I've been an AMD fan for several years because of the competitive pressure they put on Intel. I think it's one of the reasons we get the kind of bang for the buck in CPU's. So much so that CPU's are regarded as a mundane utility component of a computer, much as a power supply, a motherboard, or a copy of Windows XX.
Despite the nice price/performance ratio of the K7, it's got to be refreshed, because the fastest Pentium 4 chips can beat its top performance (after all the PR ratings and MHz are laid aside).
Conservative corporate IT buyers are quite willing to pay the relatively small extra price for P4 over a K7. Maybe a year and half ago when the K7 squeezed the PIII to the end of it's life and the K7 was the performance leader, AMD would have gotten some attention, but now it seems like the shoe is on the other foot.
Price cuts from Intel on the P4 and Celeron will keep the pressure on the K7 to where AMD really needs the Opteron.
So if the Chinese Red Army wants to release Red Flag SE-Linux 2.0 with Wine 2.0 without source code, then we'd be fine with that?
Personally, I like the "share and share alike" aspects of the GPL because of the higher force of propagation and distribution of intellectually valuable ideas that it promotes. I like when my tax dollars go into GPL code.
Since I'm not a real life web developer, I'm curious - is a reasonable course of action to push candidate fodder web pages through something like the W3 validator and have things work more of the time?
I'd really like to know, cause I'm a happy Mozilla user in a rising sea of IE 6 users. Hence, I'm advocating W3 standards rather than standards defined as "whatever the dominant application does".
But it would be some encouragement if compliance to the W3 standards (say HTML 4.01) would be sufficient to have web pages render properly in a maximal subset of browsers.
Like many other Slashdot readers I figure, "How is computing applied to humanites?"
Randomly, one specific example (rejected at least twice here) was the work of using compression algorithms to determine language and authorship of pieces of text.
However, I think one of the more fascinating aspects of the humanities is understanding more about the "human machine" and understanding how we get programmed to respond in certain ways.
Interesting that a group calling itself Software Choice is trying to tell people not to choose certain software.
Actually, it's quite typical in this day and age for a well-run advertising campaign to properly name their front organization.
Political advertisements do this all the time. You'll hear some mud-slinging ad based on emotion, innuendo and fabrications that discredit some candidate or other. At the end, you'll get the legally required announcement "Paid for by Citizens for Better Government" or "Paid for by Concerned People for Ethical Regulation". What a crok!
I wish I could remember a few specific examples. They're usually quite hilarious once you find out who is really behind the organization. Very often, the organization will choose a name that lays claim to a lot of the high-minded principles of their opposition. There's more confusion that way about who to support.
Unfortunately, though, there are probably a substantial number of people that can be hoodwinked by this subterfuge.
My wife told me straight out that she didn't want a diamond engagement ring.
She simply didn't care for such a diamond ring. She generally doesn't like jewelry, but occasionally likes a few simple pieces.
Knowing what I do about the artificial market for diamonds, I was more than happy to oblige her wishes. It was simply more money for the wedding and honeymoon, which are not cheap, either.
But do go in together to pick out wedding bands.
The experience of doing that, thinking about what kind of ring you want (Au, Ti, Ag, Pt, wide or narrow, plain or decorated, what to write on the inside - please not the LotR phrase, etc.) helps cement a relationship at least as well as getting a big rock. Or, at least, if the relationship was going to go anywhere, then the mutual ring picking experience will reinforce the same relationship dynamics.
This makes sense in a marketing move. Sun's biggest fear about linux is that folks can replace their Sparc Solaris boxen with x86 linux boxen. By marketing an x86 box (a sexy x86 box, I should add...) they can keep their hardware niche and use free software.
But it's a big shift for them to make.
Sun's accustomed to manufacturing their own in-house RISC hardware, with exacting quality specs for a small fixed-size market.
They'll need less in-house hardware expertise for a given volume of sales in the x86 world.
And they can't afford too much expensive expertise competing with Dell, HP, IBM where the margins are a lot lower than they were in the old model.
