Anyhow, Deep Blue is a Kasparov-killing machine, not a general chess-playing machine. It was tuned to Kasparov's game and no-one else's.
But this is not that different from what human contenders to the title do. They spend months training to beat the current title holder.
Kasparov made two big mistakes one was to agree to play Deep Blue without being allowed to see it play a couple of games before hand. The second was to train against a simpler program, expecting deep blue to behave the same, but faster. This is akin to say that Kasparov and I play the same chess, only that he plays faster (I only wish)....
My mom (she was in highschool when sputnik was launched) told me about pre Gagarin flights for years. Basically the story goes, that you could pick up on a ham radio the radio transmissions of from the doomed cosmonauts.
I read such an article in an old Readers Digest that I picked up at a garage sale. Now, you must keep in mind that during that time the Readers Digest was actively collaborating with the CIA to stop the "red menace".
A quick glance at the rest of the magazine articles illustrated that point handsomely. Most of the stories were peppered with anti-soviet propaganda. To give an example, in the middle of a story that had nothing to do with politics such as a tourist excursion to Texas, they would sneak a comment like "after eating that spicy taco, I felt worse than a prisoner in a soviet gulag"...
But looking at Linux technically, there is no real revolution here. Linux looks and feels like UNIX and isn't any better than a commercial version of UNIX.
While there might be a bit of hyperbole in that statement, so far what Linux has done is mostly replication of UNIX. This is no mean feat, considering how much work it involved. But Linux must definitely move beyond that and start innovating.
Can it support a better development paradigm (see OS X)? Can it provide by default a more powerful security model (a la Kerberos)? Can we at long last ditch NFS into the trashbin of history and replace it with a decent and secure file server? Could we get rid of X windows, the worst UI to adorn a desktop since Windows 2.0?
Every time there's a story about 2001, we see two basic types of flame. The first is "What's so great about 10 minutes of flashing colors?" I'm going to respond to the first in the same way I usually do. Deal with it. There's a fast-forward button. If you don't like the flashing colors, press the button.
Or see it the way it was meant to be seen: on a large movie screen (and preferably while doing acid:-).
It really irks me when somebody pans a movie with excellent photography/special effects after seeing it on the small screen.
I'm somewhat familiar with web search engine technology since its inception. Over these six years the quality of the results has gone up while the precentage coverage has remained steady.
So the main premise of the article is moot.
Also their data is flawed. The article quotes a bogus 550 billion pages which includes dynamic content not meant to be indexed. If we used a realistic definition of what a web page is, the total number of pages out there would rate in the 10-30 billion, tops.
As computational linguistics improve, as well as usage pattern agents such as copernicus and firefly are refined I expect the quality of searches to continue improving.... Just recently I came across an article demonstrating amazing automated content subject classification (coming soon to a search engine near you).
Are these "the sky is falling" printed press articles the forerunner of trolls?
Re:rootness and capabilities
on
New Linux Worm
·
· Score: 2
Breaks for cars are essential, we both agree, but having the backspace/delete keys work properly is an add-on that according to you shouldn't work out of the box...
We are not talking here about some specialized mathematical simulation software. We are talking about the security model of an OS which is touted as the system of choice for web servers, or the delete/backspace keys, which, last I checked, are used often.
To you those things are optional, which only confirms my point: when Linux sucks, the dittoheads copout with "just add package XYZ".
Re:rootness and capabilities
on
New Linux Worm
·
· Score: 2
Cars require breaks in order to function safely and effectively.
I see now the error of my ways. Car require breaks, but an OS which is touted as the best medium size web server available does not require decent security or a decent file system... No siree, you need to download it from some place else, after all why would a web server need to be secure?
Re:rootness and capabilities
on
New Linux Worm
·
· Score: 2
ctually, mainframes didn't get those sorts of attributes until the 70's
Ah ok. That makes it alright then.... (NOT)
If you want compartmentalization, ACLs, a privacy model, malcode capabilities, etc., then go to http://www.rsbac.org, patch your kernel and stop bitching.
Predictable. The usual copout of the Linux user: Just download pacakge XYZ.
Yeap, when you buy a car and it has no breaks, you don't go to the dealer and complain. No you simply walk over to Napa spare parts and download some new brakes. After all why should one assume that things should work out of the box?
That is how high the quality bar has been set by Linux dittoheads: if it doesn't work out of the box is your fault.
Re:rootness and capabilities
on
New Linux Worm
·
· Score: 2
A more legitimate qualm with the *nix model is that it is coarse-grained.
