Do you label your file names in english or with binary numbers. Do you sort them alphabetically or by INODE number? I'm betting you use english. If so then there are hierarchies of linguistic preferences that do apply. Humans read english with case insensitivity, its reasonable to expect a file system to keep its files organized according to human expectations, such as alphabetical listing and case insensitivity.
Apple tech support warns against installing the OS onto UFS. they told me specifically that some of their own utilites will not run on UFS. this has been documented on Macosxhints as well. Thus just to install the OS, you need to have an HFS partition. Other peoples apps also sometimes rely on case sensitivity as well, as do most of my perl scripts.
it wil be interesting to see what happens when apple broaches the unicode file name issue. Will they drop case sensitivity, or retain it for just english users.
Journaling is great and you should use it. It's good for both AOL grandmas and for big disk servers. For the average user it means if the worst happens, say you power down in the middle of a file table update, you disk does not get corrupted. this is good for every user.
for the huge disk servers it means when you power up after a crash you dont have to do a full file system check which could take hours on say a 400GB disk.
what is the cost? a very small amount of disk space (about 8 Megs) and about a 15% reduction in write-to-disk performance. There is no penalty for read performance.
This post is identical to a post I put up this morning on and news leak of this. amazing that this clown copied it verbatim an posted it.
No you cannot mirror 2or more xserves?
on
10.2.2 Is Coming
·
· Score: 2
There's no way to stripe across file systems with apples File system tools. Disk utility only makes raid sets out of the physical disks on the system and does not see disks on other computers.
Sorry you are wrong: raid cannot be partioned
on
10.2.2 Is Coming
·
· Score: 1, Troll
No dumbo, you are the one mounting off. You can't mix partitons on a raid server. Try it before you post.
Re:Raid 5, the missing feature
on
10.2.2 Is Coming
·
· Score: 3, Interesting
Contrary to the two preceeding reply, You CANNOT mix UFS and HFS+ partitions on the same raid drive set using apple tools. The fools who said you could did not actually try to do this else thay would know.
The confusion undoubtedly comes from the fact that you can mix a UFS and HFS+ partition on a single non-raid drive.
But you CANNOT DO IT ON A RAID 1 System. end of story. Note this is a limitation of the apple tools not the RAID system.
Dont you just love the way the fools above insiunuate that the poster is an idiot "mouthing off". Sheesh, what dorks.
Raid 5, the missing feature
on
10.2.2 Is Coming
·
· Score: 5, Informative
Journalling will be great. especially on the disk servers with 480GB worth of storage. But what the Xserves are missing is raid 5. I was pretty upset when I discovered that they only came with raid1 and raid0.
the missing raid mode is worse than it seems. The mac xserves come with 4 big IDE disks. If you want to you want the Xserve to play nice in a unix environment then its a good idea to format the disks UFS. (you dont have to, NFS works fine with HFS+, but you risk screwing yourself with the file name case insensitivity of the mac. A rare event since most people dont have important files that differ in name only in their case but it's lurking.
But wait! you cant format the whole thing UFS becausesome of the mac apps break unless they are on HFS+. So this means you need to format atleast one of the disks HFS for the OS and apps. that leaves three disks. But in RAID 1, you cant use an odd number of disks. So that leaves two disks for raid 1 UFS.
Thus the best you can do is 120GB HFS+ Raid 1 and 120GB UFS Raid 1. So out of four disks the most you can get is 120GB UFS redundant storage. Ah you say, why not just make a small HFS+ partition and let the rest be UFS. Well apple does not yet support partitioning a disk with different File systems. Thus you cant split the disk into UFS and HFS+ partitions.
Two companies are promised a partionalble raid 5 system (Xraid and NXraid) but both suddenly announced delayed shippments. My guess is they are trying to incoporate this new journaling system.
I spoke to apple about this several times. It was hinted to me to keep watching because big things were coming. I suspect these are the Journalling FS and and an outboard mass storage disk sytem. but that's a conjecture.
That's the bad news. The good news is that these Xserves are otherwise a very good deal. The throughput is better than comparably priced linux systems. Also they occupy only 1U but hold 480GB of hot swapable storage. Yes there are some NAS systems that are 1U but they are about 10 X slower in throughput, not to mention that they dont support as many services as the macs (LDAP, NFS, SAMBA, SSH, SCP, FTP, MAIL server, RSYNC,NET info, Net boot...). The macs have dual Gig-E too. ANd in a very nice move Apple will sell you a spare parts kit with everyhing you are likely to need to fix a deadXSERVE in the field.
Plus 24hour tech support.
the other nice thing about the Xserve is the construction. In addition to tool-free hot swap drives, the entire chasis slides out to the front revealing everything with no screws to undo or panels to remove. It's a clever design lacking the usual add-on slider rails of your gneric linux boxes. There's even a firewire port on the front for quick access. Another nice feature is that you dont need a terminal to set them up, they will auotmatically find the administration computer on any DNS system. And if you need to have a terminal attached, you can buy a UPS based KVM switch rather then the usual clumsy Video/mouse/keyboard KVMs.