It's a big change for them. It's a scary new market where they could either surf or flounder. Despite their late entry into Linux, Sun has some of the strongest UNIX credentials in the industry. That could be used as a selling point for them to sell Linux to customers that would dearly like some reassurance that, for example, the NIS and NFS inventors are the ones that setup their Linux deployment.
I'm guessing that some of the big decision makers got burned by some bad decisions during the heyday of the.com boom.
You have to admit they have a point. They were sold on something which
they know absolutely nothing about (have an MBA, not an BS CS)
which turned out to be a dud.
Why should they believe the next hyped up set of buzzwords coming from the IT community? (Certainly they should be skeptical of the same vendors that sold them that previous pig in a poke, whoever they happen to be - hopefully not you!)
So first dispel any illusions that every new buzzword technology is a good thing.
Also, gain some credibility with those skeptics by validating their skepticism where it was well-founded. Yes, we sunk thousands of dollars into that supposed cure-all and it was nothing but headaches. It was mistake and you're right to call it a mistake. But also point out where things went right, or perhaps unexpectedly right (eg, Joe put in an open source proxy server that was the bees knees.)
If a vendor comes painting a picture, demand references to current users, and then dig down to the worker bee level in that organization to see if things really are working. Why dig? There's probably plenty of upper level folks in the showcase example company that want to look as if they made a good decision to go with vendor Y and technology Z. The CIO doesn't want to look bad to the other C level people, so definitely dig down. I can't tell you how much money has been poured down holes as a result of an uninformed decision coming down from on high, where there is too much insulation from everyday reality of things like hung servers.
You need to back things up with solid arguments showing non-IT folks how introducing some technology helps their business' bottom line.
A worthy competitor that has implemented technology X where you can show it has had a beneficial effect is one good argument. Another argument is a detailed analysis of a small low-budget prototype roll-out: eg, we created an XML based mechanism for tech-friendly Salesman Fred to access the manufacturing database so he could know how much leadtime to let a customer know he could expect. Etc.
In the big overall scheme of things I've heard an argument made, and I believe most of it, that the unexpected growth in productivity over the past 15 years or so has been largely due to the adoption of IT. (Some growth is due to improved business processes, but I would argue that many of those processes wouldn't be possible 20 years ago given the technology of that day.) If you believe that, then stopping all further investment in IT will likely lead to a stagnation in productivity growth and profitability.
Nothing's ever that simple, of course, and there's no iron-clad arguments for adopting new technologies. There's risk, no two ways about it. But taking the risk earlier than others leads to more substantial rewards, if you can afford the investment.
I've always been intrigued by the Twiddler, but have been a little leery of the potential for carpal tunnel syndrome from using it (not that thousands of people have already gotten carpal tunnel syndrome from using two-handed QWERTY boards, but I haven't, yet).
This device looks like it might conform even better to one's hand than the flatter looking Twiddler.
What strikes me, though, is that the ultimate in one-handed keyboard comfort would be to customize the shape according to individual hands.
Once you get the basic electronics down to size, just have people go to sleep with some kind of goo in their hands that will harden into a shape that is natural for your relaxed hand.
Haven't joysticks come up with sufficient touch pads for emulating a full QWERTY board yet?
If "B" and "b" were meant to be the same, they'd look the same.
In many languages, English and German to take two examples, capital letters have had a specific usage for a much longer time than computers, or eve for a much longer time than typewriter keyboards. They are different letters and, more importantly, people believe they're different letters.
Why regress to that era 20 years ago when in some strange computer-based subculture of reality the two letters were indistinguishable?
Read your RFCs and then re-read them with a friend or two to make sure you read them right the first time.
I'd say another thing is to give some glory to people that write regression tests for RFC compliance for various applications.
Even all the stupid sounding things that people think "never" happen in real life. Those things that happen only one out of 1e7 times are the first things that the cracking crowd applies their crowbars to.
Microsoft, especially, could do with some of that kind of testing given their huge R&D budgets. It might help diminish the public black eyes they keep getting with respect to standards compliance and security vulnerabilities. Getting the mindset of being compliant to a standard rather than "we are the standard" might help them to write more watertight APIs.
But Microsoft have finaly caught up, and their installation is pretty bloody simple in 2000 and XP.
Great.