That is why it sucks, because it is too coarse grained.
The X Window System catches a lot of criticism, some of it well-deserved.
The biggest problem with X-windows is that it requires a powerful and intelligent terminal which then is treated like a dumb device. OS X has improved on this. (I gave up on berlin when they spent a few months implementing alpha transparency.)
What better interface can you get than actually getting the CLI in the GUI?
The CLI is completely unaware that there is a GUI out there. See XMLterm for the proper way to create a CLI inside a graphical user interface.
They seem to have everything I need and want, and more.
The main shells are missing a ton of things. Here's a simple one: Not remembering recently used files without full path qualification (something norton commander supported ten years ago). Here's another one: default configuration often sucks. I've used many and the default shell often has file completion off and history off. Not to talk about the whole backspace/delete rigamarole. Imagine what you would say if Microsoft Word started with the delete key disabled...
File system sucks
Are you arguing against heirarchical file systems or against the file systems themselves?
What I'm refering to here is the lack of user defined attributes on the file system, such as "this file can only be opened with application XYZ". Mainframes had those in the 60's, WinNT has user defined attributes, how long until *nix supports those by default?
But, eh, keep complaining: anything that gets me new toys to play with can't be too bad.
That's the point. Create an itch, then address it.
Re:rootness and capabilities
on
New Linux Worm
·
· Score: 3
If people stopped giving root God-like powers then problems like this wouldn't crop up.
This is one way in which Linux/Unix sucks. The security model is brain dead. It might look good compared to Windoze, but if you have ever used a mainframe you would know what I'm talking about.
Yet the Linux community seems more interested in pointing out the ways in which Linux is better than Windows instead of adressing real concerns with the *nix model... (Miguel de Icaza being the exception that proves the rule).
Here's a list
Security in *nix sucks
X-windows sucks
the xterm gui-cli interface sucks
all the shells suck (with the possible exception of zsh).
How many residents of third world countries can afford a $1500 computer?
In 1988 as a third world resident, I worked for a year saving every penny (aside from food money). At the end I was the owner of a 4.66 Turbo XT with dual floppy drives.
I'd stayed up many a night hacking (i.e. writing code) on that box...
It might be more than a coincidence that Miguel de Icaza (from Gnome fame) was a highly placed computer consultant for the political party from which the City Major belongs to.
Geeks tend to constantly overestimate their own ability in what amounts to severe egotism. They assume that if someone else could create something, then surely it wouldn't been no trouble to them.
The XOR patent for cursor display was the favourite whipping boy of the anti-software patent movement (you can still find the odd reference here and there). XOR-ing cursors was according to the anti-software patent folks the "obvious" and "natural way" to do it.
Then at some point a researcher took the care of examining how exactly cursor display was done in all other graphic systems in 1978 when the XOR patent application was filed (there weren't that many) and all those he could find used a copy buffer and not the "obvious" XOR idea.
IIRC, the XOR idea was so quickly adopted that by 1981 it was prevalent and even tought in school. But no documented use before 1980 or so could be found.
In 1989 corporations with deep pockets challenged the XOR patent. It was duly reexamined and held valid by the PTO.
It does protect the author from someone copying their work verbatim and passing it off as their own, but it certainly doesn't protect them from another author having a Troll/Elf fight with words were the Elf is victorious, in their own book.
Neither should a patent. Until recently you could only patent specific implementations. Small variations on a discovery would circumvent the patent, just like small variations on a plot would circumvent copyright...
All this argues for is a narrow interpretation of what is patented be it software or hardware. This is not an argument against software patents per se.
There are a number of reasons why people are opposed to software patents. For one, it increases the barrier to entry into the software market.
Indeed, but that applies equally to the wooden widget market. Yet you don't see anyone here complaining about the barriers to entry to the wooden widget market due to patents.
The second reason there's a lot of opposition to patents is the fact they last 21 years, and in the computer industry, the landscape can drastically change in 6 months. In that light, software patents that last 21 years seems like overkill.
I agree with this one. All this argues for is a smaller time lapse for software patents with which, again, I agree 100%. But again the solution to patents that last too long is not no patents whatsoever, but rather shorter patents for software (in the same spirit, but in opposite effect to patents which are *extended* for drugs with long clinic trials).
The most recent indication of this seems to be Rambus, Inc and the JEDEC forum, where Rambus didn't feel it was nessecary to inform participants of their pending patent on SDRAM, and DDR technologies.
