive. It's much more sensitive to environment and more fragile than a tape.
A file system as a backup is not very useful in the real world, I'm sorry to tell you. Users never say "I lost this file on this date; please restore it." Instead, it's "Well, I know I had this file last month, but I need the last version of it I edited, which was sometime in the last three weeks. Can you find the last version before I deleted it accidentally?"
Also, just copying data to a drive is wasteful. Why do I want to use 2 drives every time I back up, especially when I'm doing daily backups? The connectors on a hard drive aren't rated for that many changes normally -- I've had power connectors fail on drives that were used for testing in various machines. It's too fragile, and too damned expensive for real work. At home? Sure. But it does not compete with SDLT or LTO, or Mammoth2 for that matter.
If you think that seagate has a good warranty record, you never had to fucking deal with the original Barracudas. I had one conversation that went like this:
Me: "Hi, I have another dead Barracuda. It's a ST12550N, serial number XXXXXX."
Me: "Really. I'm holding it in my hand. That's serial XXXXXX."
ST: "No, that drive is in Singapore in our warehouse by my records. Apparently you have a stolen drive, what was your name again?"
Me: "This fucking drive was a warranty replacement that I got last month, and it's already dead. "
And so forth. This went on for about two hours, with Seagate telling me that it was not possible that I had a drive their computer said was in Singapore.
Over the course of a year, we had over 30 failures of SCSI Barracudas, mostly ST12550N (Yes, I do still remember the model number.) The drive changed several times, giving a different number of sectors with each firmware rev and each warranty replacement, which made it hell to use them in a RAID array (and suicide not to). We had to send two off for replacement at a time, and pray that we got two that had the same number of sectors... and rebuild all of our RAID-1 arrays periodically with new disks just so that we could pair them. Granted, that was mostly the fault of the DPT controllers (PM2122 EISA, with 8M of cache and hardware RAID in 1993. w00t.)
Still, the replacements were sometimes larger, sometimes smaller, and any RAID system would have been fux0r3d by these drives shrinking. As I said, we had over 30 failures, but we only owned 24 drives! I know that Seagate has improved now, and I use their drives again, but it took years.
The point of the rant? Seagate's warranty track record is not outstanding. At all.
("Your honor, I was flat broke but I needed a car, so I just took one from the lot. I felt I was entitled to it because I couldn't afford one and I really needed it.")
You've just made one distinction. By taking that car off the lot, I have deprived another person of their property. If I download a digital copy of "Christie Road" by Green Day, who have I deprived of property? By the same logic, I'm stealing from an author by using a library.
No, I don't know how to fix the industry so that the laws make sense, but keep in mind that copyright was originally 20 years. The classic example is the buggy whip industry and Henry Ford.
"There has grown up in the minds of certain groups in this country the
notion that because a man or a corporation has made a profit out of
the public for a number of years, the government and the courts are
charged with the duty of guaranteeing such profit in the future, even
in the face of changing circumstances and contrary public interest.
This strange doctrine is not supported by statute nor common law.
Neither individuals nor corporations have any right to come into court
and ask that the clock of history be stopped, or turned back, for
their private benefit."
 -- Robert Heinlein
Slashdot Mirror
The guy who wrote this is now officially nominated to join the Order of the Tinfoil Hat.
Evidence, please. I would like a citation. I don't like the current administration, but you can't expect to be taken seriously making a charge like that without evidence.
I enjoy MS bashing as much as any Linux and *BSD (and Solaris) user, but this is wrong. The box this document was on was not cracked -- it was fulfilling its intended purpose as a public file server, allowing uploads to MS support and downloads by customers of files left there by the support staff. The security failing was in policy, not a technical problem.
The server did what it was supposed to do -- staff members at MS who put a confidential document on a public server as a way of subverting policies related to internal filesharing.
The boxes that MS puts on the 'net are generally fairly well secured. Most problems are due to people, not systems. (Yes, even though I am a UN*X bigot, I'm saying that it is possible to secure a Windows box on the big bad Interweb. Possible doesn't mean easy.)
You should patent this idea. Now.
Then, use the patent as a club to prevent other people from doing it! Yes! Victory is ours!
If this idea is placed into use, expect hyperinflation.
X Windows != [KDE|GNOME]
There are many, many fine options that are very lightweight by comparison to either of those bloated pigs (I prefer Gnome, generally, but I'm just as happy in WindowMaker or a slim FVWM2 config).
Think before replying.