Anyhow the bottom line is this as soon as a partionalble journaled raid 5 system is avaliable the Xserves will be one of the least expensivie full featured HIGH QUALITY 1U half terrabyte disk servers you can own. (note I said High quality). I just wish they would hurry up since I have two of these cooling their heels waiting for raid 5.
from the alta vista web site:
Fresh, relevant results: AltaVista refreshes 50% of the results daily. Results include PDF files as well as Web pages, images, audio and video files.
How is this possible. Surely you can't poll 50% of the web every day. Nor could you even poll 50% of the spiderable web every day. This seems absurd, but its their number one ranked improvement.
AltaVista Shortcuts and AltaVista Shortcut Answers find results on Web pages that are usually invisible to search engines. (on the U.S. Site only)
Umm does this mean Alta Vista is going to start ignoring ROBOTS.TXT permissions? I dont think they are talking about PDF documents because they called that feature out in a separate bullet. So what is the Invisi-weba dn why do only they have access to it?
Here is the transcript of the whitehouse breifing when poindexter was appointed. Notice that Ari fleschier seemed to think that announcing the gas efficient vehicle promotion was news, poindexter was not.
The James S. Brady Briefing Room
12:45 P.M. EST
MR. FLEISCHER: Good afternoon. I'll give you
a report on the President's day, then I'll be happy to take any
questions you have.
The President this morning had a briefing with the Central
Intelligence Agency and the FBI, to go over the latest developments on
the war on terrorism. On national security matters, convened a meeting
of the National Security Council.
Early this morning, the President held an event on the South Lawn,
where he promoted several energy efficient vehicles that are hybrid
uses of cars. It's part of the President's overall energy
program, which focuses on both conservation and on increased production
of domestic energy sources.
Later today, the President will travel to participate in a live
radio address in honor of the 60th annual celebration of the Voice of
America. That will be an event here in Washington, at the
Voice of America headquarters.
And the final public event on the President's schedule today is in
the East Room, the President will meet with the governors, who are here
for their annual National Governors' Association
conference. The President's remarks will focus on the
efforts the federal government is hoping to help states
with -- homeland security, as well as focusing on education
reform and on welfare reform.
That's a summary of the President's day. Helen.
Q Ari, why would this administration choose
a man for couterterrorism who is so associated with the dark side of
the Iran Contra scandal, Admiral Poindexter?
MR. FLEISCHER: When you say, choose him for
counterterrorism, can you be more specific?
Q He's in the Pentagon, he's been appointed
head of DARPA, which is a counterterrorist office, developing plans,
demonstrations with information.
MR. FLEISCHER: I'm not aware of any appointment.
Q Yet.
MR. FLEISCHER: Let me just say about Admiral Poindexter,
Admiral Poindexter is somebody who this administration thinks is an
outstanding American and an outstanding citizen who has done a very
good job in what he has done for our country, serving in the military.
Q How can you say that, when he told Colonel
North to lie?
MR. FLEISCHER: Helen, I think your views on Iran Contra
are well-known, but the President does believe that Admiral Poindexter
served --
Q It isn't my view, this is the prosecutor
for the United States.
MR. FLEISCHER: I understand. The President
thinks that Admiral Poindexter has served our nation very well.
Q Really?
MR. FLEISCHER: That's the President's thoughts.
Q Do you know his record?
MR. FLEISCHER: I'm sure you will inform me.
Q I don't have to, all you have to do is
look it up.
In deed one of his recent employments was to create a bulliten Board System. His last employer SYNTEK was working on exactly such an information awareness system.
Education
Ph.D., Physics, California Institute of Technology, 1964
M.S., Physics, California Institute of Technology, 1961
B.S., Engineering, U.S. Naval Academy, 1958
Experience Summary
Brings a unique blend of experience to problems from the highest levels of government to
the laboratory. Demonstrated exceptional management and command ability ranging
from naval operations to the national security of the United States. Noted for creative
solutions to difficult issues and ability to quickly grasp the essence of new tasks. Goal
oriented.
Professional Background
DARPA Information Awareness Office, Director 2002 - Present
The mission of the Information Awareness Office (IAO) is to imagine, develop, apply,
integrate, demonstrate and transition information technologies, components, and
prototype closed-loop information systems that will counter asymmetric threats (most
notably, terrorist threats) by achieving total information awareness: enabling preemption;
national security warning; and, national security decision making.
SYNTEK Technologies, Senior Vice President 1996 - 2002
SYNTEK is a small high technology firm with contracts in domestic and international
defense and commercial business. Poindexter was responsible for high-level advice on
management and direction of information systems projects (for example Defense
Advanced Research Project Agency's Project Genoa).