This is not-exclusive to software. In fact the patent you have in mind is a half about hardware, so once again, if anything this an argument against *all* patents, not simply software ones.
The most frustrating part, however, is the fact that patents aren't usually used to help independent inventors. In most cases, patents are obtained by big businesses to either cripple their competition, or to raise the barrier to entry to prevent new competitors.
Again I agree, and again this is not exclusive to software patents. Yet another case of a clear flaw in the patent system (like stupid patents) which is ascribed to *software* patents when in reality aflicts the entire patent system.
I must be suicidal, defending patents in/. Already lost karma to some loser who called my previous reasoned posting flame-bait... Oh well is just stupid karma points....
So, by your argument, I should be able to labor for a year in my garage, write a really good book using some neat plot devices, and then prevent anybody else from using those plot devices for seventeen years?
In fact you are. Is called copyright and it lasts a lot longer than 17 years....
If you spend a year in your garage working on a project and come out with an innovative wooden widget you can walk to the patent office and obtain a 17 year monopoly on it.
If on the other hand you spend a year in your garage developing some innovative thing that had not been done before, but happens to be in software, many/.'ers propose that you shouldn't be able to patent it.
Why the disparity? Here are some possible reasons:
The majority of/.ers here are developers as opposed to inventors, and therefore patents would make their life harder.
Knee-jerk reaction against the establishment from a very independent crowd.
Misguided opposition to stupid patents which is translated to "no patents" instead of "no stupid patents".
My browser (NS4.75) core-dumped on me. Here it goes again:
From personal experience as an employeer, we tend to ask co-ops to perform as difficult a job as they are technically qualified to do.
Having said that, every employee, co-op or not, will find him/herself performing boring tasks very so often.
As Dilbert says, this is the reason why the pay you: if work was all fun they would charge admission prices at the entrance!
This is not to say that there aren't unenlightened co-op employers out there who think of co-op students as under-aged janitors, as opposed to trainees, but from what you write yours doesn't seem to be one.
The schools I'm familiar with keep tabs on co-op employers, and if they treat their students as sources of menial labor with no training component the employer gets an earful.
From personal experience as an employeer, we tend to ask co-ops to perform as difficult a job as they are technically qualified to do.
Having said that, every employee, co-op or not, will find him/herself performing boring tasks very so often.
As Dilbert says, this is the reason why the pay you: if work was all fun they would charge admission prices at the entrance!
This is not to say that there aren't unenlightened co-op employers out there who think of co-op students as under-aged janitors, as opposed to trainees, but from what you write yours doesn't seem to be one.
The schools I'm familiar with keep tabs on co-op employers, and if they treat their students as sources of menial labor with no training component the employer gets an earful.
Slashdot readers seem more interested in wonders of technology rather than wonders of nature.
The pictures from the website are amazing, but there is no way in which us slashdot geeks can take pride on that. Now, crashing aluminum foil on an asteroid is something any os could have worked on, thus its clear interest to this audience...
(20 points) Cite evidence that ancient Egyptian society was composed of Africans rather than Caucasians, and explain the impact of this anthropological theory.
Everybody knows that ancient Egyptians were Caucasians, in fact I remember quite distinctively from my history lessons that Cleopatra had blue eyes and was married to Richard Burton.
Slashdot Mirror
For internet TV to be useful it would have to be (a) on demand and (b) downloadable to the TV set , not the computer monitor.
But this is not that different from what human contenders to the title do. They spend months training to beat the current title holder.
Kasparov made two big mistakes one was to agree to play Deep Blue without being allowed to see it play a couple of games before hand. The second was to train against a simpler program, expecting deep blue to behave the same, but faster. This is akin to say that Kasparov and I play the same chess, only that he plays faster (I only wish)....
I read such an article in an old Readers Digest that I picked up at a garage sale. Now, you must keep in mind that during that time the Readers Digest was actively collaborating with the CIA to stop the "red menace".
A quick glance at the rest of the magazine articles illustrated that point handsomely. Most of the stories were peppered with anti-soviet propaganda. To give an example, in the middle of a story that had nothing to do with politics such as a tourist excursion to Texas, they would sneak a comment like "after eating that spicy taco, I felt worse than a prisoner in a soviet gulag"...
While there might be a bit of hyperbole in that statement, so far what Linux has done is mostly replication of UNIX. This is no mean feat, considering how much work it involved. But Linux must definitely move beyond that and start innovating.