Now that it is mandated that the OS be installed at the factory and fewer consumers actually get the installation CD it's getting easier to install. That makes sense.
It would seem that making installation of MS products more difficult would be a better tactic against "piracy" than making installation of 2K/XP easier.
That way, anyone buying a sham copy of the OS in Thailand would have no use for it unless they were willing to spend serious time, money and intellect on the installation process.
Reminds me of a law enforcement tactic that they bring out every few years.
Publish in the newspaper a list of names that include some people wanted on outstanding warrants that can't be tracked down, have no fixed address, etc.
But the advertisement is not "Wanted" rather, it's "Winners of Prizes in our Store Competition! Show up Saturday at 10 am to claim your prize!"
People actually deliver themselves to the right address, where they are cuffed.
I've known FORTRAN since...well...for longer than many slashdot readers have been alive. Done big, numerically intensive projects in it for many years in the 1980s.
I've also done C (1990s) and C++ (2000s).
Ten years ago the criticism about speed was true - that compiled FORTRAN would beat C++ to pieces. Not anymore, unless you're committing newbie mistakes in C++.
If I were you I'd make the best of all worlds. I'd use Python for upper level logic in a clean syntax, in a quick scripting environment.
Then, if there's numerically intensive loops in the lower reaches of your code that get executed billions of times, go ahead and use FORTRAN. Especially when you're doing something like computing eigenvalues for specially shaped matrices, etc., where chances are someone has already written a FORTRAN subroutien to do it well.
Finally, use something like SIP or SWIG to connect the upper level Python to the lower level FORTRAN. Look, too, at Numerical Python and SciPy for others that have been down this road.
There's a lot to be gained from all those netlib routines that have withstood the test of time and been optimized to such an extent that even the FORTRAN programmers surrender, and drop down to call BLAS routines that are best written in assembler.
Using FORTRAN for the whole project is asking for pain once you start looking at things like parsing character input, connecting to network, linking into system libraries written in C. Been there, done that, have the scars.
The main reason is because the user want Windows. The team here would very much like to go Linux, but the users are the real hold-up. Honestly, $300000 / year could do a lot here.
You might be surprised at what your users "want".
Sure, many have invested their precious time in climbing the learning curves of Word, Excel, Outlook, etc. and can't be bothered to learn alternative open source applications.
But others, usually the more technically adventurous types (like yourself) are willing to try out something new, to invest the time just in case there happens to be a reward for the risk.
Most sites just use Linux as an under-the-radar server OS that is cheap and reliable.
But do take the next step of building up a nice desktop version for your site. With a little tuning, the new Linux desktops can be made into something productive for your users.
After a while, others will notice the new boxes and Linux growth will sell itself as people begin to ask questions that never get asked in the monoculture environment where there are no alternatives.
wether or not MS abused it's monopoly by also distributing a tcp/ip stack, a web browser, a media player,
No one disputes that a TCP/IP stack, a web browser or a media player is not a useful thing to have on a computer.
If "utility" to the consumer were the only criterion upon which to evaluate improvements to Windows, then there would be no problem in bundling other "useful" items along with Windows, such as MSN,....(a few years go by)..., and MS House [was a Ted Rall cartoon about this a few years back], which "works best with Windows".
You might rather have a choice at who sells you your house, but given a choice between a "house with computer" or a "computerless house", you'd probably cave-in and buy the Service Agreement from Microsoft that gives upgrades to the latest house features. (There's no other way to buy MS House except on their terms.) There is absolutely no question that the house represents a added value to the Windows experience; consumers want them and Microsoft is providing what the consumer finds useful.
Microsoft has been found guilty of using it's monopoly position in one market to compete unfairly in another. That's the issue - not whether or not the add-on gizmo's have provided some non-zero utility to the consumer.
Slashdot Mirror
Why the push for MHz and not multiprocessor systems?
An excellent question, and one with an answer, too!
Most systems already have multiple processors!
Don't believe me? Hint: look on the GPU, or even on the high end NIC or RAID controllers.
I've been an AMD fan for several years because of the competitive pressure they put on Intel. I think it's one of the reasons we get the kind of bang for the buck in CPU's. So much so that CPU's are regarded as a mundane utility component of a computer, much as a power supply, a motherboard, or a copy of Windows XX.