TP Systems, Inc., Co-founder 1990 - 1996
TP Systems was a software development firm specializing in commercial software for the
IBM PC's and compatibles. Poindexter was the chief designer and programmer.
Development included a symbolic debugger for multi-tasking environments, a BBS
communications program plus numerous utility programs.
Elkins Group, Consultant 1993 - 1996
Elkins was a business alliance with Electronic Data Systems (EDS), has developed Elkins
Interactive Training Network (EITN), a satellite based training delivery system that has
world-wide marketing potential. Poindexter was the Chairman of the Maritime Advisory
Committee and a member of the Board of Directors. He also provided advice on strategic
planning.
Presearch, Inc., Senior Scientist 1988 - 1989
Presearch had primarily been involved with defense studies and analysis. Faced with
anticipated defense budget reductions, Poindexter joined the firm to develop new
....
and so on.
No where does he mention his crimes.
You are sort of right. I went and looked it up. True his conviction was overturned on appeal. But the reason was not because he was innocent. Rather the grounds for overturning the conviction were that his testominy before congress had immunized him from prosecution. He is a convicted felon. He just had the conviction overturned but no one, no even poindexter or oliver north, denies he committed a felony.
One of the sad ironies here is that poindexter was a convicted felon. (later pardoned I believe) He was the tip-top head of the IRAN contra scam during the reagan administration. (oliver north's boss who claimed he never informed ronnie about his activities). And of course what did he do in the reagan adminstration. Well for one he tried to destroy all the information linking him to the crime. Destruction of records is a crime. How did he get caught. He did not know the white house e-mail system had backups!
Since he was released from his sentence, Poindexter has been working in private industry for a military consulting firm on EXACTLY the technology being proposed here. Setting aside the ethics of using your position to bebefit your former employer, is poindexter the man to decide what to do?
if we take admiral P at his word, that he never told ronny he was commiting crimes then we know he's a yahoo that cant be trusted to follow orders, laws or safegaurds on private information.
This is in keeping with the Bush admin's policy of appointing inappropriate people.
Sec DOE: Abrahams, only former experience was proposing in congress to eleminate the DOE.
Sec Interior: Gail norton, fromer mining industry lobbist
SEC head: former lobiest/lawyer for big 5 accounting firms.
Sec EPA: Christie Whitman, former govenor of NJ, reportedly accepts donations for chem industry.
Sec Labor: Can you even name the sec labor, do we still have one?
Sec Army: Former Enron top official, accepts free ski vacations from ENRON, then proposes to outsource the Utilities on Military bases to enron.
and so on....
Surprisingly, This lawsuit is good for microsoft in many ways. And in fact the single most advantageos thing microsoft can do is fight this hard and LOSE!. This immesnely strengthens there position in net dominance.
Here's why. First suppose they fight hard and lose. Now the patent is entrenched. This hurts Netscape, Sun, Apple( via quicktime), KDE, and IE equally. In fact its a body blow to all of them. But now what, well two things can happen
Scenario 1: Microsoft Buys the company to get the now validated patent (say for 1 billion), and then puts JAVA and Netscape out of the embedded app bussiness.
Scenario 2: Microsoft is unable to but the patent. But they dont care! why because they have the.NET strategy..NET escapes this patent. And by abandoning IE's EMBED and APPLET tags who do they hurt? JAVA and Quicktime get hurt. MS who does not get hurt. All of their embedables work fine under.NET. And as far as browser wars go, this is great for Microsoft since Netscape wont be able to use Embeds. Everyon will want to use.NET.
MS wins.
Am I crazy? No. Microsoft has already abandon support for the APPLET Tag in windows XP. And they have announced they will not be supporting the Quicktime EMBED's too. Basically they are phasing out everyone eleses Embedable objects as they prepare the way for.NET. This is also why it was critical for MS to say IE was "an integral part of the OS" despite the fact that it wasn't. It's going to be under.NET.
So their optimal strategy is to fight it n the grounds that OLE preceeded this and thus was prior art. If they win, there's also a good chance they could effectively own the patent itself since they would now have shown that EMBED's are covered under their OLE patent. On the other hand if they lose after fighting a good fight, no one elese will have stronger grounds for Prior art claims, the patent will be in force. and as explained above MS wins under that scenrio too.
No. Since most of my code would not altivec optimize I could not use G4s for the cluster. I ended up buying mostly low power P3s (since that way I dont have to cool them) and small number of P4s . The p3s give me bang for the buck when the algorithms parallelize well (e.g. monte carlo calculations). When things dont parallelize well I use (fewer) faster cpus like the xeon P4s. And finally for those few cases where the P4s behave wildly non-optimal I can always use the P3s.
I did buy macs for the desk top front end since they play well with linux and NFS, yet as we all know, have great desktops and run microsoft office.