Agreed. The University has been trying to set up a local peering point in Minneapolis, but not all ISPs (esp. Time Warner) seem to be interested in participating. I need to send off another letter to them... :)
Not bloody likely. I'm in Minneapolis, MN. Here's my traceroute to the University of Minnesota:
traceroute to 128.101.101.101 (128.101.101.101), 30 hops max, 38 byte packets
1 xxxxxxxxxxx.rr.com (24.xxx.xxx.xxx) 1.480 ms 1.212 ms 3.600 ms
2 10.y.y.y (10.y.y.y) 12.314 ms 19.837 ms 8.476 ms
3 mplsmn01-rtr2-srp-2-0.mn.rr.com (24.26.162.2) 19.682 ms 9.169 ms 8.995 ms
4 mplsmn01-rtr1-srp-2-0.mn.rr.com (24.26.162.1) 20.112 ms 12.612 ms 12.008 ms
5 pop1-chi-P3-1.atdn.net (66.185.141.89) 28.199 ms 26.546 ms 23.704 ms
6 bb1-chi-P0-0.atdn.net (66.185.141.84) 24.655 ms 25.107 ms 36.789 ms
7 bb1-kcy-P7-0.atdn.net (66.185.152.125) 40.153 ms 38.884 ms 36.182 ms
8 bb2-kcy-P1-0.atdn.net (66.185.152.127) 38.371 ms 71.896 ms 48.152 ms
9 bb2-den-P7-0.atdn.net (66.185.152.188) 48.200 ms 48.099 ms 50.597 ms
10 bb1-den-P1-0.atdn.net (66.185.152.136) 48.182 ms 48.030 ms 56.077 ms
11 bb1-sun-P5-0.atdn.net (66.185.152.253) 74.332 ms 72.269 ms 73.656 ms
12 bb2-sun-P1-0.atdn.net (66.185.152.1) 104.375 ms 73.225 ms 73.054 ms
13 bb2-las-P7-0.atdn.net (66.185.152.22) 79.735 ms 81.554 ms 80.461 ms
14 pop2-las-P1-0.atdn.net (66.185.137.163) 91.439 ms 78.519 ms 92.356 ms
15 aol-gw.la2ca.ip.att.net (192.205.32.101) 98.355 ms 79.452 ms 81.495 ms
16 gbr3-p50.la2ca.ip.att.net (12.123.28.130) 83.982 ms 99.443 ms 93.248 ms
17 gbr4-p20.sffca.ip.att.net (12.122.2.69) 92.254 ms 90.989 ms 112.171 ms
18 gbr3-p50.dvmco.ip.att.net (12.122.2.66) 111.926 ms 110.579 ms 110.642 ms
19 gbr1-p100.dvmco.ip.att.net (12.122.5.18) 115.916 ms 111.989 ms 111.105 ms
20 gar2-p360.dvmco.ip.att.net (12.123.36.137) 111.924 ms 111.556 ms 112.587 ms
21 12.124.158.46 (12.124.158.46) 115.931 ms 120.008 ms 118.364 ms
22 den-core-02.tamerica.net (205.171.16.17) 116.331 ms 117.854 ms 115.497 ms
23 min-core-02.tamerica.net (205.171.8.98) 151.716 ms 141.178 ms 144.119 ms
24 min-edge-01.inet.qwest.net (205.171.128.10) 156.578 ms 141.673 ms 152.590 ms
25 65.121.10.62 (65.121.10.62) 151.691 ms 141.701 ms 242.474 ms
26 tc2-qtr.northernlights.gigapop.net (192.42.152.129) 145.372 ms 144.367 ms 141.991 ms
27 tc3x.router.umn.edu (160.94.26.97) 144.602 ms 143.957 ms 147.239 ms
28 ntc-1-rsmx.rswitch.umn.edu (160.94.26.1) 144.811 ms 148.737 ms 144.713 ms
29 ns.nts.umn.edu (128.101.101.101) 145.145 ms 161.426 ms 144.250 ms
Note: the private network (10.0.0.0/8) is not mine -- it's Time Warner's.
Even in the same state, I'm bouncing through 26 hops to reach the U of MN's border. More to the point, if I'm reading this right, the path on atdn.net is MSP-> Chicago-> Kansas City-> Denver-> sun(?)-> Las Vegas-> L.A.-> San Francisco-> Denver (again)-> Finally, back to Minnesota.
Jebus, that sucks.
The basic idea is that a restriction on speech, such as the FBI actions in the libraries, will affect my speech and ability to speak in the future. Read the linked page; it explains it better than I can.
I would suggest first using gcc on all UN*X platforms, and also trying out something like ant instead of the various forms of make you're dealing with now.
Also, have you considered using a library like Qt to handle most of the porting details? It's not free, but it is good if you can deal with it's oddities (I personally consider preprocessing to be evil).