Can it support a better development paradigm (see OS X)? Can it provide by default a more powerful security model (a la Kerberos)? Can we at long last ditch NFS into the trashbin of history and replace it with a decent and secure file server? Could we get rid of X windows, the worst UI to adorn a desktop since Windows 2.0?
Or see it the way it was meant to be seen: on a large movie screen (and preferably while doing acid :-).
It really irks me when somebody pans a movie with excellent photography/special effects after seeing it on the small screen.
I downloaded the beta version and it works! It even timestamps the files properly. This is amazing!
I'm somewhat familiar with web search engine technology since its inception. Over these six years the quality of the results has gone up while the precentage coverage has remained steady.
So the main premise of the article is moot.
Also their data is flawed. The article quotes a bogus 550 billion pages which includes dynamic content not meant to be indexed. If we used a realistic definition of what a web page is, the total number of pages out there would rate in the 10-30 billion, tops.
As computational linguistics improve, as well as usage pattern agents such as copernicus and firefly are refined I expect the quality of searches to continue improving.... Just recently I came across an article demonstrating amazing automated content subject classification (coming soon to a search engine near you).
Are these "the sky is falling" printed press articles the forerunner of trolls?
We are not talking here about some specialized mathematical simulation software. We are talking about the security model of an OS which is touted as the system of choice for web servers, or the delete/backspace keys, which, last I checked, are used often.
To you those things are optional, which only confirms my point: when Linux sucks, the dittoheads copout with "just add package XYZ".
I see now the error of my ways. Car require breaks, but an OS which is touted as the best medium size web server available does not require decent security or a decent file system... No siree, you need to download it from some place else, after all why would a web server need to be secure?
Ah ok. That makes it alright then.... (NOT)
If you want compartmentalization, ACLs, a privacy model, malcode capabilities, etc., then go to http://www.rsbac.org, patch your kernel and stop bitching.
Predictable. The usual copout of the Linux user: Just download pacakge XYZ.
Yeap, when you buy a car and it has no breaks, you don't go to the dealer and complain. No you simply walk over to Napa spare parts and download some new brakes. After all why should one assume that things should work out of the box?
That is how high the quality bar has been set by Linux dittoheads: if it doesn't work out of the box is your fault.
That is why it sucks, because it is too coarse grained.
The X Window System catches a lot of criticism, some of it well-deserved.
The biggest problem with X-windows is that it requires a powerful and intelligent terminal which then is treated like a dumb device. OS X has improved on this. (I gave up on berlin when they spent a few months implementing alpha transparency.)
What better interface can you get than actually getting the CLI in the GUI?
The CLI is completely unaware that there is a GUI out there. See XMLterm for the proper way to create a CLI inside a graphical user interface.
They seem to have everything I need and want, and more.
The main shells are missing a ton of things. Here's a simple one: Not remembering recently used files without full path qualification (something norton commander supported ten years ago). Here's another one: default configuration often sucks. I've used many and the default shell often has file completion off and history off. Not to talk about the whole backspace/delete rigamarole. Imagine what you would say if Microsoft Word started with the delete key disabled...
File system sucks
Are you arguing against heirarchical file systems or against the file systems themselves?
What I'm refering to here is the lack of user defined attributes on the file system, such as "this file can only be opened with application XYZ". Mainframes had those in the 60's, WinNT has user defined attributes, how long until *nix supports those by default?
But, eh, keep complaining: anything that gets me new toys to play with can't be too bad.
That's the point. Create an itch, then address it.
This is one way in which Linux/Unix sucks. The security model is brain dead. It might look good compared to Windoze, but if you have ever used a mainframe you would know what I'm talking about.
Yet the Linux community seems more interested in pointing out the ways in which Linux is better than Windows instead of adressing real concerns with the *nix model... (Miguel de Icaza being the exception that proves the rule).
Here's a list
- Security in *nix sucks
- X-windows sucks
- the xterm gui-cli interface sucks
- all the shells suck (with the possible exception of zsh).
- file system in *nix sucks
- netscape in *nix sucks
any others?Flame away
In 1988 as a third world resident, I worked for a year saving every penny (aside from food money). At the end I was the owner of a 4.66 Turbo XT with dual floppy drives.
I'd stayed up many a night hacking (i.e. writing code) on that box...
Shareware, publicly available in source code format.
# Filename: vote1.cgi
# Version: 3.5
# Author: Richard Ian-Frese
# e-mail: rif@u.washington.edu
# www: http://staff.washington.edu/rif
# Date created: 17-Nov-95
# Last updated: 04-Mar-99
# Tested running: Perl v4.036, v5.003
# Server protocol: HTTP/1.0
# Server software: NCSA v1.4.2, Apache v1.1.1, NT v4.00
It might be more than a coincidence that Miguel de Icaza (from Gnome fame) was a highly placed computer consultant for the political party from which the City Major belongs to.