I think the only mistake I made was buying mac x-serves for some of the disk servers. The xserves appear to be very well built and cost lest than an equivalent quality intel disk server. (note the word equivalent). They definitely can sustain high disk throughput. What I failed to comprehend was that xserve does not yet support raid 5 and one of the disks has to be configured as HFS+. this massively cuts down the usable disk space, especially if youwant redundant raid. as a result I had to get an alternative linux box as a disk server until apple or a 3rd party releases a raid 5 for xserve.
Scientific Benchmarking
on
Is Mac OS X Slow?
·
· Score: 5, Interesting
I recently spent some time evaluating mac,athalon,pentium,athalon benchmaks using scientific fortran that i had written for protein structure analysis. We were buying a large 500 processor cluster so I wanted to get it right.
Since I was buying a cluster my criteria was not single processor speed but speed per dollar what i found was mildy surprising.
For programs that could take advatage of the altivec chip inside the G4, the mac was about a factor of 2 cheaper per run time than the P4 and athalons. On the otherhand with the Altivec turned off the mac was about a factor of 2 more expensive per run time. I note that this was not done on code optimised for the altivec but was just generic fortran passed through an automatic vector pre-processor program for compile time optimization.
Of all the processors I tested, P3, p4, athalon, the P4 had the wildest variations in benchmarking. that is all the other proceesors seemed to have constant scaling factors in speed as the applications varied. but the p4 variev by over a factor of 3 from the others both faster and slower. I assume this has something to do with the very long pipeline, and the hyper threading, and the size of the caches. But even taking these into account I found it highly unpredictable which applications would run faster or slower (that is ones that might logically have more cache misses did not neccessary degrade)
.
In the end I decided the P3 has the most bang for the buck , though falling cpu prices might shift that conclusion to the athalon. The problem I encountered with the athalon was a higher down time for the cluster units due to thermal faliure., so thats a hidden cost. The apples NEVER failed in any thermal tests so thats a hidden plus.
Now this analysis does not factor in other things like Graphics speed other factors more important to users than sceintific apps. However when I compare my molecular visualization grpahics before and after the release of 10.2 I have to say the mac is insanely fast for graphics now wheere before it was intolerably slow.
There seems to be an eternal conflict between what is the province of the OS and the app. The same arguments keep happening. When I first started programming computers in the early 70s unix/vms/prime the OS on the mainframe defined your interaction.
then along came the micros like the altair8800. With these there was no operating system per se. Usually you just loaded interpeted basic from a bootloader and then managed your disk or tape using program you wrote in basic. For example PIP was the name of a program for reading files on a floppy. you loaded pip off tape and then you could access the floppy. when you were done you got rid of PIP. there was no OS only indiviual ad hoc programs. You could say the programming envionment was the OS but givine that your programs were doing peek and poke instructions instead of using an API i'd say there was no OS.
then with the rise of more memory and disks, apple, commodor, trs-80, and IBM started to emulate the mini-computers which had those VMS/CPM/DOS operating systems. then we graduated to Windows and macOS with real APIs.
then what, well along came the Browser, and the idea that one could replace the OS and APIs with a new sort of middleware that would be platform independent. Mosiac and netscape could open text files, and even do many operatiung system inteaction functions like telnet. (KDE extends the metafore to launching programs from a browser inteface).
the other fork was java, which combined a programming language with all the high level functions needed to act like an OS. in other words it was both a language and an API rolled into one. Sun began to talk about how JAVA was really a new kind of OS. you no longer needed to worry about what the computer or OS running underneat was. JAVA was you environment.
so now someone is saying OSs are dead. whoopee. I've seen it before.
Well, it was all pretty sane up to that point; it's nice to know our opponents are still smoking crack -ESR
I'm sorry, but if I was an IT manager of a corporation and if I was reading this commentary I would assume no credibility to ESR after this comment. Grow Up.
I'm sure you would not even raise an eyebrow at the term "pipe dream". Yet of course, that is the very meaning and origin of the term. Smoking opium or crack leads to an irrational overly rosy pipe dream. "smoking crack" is an attmept to update the lexicon in a mildly shocking manner. A bit risque but not rabid.
In the case of Wicard's wheat, the supreme court ruled that if a man was growing his own wheat then obvioulsy he would not be buying any wheat, and his failure to consume affected interstate commerce. Therefore the Feds have the authority to regulate what and how much he can grow. And from there it was all downhill.
Congress's ability to make laws the regulate personal behaviour and practices entirely within a state ALL stem from the constitution's allowance for the feds to regulate inter-state commerce. And this was originally put in the constitution as a sweetener to join the union (i.e joint a free trade zone! much like reason everyone joined the EU or why nafta happened. scary).
Excerpted from www.fff.org:
Enter Roscoe Filburn, an Ohio dairy and poultry farmer, who raised a small quantity of winter wheat -- some to sell, some to feed his livestock, and some to consume. In 1940, under authority of the Agricultural Adjustment Act, the central government told Mr. Filburn that for the next year he would be limited to planting 11 acres of wheat and harvesting 20 bushels per acre. He harvested 12 acres over his allotment for consumption on his own property. When the government fined him, Mr. Filburn refused to pay.