Despite the nice price/performance ratio of the K7, it's got to be refreshed, because the fastest Pentium 4 chips can beat its top performance (after all the PR ratings and MHz are laid aside).
Conservative corporate IT buyers are quite willing to pay the relatively small extra price for P4 over a K7. Maybe a year and half ago when the K7 squeezed the PIII to the end of it's life and the K7 was the performance leader, AMD would have gotten some attention, but now it seems like the shoe is on the other foot.
Price cuts from Intel on the P4 and Celeron will keep the pressure on the K7 to where AMD really needs the Opteron.Lest this happen to you.
So if the Chinese Red Army wants to release Red Flag SE-Linux 2.0 with Wine 2.0 without source code, then we'd be fine with that?
Personally, I like the "share and share alike" aspects of the GPL because of the higher force of propagation and distribution of intellectually valuable ideas that it promotes. I like when my tax dollars go into GPL code.
There's still a net loss, but the real phenomenon appears to be a shifting of phytoplankton from north to south.
You mean, shifting in a direction opposite to where most developed countries are located [except for Oz]?
Since I'm not a real life web developer, I'm curious - is a reasonable course of action to push candidate fodder web pages through something like the W3 validator and have things work more of the time?
I'd really like to know, cause I'm a happy Mozilla user in a rising sea of IE 6 users. Hence, I'm advocating W3 standards rather than standards defined as "whatever the dominant application does".
But it would be some encouragement if compliance to the W3 standards (say HTML 4.01) would be sufficient to have web pages render properly in a maximal subset of browsers.
Like many other Slashdot readers I figure, "How is computing applied to humanites?"
Randomly, one specific example (rejected at least twice here) was the work of using compression algorithms to determine language and authorship of pieces of text.
However, I think one of the more fascinating aspects of the humanities is understanding more about the "human machine" and understanding how we get programmed to respond in certain ways.
Interesting that a group calling itself Software Choice is trying to tell people not to choose certain software.
Actually, it's quite typical in this day and age for a well-run advertising campaign to properly name their front organization.
Political advertisements do this all the time. You'll hear some mud-slinging ad based on emotion, innuendo and fabrications that discredit some candidate or other. At the end, you'll get the legally required announcement "Paid for by Citizens for Better Government" or "Paid for by Concerned People for Ethical Regulation". What a crok!
I wish I could remember a few specific examples. They're usually quite hilarious once you find out who is really behind the organization. Very often, the organization will choose a name that lays claim to a lot of the high-minded principles of their opposition. There's more confusion that way about who to support.
Unfortunately, though, there are probably a substantial number of people that can be hoodwinked by this subterfuge.
My wife told me straight out that she didn't want a diamond engagement ring.
She simply didn't care for such a diamond ring. She generally doesn't like jewelry, but occasionally likes a few simple pieces.
Knowing what I do about the artificial market for diamonds, I was more than happy to oblige her wishes. It was simply more money for the wedding and honeymoon, which are not cheap, either.
But do go in together to pick out wedding bands.
The experience of doing that, thinking about what kind of ring you want (Au, Ti, Ag, Pt, wide or narrow, plain or decorated, what to write on the inside - please not the LotR phrase, etc.) helps cement a relationship at least as well as getting a big rock. Or, at least, if the relationship was going to go anywhere, then the mutual ring picking experience will reinforce the same relationship dynamics.
This makes sense in a marketing move. Sun's biggest fear about linux is that folks can replace their Sparc Solaris boxen with x86 linux boxen. By marketing an x86 box (a sexy x86 box, I should add...) they can keep their hardware niche and use free software.
But it's a big shift for them to make.
Sun's accustomed to manufacturing their own in-house RISC hardware, with exacting quality specs for a small fixed-size market.
They'll need less in-house hardware expertise for a given volume of sales in the x86 world.
And they can't afford too much expensive expertise competing with Dell, HP, IBM where the margins are a lot lower than they were in the old model.