The XOR patent for cursor display was the favourite whipping boy of the anti-software patent movement (you can still find the odd reference here and there). XOR-ing cursors was according to the anti-software patent folks the "obvious" and "natural way" to do it.
Then at some point a researcher took the care of examining how exactly cursor display was done in all other graphic systems in 1978 when the XOR patent application was filed (there weren't that many) and all those he could find used a copy buffer and not the "obvious" XOR idea.
IIRC, the XOR idea was so quickly adopted that by 1981 it was prevalent and even tought in school. But no documented use before 1980 or so could be found.
In 1989 corporations with deep pockets challenged the XOR patent. It was duly reexamined and held valid by the PTO.
Don Knuth solved a math problem at the beginning of the school year in High School. He got an A on the course and was excused from any extra work.
Neither should a patent. Until recently you could only patent specific implementations. Small variations on a discovery would circumvent the patent, just like small variations on a plot would circumvent copyright...
All this argues for is a narrow interpretation of what is patented be it software or hardware. This is not an argument against software patents per se.
Indeed, but that applies equally to the wooden widget market. Yet you don't see anyone here complaining about the barriers to entry to the wooden widget market due to patents.
The second reason there's a lot of opposition to patents is the fact they last 21 years, and in the computer industry, the landscape can drastically change in 6 months. In that light, software patents that last 21 years seems like overkill.
I agree with this one. All this argues for is a smaller time lapse for software patents with which, again, I agree 100%. But again the solution to patents that last too long is not no patents whatsoever, but rather shorter patents for software (in the same spirit, but in opposite effect to patents which are *extended* for drugs with long clinic trials).
The most recent indication of this seems to be Rambus, Inc and the JEDEC forum, where Rambus didn't feel it was nessecary to inform participants of their pending patent on SDRAM, and DDR technologies.
This is not-exclusive to software. In fact the patent you have in mind is a half about hardware, so once again, if anything this an argument against *all* patents, not simply software ones.
The most frustrating part, however, is the fact that patents aren't usually used to help independent inventors. In most cases, patents are obtained by big businesses to either cripple their competition, or to raise the barrier to entry to prevent new competitors.
Again I agree, and again this is not exclusive to software patents. Yet another case of a clear flaw in the patent system (like stupid patents) which is ascribed to *software* patents when in reality aflicts the entire patent system.
I must be suicidal, defending patents in /. Already lost karma to some loser who called my previous reasoned posting flame-bait... Oh well is just stupid karma points....
In fact you are. Is called copyright and it lasts a lot longer than 17 years....
If on the other hand you spend a year in your garage developing some innovative thing that had not been done before, but happens to be in software, many /.'ers propose that you shouldn't be able to patent it.
Why the disparity? Here are some possible reasons:
From personal experience as an employeer, we tend to ask co-ops to perform as difficult a job as they are technically qualified to do.
Having said that, every employee, co-op or not, will find him/herself performing boring tasks very so often.
As Dilbert says, this is the reason why the pay you: if work was all fun they would charge admission prices at the entrance!
This is not to say that there aren't unenlightened co-op employers out there who think of co-op students as under-aged janitors, as opposed to trainees, but from what you write yours doesn't seem to be one.
The schools I'm familiar with keep tabs on co-op employers, and if they treat their students as sources of menial labor with no training component the employer gets an earful.
From personal experience as an employeer, we tend to ask co-ops to perform as difficult a job as they are technically qualified to do. Having said that, every employee, co-op or not, will find him/herself performing boring tasks very so often. As Dilbert says, this is the reason why the pay you: if work was all fun they would charge admission prices at the entrance! This is not to say that there aren't unenlightened co-op employers out there who think of co-op students as under-aged janitors, as opposed to trainees, but from what you write yours doesn't seem to be one. The schools I'm familiar with keep tabs on co-op employers, and if they treat their students as sources of menial labor with no training component the employer gets an earful.
The pictures from the website are amazing, but there is no way in which us slashdot geeks can take pride on that. Now, crashing aluminum foil on an asteroid is something any os could have worked on, thus its clear interest to this audience...
Everybody knows that ancient Egyptians were Caucasians, in fact I remember quite distinctively from my history lessons that Cleopatra had blue eyes and was married to Richard Burton.