Wickard v. Filburn got to the Supreme Court, and in 1942, the justices unanimously ruled against the farmer. The government claimed that if Mr. Filburn grew wheat for his own use, he would not be buying it -- and that affected interstate commerce. It also argued that if the price of wheat rose, which is what the government wanted, Mr. Filburn might be tempted to sell his surplus wheat in the interstate market, thwarting the government's objective. The Supreme Court bought it.
The Court's opinion must be quoted to be believed:
[The wheat] supplies a need of the man who grew it which would otherwise be reflected by purchases in the open market. Home-grown wheat in this sense competes with wheat in commerce.
As Epstein commented, "Could anyone say with a straight face that the consumption of home-grown wheat is 'commerce among the several states?'" For good measure, the Court justified the obvious sacrifice of Mr. Filburn's freedom and interests to the unnamed farmers being protected:
It is of the essence of regulation that it lays a restraining hand on the self-interest of the regulated and that advantages from the regulation commonly fall to others.
After Wickard , everything is mere detail. The entire edifice of civil rights legislation stands on the commerce power. Under this maximum commerce power, the government has been free to regulate nearly everything, including a restaurant owner's bigotry. The Court has held that if Congress sees a connection to interstate commerce, it is not its role to second guess.
Slashdot Mirror
Do you label your file names in english or with binary numbers. Do you sort them alphabetically or by INODE number? I'm betting you use english. If so then there are hierarchies of linguistic preferences that do apply. Humans read english with case insensitivity, its reasonable to expect a file system to keep its files organized according to human expectations, such as alphabetical listing and case insensitivity.
it wil be interesting to see what happens when apple broaches the unicode file name issue. Will they drop case sensitivity, or retain it for just english users.
for the huge disk servers it means when you power up after a crash you dont have to do a full file system check which could take hours on say a 400GB disk.
what is the cost? a very small amount of disk space (about 8 Megs) and about a 15% reduction in write-to-disk performance. There is no penalty for read performance.
Raid sets cannot be mixed file system partions. your comment only applies to single disk partions which can be mixed. Raid sets cannot
Hi this is Goobah99. No I did not repeat my own post. someone copied it verbatim ans reposted it as an AC. Weird.
This post is identical to a post I put up this morning on and news leak of this. amazing that this clown copied it verbatim an posted it.
There's no way to stripe across file systems with apples File system tools. Disk utility only makes raid sets out of the physical disks on the system and does not see disks on other computers.
you're right.gotta relax.
No dumbo, you are the one mounting off. You can't mix partitons on a raid server. Try it before you post.
The confusion undoubtedly comes from the fact that you can mix a UFS and HFS+ partition on a single non-raid drive. But you CANNOT DO IT ON A RAID 1 System. end of story. Note this is a limitation of the apple tools not the RAID system.
Dont you just love the way the fools above insiunuate that the poster is an idiot "mouthing off". Sheesh, what dorks.
the missing raid mode is worse than it seems. The mac xserves come with 4 big IDE disks. If you want to you want the Xserve to play nice in a unix environment then its a good idea to format the disks UFS. (you dont have to, NFS works fine with HFS+, but you risk screwing yourself with the file name case insensitivity of the mac. A rare event since most people dont have important files that differ in name only in their case but it's lurking.
But wait! you cant format the whole thing UFS becausesome of the mac apps break unless they are on HFS+. So this means you need to format atleast one of the disks HFS for the OS and apps. that leaves three disks. But in RAID 1, you cant use an odd number of disks. So that leaves two disks for raid 1 UFS.
Thus the best you can do is 120GB HFS+ Raid 1 and 120GB UFS Raid 1. So out of four disks the most you can get is 120GB UFS redundant storage. Ah you say, why not just make a small HFS+ partition and let the rest be UFS. Well apple does not yet support partitioning a disk with different File systems. Thus you cant split the disk into UFS and HFS+ partitions.
Two companies are promised a partionalble raid 5 system (Xraid and NXraid) but both suddenly announced delayed shippments. My guess is they are trying to incoporate this new journaling system.
I spoke to apple about this several times. It was hinted to me to keep watching because big things were coming. I suspect these are the Journalling FS and and an outboard mass storage disk sytem. but that's a conjecture.
That's the bad news. The good news is that these Xserves are otherwise a very good deal. The throughput is better than comparably priced linux systems. Also they occupy only 1U but hold 480GB of hot swapable storage. Yes there are some NAS systems that are 1U but they are about 10 X slower in throughput, not to mention that they dont support as many services as the macs (LDAP, NFS, SAMBA, SSH, SCP, FTP, MAIL server, RSYNC,NET info, Net boot ...). The macs have dual Gig-E too. ANd in a very nice move Apple will sell you a spare parts kit with everyhing you are likely to need to fix a deadXSERVE in the field.