It's a big change for them. It's a scary new market where they could either surf or flounder. Despite their late entry into Linux, Sun has some of the strongest UNIX credentials in the industry. That could be used as a selling point for them to sell Linux to customers that would dearly like some reassurance that, for example, the NIS and NFS inventors are the ones that setup their Linux deployment.
Only at some companies.
I'm guessing that some of the big decision makers got burned by some bad decisions during the heyday of the .com boom.
You have to admit they have a point. They were sold on something which
- they know absolutely nothing about (have an MBA, not an BS CS)
- which turned out to be a dud.
Why should they believe the next hyped up set of buzzwords coming from the IT community? (Certainly they should be skeptical of the same vendors that sold them that previous pig in a poke, whoever they happen to be - hopefully not you!)So first dispel any illusions that every new buzzword technology is a good thing.
Also, gain some credibility with those skeptics by validating their skepticism where it was well-founded. Yes, we sunk thousands of dollars into that supposed cure-all and it was nothing but headaches. It was mistake and you're right to call it a mistake. But also point out where things went right, or perhaps unexpectedly right (eg, Joe put in an open source proxy server that was the bees knees.)
If a vendor comes painting a picture, demand references to current users, and then dig down to the worker bee level in that organization to see if things really are working. Why dig? There's probably plenty of upper level folks in the showcase example company that want to look as if they made a good decision to go with vendor Y and technology Z. The CIO doesn't want to look bad to the other C level people, so definitely dig down. I can't tell you how much money has been poured down holes as a result of an uninformed decision coming down from on high, where there is too much insulation from everyday reality of things like hung servers.
You need to back things up with solid arguments showing non-IT folks how introducing some technology helps their business' bottom line.
A worthy competitor that has implemented technology X where you can show it has had a beneficial effect is one good argument. Another argument is a detailed analysis of a small low-budget prototype roll-out: eg, we created an XML based mechanism for tech-friendly Salesman Fred to access the manufacturing database so he could know how much leadtime to let a customer know he could expect. Etc.
In the big overall scheme of things I've heard an argument made, and I believe most of it, that the unexpected growth in productivity over the past 15 years or so has been largely due to the adoption of IT. (Some growth is due to improved business processes, but I would argue that many of those processes wouldn't be possible 20 years ago given the technology of that day.) If you believe that, then stopping all further investment in IT will likely lead to a stagnation in productivity growth and profitability.
Nothing's ever that simple, of course, and there's no iron-clad arguments for adopting new technologies. There's risk, no two ways about it. But taking the risk earlier than others leads to more substantial rewards, if you can afford the investment.
I've always been intrigued by the Twiddler, but have been a little leery of the potential for carpal tunnel syndrome from using it (not that thousands of people have already gotten carpal tunnel syndrome from using two-handed QWERTY boards, but I haven't, yet).
This device looks like it might conform even better to one's hand than the flatter looking Twiddler.
What strikes me, though, is that the ultimate in one-handed keyboard comfort would be to customize the shape according to individual hands.
Once you get the basic electronics down to size, just have people go to sleep with some kind of goo in their hands that will harden into a shape that is natural for your relaxed hand.
Haven't joysticks come up with sufficient touch pads for emulating a full QWERTY board yet?
of an earlier announcement of a vulnerability here found by some folks at Bell Labs.
So is this new (albeit social engineering) vulnerability just "asking the million questions" in one shot?
If "B" and "b" were meant to be the same, they'd look the same.
In many languages, English and German to take two examples, capital letters have had a specific usage for a much longer time than computers, or eve for a much longer time than typewriter keyboards. They are different letters and, more importantly, people believe they're different letters.
Why regress to that era 20 years ago when in some strange computer-based subculture of reality the two letters were indistinguishable?
Read your RFCs and then re-read them with a friend or two to make sure you read them right the first time.
I'd say another thing is to give some glory to people that write regression tests for RFC compliance for various applications.
Even all the stupid sounding things that people think "never" happen in real life. Those things that happen only one out of 1e7 times are the first things that the cracking crowd applies their crowbars to.
Microsoft, especially, could do with some of that kind of testing given their huge R&D budgets. It might help diminish the public black eyes they keep getting with respect to standards compliance and security vulnerabilities. Getting the mindset of being compliant to a standard rather than "we are the standard" might help them to write more watertight APIs.