Good luck.
Or better, for safety, 'exec /bin/cat /dev/random'.
The end result of your solution (just buy the IA16 card and flash it with the OF BIOS) is that there will not be a OF BIOS available at all, since it will no longer be profitable for ATI to develop it. Think about the consequences of your actions.
I can build a fileserver a hell of a lot cheaper with Dell hardware than Sun hardware, and this lets me run the best OS for the job in my environment.
I know and understand this. I was making a joke. :-P I guess that jokes do closely resemble trolls...
An "infosec professional" who runs Lookout? Please....
No recompile necessary. Since 6.x.
Of course, I'm a bastage.
Dammit, people, it's not that hard to be polite. You don't need to be reachable immediately at the press of a button all the time.
And funny enough, the Wired article was by the same author... See Politech for more information from this guy.
You need more memory, son. A HALF-MINUTE to pull it out of swap!? Sheeeiit....
Aren't SSE ops lower precision? I'm guessing that the original poster was more interested in performance for scientific computing of some kind, where precision matters.
DVD-R will work fine... iff you have the time. Using (as an example) a Exabyte M2, you get 12MB/s raw (uncompressed) writing to the tape. That's 720MB/m, or about 40GB/hour. Each tape is 60G, so I need about 5 tapes to back up my 320G disk fully. I can do that in a day with a single drive, at about $1/G for the tape and (admittedly) $3000 or so for the drive.
Based on what I remember and what I've read, a 2X dvd-r(w) drive will take about 30 minutes to write a 4.7G disk. So, we get about 10G per hour, or 32 hours (and 64 disk switches) to back up our 320G of (non-compressable) data. We'll assume that the data is non-compressable video or music, since that's the most likely uses for this size disk in IDE.
Additional issues with backup to DVD-R include longevity of media -- does anybody really know yet whether I could read my DVD-R I write today in 10 years? I know that I can still read tapes that old, if the format is chosen appropriately. Hell, at work I've got a 9-track drive that gets active use with researchers wanting access to data they collected in the early 80's and have left in a file cabinet since.
I'm not saying that the average user is going to need to back it up in a safe enough way (ie, at least two copies offsite, tested, daily incrementals, etc.) but that is how I look at the problem..
Most of the replies to my original comment have focused on the idea that most people don't need a true backup, but I can't let someone call a RAID-1 array a "backup" without disputing it.
Just remember, when reading this comment, that I'm a paranoid fuck and that's how I look at the world...
A file system as a backup is not very useful in the real world, I'm sorry to tell you. Users never say "I lost this file on this date; please restore it." Instead, it's "Well, I know I had this file last month, but I need the last version of it I edited, which was sometime in the last three weeks. Can you find the last version before I deleted it accidentally?"
Also, just copying data to a drive is wasteful. Why do I want to use 2 drives every time I back up, especially when I'm doing daily backups? The connectors on a hard drive aren't rated for that many changes normally -- I've had power connectors fail on drives that were used for testing in various machines. It's too fragile, and too damned expensive for real work. At home? Sure. But it does not compete with SDLT or LTO, or Mammoth2 for that matter.
If you think that seagate has a good warranty record, you never had to fucking deal with the original Barracudas. I had one conversation that went like this:
And so forth. This went on for about two hours, with Seagate telling me that it was not possible that I had a drive their computer said was in Singapore.Over the course of a year, we had over 30 failures of SCSI Barracudas, mostly ST12550N (Yes, I do still remember the model number.) The drive changed several times, giving a different number of sectors with each firmware rev and each warranty replacement, which made it hell to use them in a RAID array (and suicide not to). We had to send two off for replacement at a time, and pray that we got two that had the same number of sectors... and rebuild all of our RAID-1 arrays periodically with new disks just so that we could pair them. Granted, that was mostly the fault of the DPT controllers (PM2122 EISA, with 8M of cache and hardware RAID in 1993. w00t.)
Still, the replacements were sometimes larger, sometimes smaller, and any RAID system would have been fux0r3d by these drives shrinking. As I said, we had over 30 failures, but we only owned 24 drives! I know that Seagate has improved now, and I use their drives again, but it took years.
The point of the rant? Seagate's warranty track record is not outstanding. At all.
You've just made one distinction. By taking that car off the lot, I have deprived another person of their property. If I download a digital copy of "Christie Road" by Green Day, who have I deprived of property? By the same logic, I'm stealing from an author by using a library.
No, I don't know how to fix the industry so that the laws make sense, but keep in mind that copyright was originally 20 years. The classic example is the buggy whip industry and Henry Ford.