Plus 24hour tech support.
the other nice thing about the Xserve is the construction. In addition to tool-free hot swap drives, the entire chasis slides out to the front revealing everything with no screws to undo or panels to remove. It's a clever design lacking the usual add-on slider rails of your gneric linux boxes. There's even a firewire port on the front for quick access. Another nice feature is that you dont need a terminal to set them up, they will auotmatically find the administration computer on any DNS system. And if you need to have a terminal attached, you can buy a UPS based KVM switch rather then the usual clumsy Video/mouse/keyboard KVMs.
Anyhow the bottom line is this as soon as a partionalble journaled raid 5 system is avaliable the Xserves will be one of the least expensivie full featured HIGH QUALITY 1U half terrabyte disk servers you can own. (note I said High quality). I just wish they would hurry up since I have two of these cooling their heels waiting for raid 5.
Fresh, relevant results: AltaVista refreshes 50% of the results daily. Results include PDF files as well as Web pages, images, audio and video files.
How is this possible. Surely you can't poll 50% of the web every day. Nor could you even poll 50% of the spiderable web every day. This seems absurd, but its their number one ranked improvement.
AltaVista Shortcuts and AltaVista Shortcut Answers find results on Web pages that are usually invisible to search engines. (on the U.S. Site only)
Umm does this mean Alta Vista is going to start ignoring ROBOTS.TXT permissions? I dont think they are talking about PDF documents because they called that feature out in a separate bullet. So what is the Invisi-weba dn why do only they have access to it?
this soundslike vacuous hype.
mod up the parent, its a great primary source.
Well she's close. He does head the IAO sector of DARPA. IAO = information awareness office
The James S. Brady Briefing Room
12:45 P.M. EST
MR. FLEISCHER: Good afternoon. I'll give you a report on the President's day, then I'll be happy to take any questions you have.
The President this morning had a briefing with the Central Intelligence Agency and the FBI, to go over the latest developments on the war on terrorism. On national security matters, convened a meeting of the National Security Council.
Early this morning, the President held an event on the South Lawn, where he promoted several energy efficient vehicles that are hybrid uses of cars. It's part of the President's overall energy program, which focuses on both conservation and on increased production of domestic energy sources.
Later today, the President will travel to participate in a live radio address in honor of the 60th annual celebration of the Voice of America. That will be an event here in Washington, at the Voice of America headquarters.
And the final public event on the President's schedule today is in the East Room, the President will meet with the governors, who are here for their annual National Governors' Association conference. The President's remarks will focus on the efforts the federal government is hoping to help states with -- homeland security, as well as focusing on education reform and on welfare reform.
That's a summary of the President's day. Helen.
Q Ari, why would this administration choose a man for couterterrorism who is so associated with the dark side of the Iran Contra scandal, Admiral Poindexter?
MR. FLEISCHER: When you say, choose him for counterterrorism, can you be more specific?
Q He's in the Pentagon, he's been appointed head of DARPA, which is a counterterrorist office, developing plans, demonstrations with information.
MR. FLEISCHER: I'm not aware of any appointment.
Q Yet.
MR. FLEISCHER: Let me just say about Admiral Poindexter, Admiral Poindexter is somebody who this administration thinks is an outstanding American and an outstanding citizen who has done a very good job in what he has done for our country, serving in the military.
Q How can you say that, when he told Colonel North to lie?
MR. FLEISCHER: Helen, I think your views on Iran Contra are well-known, but the President does believe that Admiral Poindexter served --
Q It isn't my view, this is the prosecutor for the United States.
MR. FLEISCHER: I understand. The President thinks that Admiral Poindexter has served our nation very well.
Q Really?
MR. FLEISCHER: That's the President's thoughts.
Q Do you know his record?
MR. FLEISCHER: I'm sure you will inform me.
Q I don't have to, all you have to do is look it up.
In deed one of his recent employments was to create a bulliten Board System. His last employer SYNTEK was working on exactly such an information awareness system.
M.S., Physics, California Institute of Technology, 1961
B.S., Engineering, U.S. Naval Academy, 1958 Experience Summary
Brings a unique blend of experience to problems from the highest levels of government to the laboratory. Demonstrated exceptional management and command ability ranging from naval operations to the national security of the United States. Noted for creative solutions to difficult issues and ability to quickly grasp the essence of new tasks. Goal oriented. Professional Background DARPA Information Awareness Office, Director 2002 - Present
The mission of the Information Awareness Office (IAO) is to imagine, develop, apply, integrate, demonstrate and transition information technologies, components, and prototype closed-loop information systems that will counter asymmetric threats (most notably, terrorist threats) by achieving total information awareness: enabling preemption; national security warning; and, national security decision making.
SYNTEK Technologies, Senior Vice President 1996 - 2002
SYNTEK is a small high technology firm with contracts in domestic and international defense and commercial business. Poindexter was responsible for high-level advice on management and direction of information systems projects (for example Defense Advanced Research Project Agency's Project Genoa).
TP Systems, Inc., Co-founder 1990 - 1996
TP Systems was a software development firm specializing in commercial software for the IBM PC's and compatibles. Poindexter was the chief designer and programmer. Development included a symbolic debugger for multi-tasking environments, a BBS communications program plus numerous utility programs.
Elkins Group, Consultant 1993 - 1996
Elkins was a business alliance with Electronic Data Systems (EDS), has developed Elkins Interactive Training Network (EITN), a satellite based training delivery system that has world-wide marketing potential. Poindexter was the Chairman of the Maritime Advisory Committee and a member of the Board of Directors. He also provided advice on strategic planning.
Presearch, Inc., Senior Scientist 1988 - 1989 Presearch had primarily been involved with defense studies and analysis. Faced with anticipated defense budget reductions, Poindexter joined the firm to develop new
and so on. No where does he mention his crimes.
You are sort of right. I went and looked it up. True his conviction was overturned on appeal. But the reason was not because he was innocent. Rather the grounds for overturning the conviction were that his testominy before congress had immunized him from prosecution. He is a convicted felon. He just had the conviction overturned but no one, no even poindexter or oliver north, denies he committed a felony.
Since he was released from his sentence, Poindexter has been working in private industry for a military consulting firm on EXACTLY the technology being proposed here. Setting aside the ethics of using your position to bebefit your former employer, is poindexter the man to decide what to do?
if we take admiral P at his word, that he never told ronny he was commiting crimes then we know he's a yahoo that cant be trusted to follow orders, laws or safegaurds on private information.
This is in keeping with the Bush admin's policy of appointing inappropriate people.
Sec DOE: Abrahams, only former experience was proposing in congress to eleminate the DOE.
Sec Interior: Gail norton, fromer mining industry lobbist
SEC head: former lobiest/lawyer for big 5 accounting firms.
Sec EPA: Christie Whitman, former govenor of NJ, reportedly accepts donations for chem industry.
Sec Labor: Can you even name the sec labor, do we still have one?
Sec Army: Former Enron top official, accepts free ski vacations from ENRON, then proposes to outsource the Utilities on Military bases to enron.
and so on....
Here's why. First suppose they fight hard and lose. Now the patent is entrenched. This hurts Netscape, Sun, Apple( via quicktime), KDE, and IE equally. In fact its a body blow to all of them. But now what, well two things can happen
Scenario 1: Microsoft Buys the company to get the now validated patent (say for 1 billion), and then puts JAVA and Netscape out of the embedded app bussiness. .NET strategy. .NET escapes this patent. And by abandoning IE's EMBED and APPLET tags who do they hurt? JAVA and Quicktime get hurt. MS who does not get hurt. All of their embedables work fine under .NET. And as far as browser wars go, this is great for Microsoft since Netscape wont be able to use Embeds. Everyon will want to use .NET.
MS wins.
Scenario 2: Microsoft is unable to but the patent. But they dont care! why because they have the
Am I crazy? No. Microsoft has already abandon support for the APPLET Tag in windows XP. And they have announced they will not be supporting the Quicktime EMBED's too. Basically they are phasing out everyone eleses Embedable objects as they prepare the way for .NET. This is also why it was critical for MS to say IE was "an integral part of the OS" despite the fact that it wasn't. It's going to be under .NET.
So their optimal strategy is to fight it n the grounds that OLE preceeded this and thus was prior art. If they win, there's also a good chance they could effectively own the patent itself since they would now have shown that EMBED's are covered under their OLE patent. On the other hand if they lose after fighting a good fight, no one elese will have stronger grounds for Prior art claims, the patent will be in force. and as explained above MS wins under that scenrio too.
I did buy macs for the desk top front end since they play well with linux and NFS, yet as we all know, have great desktops and run microsoft office.
I think the only mistake I made was buying mac x-serves for some of the disk servers. The xserves appear to be very well built and cost lest than an equivalent quality intel disk server. (note the word equivalent). They definitely can sustain high disk throughput. What I failed to comprehend was that xserve does not yet support raid 5 and one of the disks has to be configured as HFS+. this massively cuts down the usable disk space, especially if youwant redundant raid. as a result I had to get an alternative linux box as a disk server until apple or a 3rd party releases a raid 5 for xserve.
Since I was buying a cluster my criteria was not single processor speed but speed per dollar what i found was mildy surprising. For programs that could take advatage of the altivec chip inside the G4, the mac was about a factor of 2 cheaper per run time than the P4 and athalons. On the otherhand with the Altivec turned off the mac was about a factor of 2 more expensive per run time. I note that this was not done on code optimised for the altivec but was just generic fortran passed through an automatic vector pre-processor program for compile time optimization.
Of all the processors I tested, P3, p4, athalon, the P4 had the wildest variations in benchmarking. that is all the other proceesors seemed to have constant scaling factors in speed as the applications varied. but the p4 variev by over a factor of 3 from the others both faster and slower. I assume this has something to do with the very long pipeline, and the hyper threading, and the size of the caches. But even taking these into account I found it highly unpredictable which applications would run faster or slower (that is ones that might logically have more cache misses did not neccessary degrade)
. In the end I decided the P3 has the most bang for the buck , though falling cpu prices might shift that conclusion to the athalon. The problem I encountered with the athalon was a higher down time for the cluster units due to thermal faliure., so thats a hidden cost. The apples NEVER failed in any thermal tests so thats a hidden plus.
Now this analysis does not factor in other things like Graphics speed other factors more important to users than sceintific apps. However when I compare my molecular visualization grpahics before and after the release of 10.2 I have to say the mac is insanely fast for graphics now wheere before it was intolerably slow.
then along came the micros like the altair8800. With these there was no operating system per se. Usually you just loaded interpeted basic from a bootloader and then managed your disk or tape using program you wrote in basic. For example PIP was the name of a program for reading files on a floppy. you loaded pip off tape and then you could access the floppy. when you were done you got rid of PIP. there was no OS only indiviual ad hoc programs. You could say the programming envionment was the OS but givine that your programs were doing peek and poke instructions instead of using an API i'd say there was no OS.
then with the rise of more memory and disks, apple, commodor, trs-80, and IBM started to emulate the mini-computers which had those VMS/CPM/DOS operating systems. then we graduated to Windows and macOS with real APIs.
then what, well along came the Browser, and the idea that one could replace the OS and APIs with a new sort of middleware that would be platform independent. Mosiac and netscape could open text files, and even do many operatiung system inteaction functions like telnet. (KDE extends the metafore to launching programs from a browser inteface).
the other fork was java, which combined a programming language with all the high level functions needed to act like an OS. in other words it was both a language and an API rolled into one. Sun began to talk about how JAVA was really a new kind of OS. you no longer needed to worry about what the computer or OS running underneat was. JAVA was you environment.
so now someone is saying OSs are dead. whoopee. I've seen it before.
Well, it was all pretty sane up to that point; it's nice to know our opponents are still smoking crack -ESR
I'm sorry, but if I was an IT manager of a corporation and if I was reading this commentary I would assume no credibility to ESR after this comment. Grow Up.
I'm sure you would not even raise an eyebrow at the term "pipe dream". Yet of course, that is the very meaning and origin of the term. Smoking opium or crack leads to an irrational overly rosy pipe dream. "smoking crack" is an attmept to update the lexicon in a mildly shocking manner. A bit risque but not rabid.
Congress's ability to make laws the regulate personal behaviour and practices entirely within a state ALL stem from the constitution's allowance for the feds to regulate inter-state commerce. And this was originally put in the constitution as a sweetener to join the union (i.e joint a free trade zone! much like reason everyone joined the EU or why nafta happened. scary).
Excerpted from www.fff.org: Enter Roscoe Filburn, an Ohio dairy and poultry farmer, who raised a small quantity of winter wheat -- some to sell, some to feed his livestock, and some to consume. In 1940, under authority of the Agricultural Adjustment Act, the central government told Mr. Filburn that for the next year he would be limited to planting 11 acres of wheat and harvesting 20 bushels per acre. He harvested 12 acres over his allotment for consumption on his own property. When the government fined him, Mr. Filburn refused to pay. Wickard v. Filburn got to the Supreme Court, and in 1942, the justices unanimously ruled against the farmer. The government claimed that if Mr. Filburn grew wheat for his own use, he would not be buying it -- and that affected interstate commerce. It also argued that if the price of wheat rose, which is what the government wanted, Mr. Filburn might be tempted to sell his surplus wheat in the interstate market, thwarting the government's objective. The Supreme Court bought it. The Court's opinion must be quoted to be believed: [The wheat] supplies a need of the man who grew it which would otherwise be reflected by purchases in the open market. Home-grown wheat in this sense competes with wheat in commerce. As Epstein commented, "Could anyone say with a straight face that the consumption of home-grown wheat is 'commerce among the several states?'" For good measure, the Court justified the obvious sacrifice of Mr. Filburn's freedom and interests to the unnamed farmers being protected: It is of the essence of regulation that it lays a restraining hand on the self-interest of the regulated and that advantages from the regulation commonly fall to others. After Wickard , everything is mere detail. The entire edifice of civil rights legislation stands on the commerce power. Under this maximum commerce power, the government has been free to regulate nearly everything, including a restaurant owner's bigotry. The Court has held that if Congress sees a connection to interstate commerce, it is not its role to second guess.