But Microsoft have finaly caught up, and their installation is pretty bloody simple in 2000 and XP.
Great.
Now that it is mandated that the OS be installed at the factory and fewer consumers actually get the installation CD it's getting easier to install. That makes sense.
It would seem that making installation of MS products more difficult would be a better tactic against "piracy" than making installation of 2K/XP easier.
That way, anyone buying a sham copy of the OS in Thailand would have no use for it unless they were willing to spend serious time, money and intellect on the installation process.
Companies shouldn't give out loans to board members
I think that spells
just as much as having accounting firms provide dual services of auditing books and "consulting".See Molly Ivin's column for a full rant.
Reminds me of a law enforcement tactic that they bring out every few years.
Publish in the newspaper a list of names that include some people wanted on outstanding warrants that can't be tracked down, have no fixed address, etc.
But the advertisement is not "Wanted" rather, it's "Winners of Prizes in our Store Competition! Show up Saturday at 10 am to claim your prize!"
People actually deliver themselves to the right address, where they are cuffed.
Edsger Dijkstra (Interestingly enough, Dijkstra died today.)
Yeah, I saw that. Sad, losing a luminary like that.
And pointedly relevent to this discussion, since FORTRAN used to use GOTO statements for branching.
If you're considering FORTRAN, then beware the GOTO as Edsger pointed out in this classic.
I've known FORTRAN since...well...for longer than many slashdot readers have been alive. Done big, numerically intensive projects in it for many years in the 1980s.
I've also done C (1990s) and C++ (2000s).
Ten years ago the criticism about speed was true - that compiled FORTRAN would beat C++ to pieces. Not anymore, unless you're committing newbie mistakes in C++.
If I were you I'd make the best of all worlds. I'd use Python for upper level logic in a clean syntax, in a quick scripting environment.
Then, if there's numerically intensive loops in the lower reaches of your code that get executed billions of times, go ahead and use FORTRAN. Especially when you're doing something like computing eigenvalues for specially shaped matrices, etc., where chances are someone has already written a FORTRAN subroutien to do it well.
Finally, use something like SIP or SWIG to connect the upper level Python to the lower level FORTRAN. Look, too, at Numerical Python and SciPy for others that have been down this road.
There's a lot to be gained from all those netlib routines that have withstood the test of time and been optimized to such an extent that even the FORTRAN programmers surrender, and drop down to call BLAS routines that are best written in assembler.
Using FORTRAN for the whole project is asking for pain once you start looking at things like parsing character input, connecting to network, linking into system libraries written in C. Been there, done that, have the scars.
Amend the UCITA so that all software sold is required either:
- to provide a warranty, or
- to provide full open access to the source code so the user may modify it as they see fit.
completely at the pleasure of the software author or vendor.is it possible MS is changing?
Guaranteed they are changing.
"You want interoperable protocols?"
"We got interoperable protocols. Here, they're yours now, too! Use `em all you want under the terms of 12 page anti-GPL license!"
"BTW, if you liked those protocols, we'll get busy making more! (In fact, they're already here, getting perfected until just the right time!"
The main reason is because the user want Windows. The team here would very much like to go Linux, but the users are the real hold-up. Honestly, $300000 / year could do a lot here.
You might be surprised at what your users "want".
Sure, many have invested their precious time in climbing the learning curves of Word, Excel, Outlook, etc. and can't be bothered to learn alternative open source applications.
But others, usually the more technically adventurous types (like yourself) are willing to try out something new, to invest the time just in case there happens to be a reward for the risk.
Most sites just use Linux as an under-the-radar server OS that is cheap and reliable.
But do take the next step of building up a nice desktop version for your site. With a little tuning, the new Linux desktops can be made into something productive for your users.
After a while, others will notice the new boxes and Linux growth will sell itself as people begin to ask questions that never get asked in the monoculture environment where there are no alternatives.
If you look in the average inbox, you'll find a fair quantity of Unsolicited Bulk Email (aka "spam").
wether or not MS abused it's monopoly by also distributing a tcp/ip stack, a web browser, a media player,
No one disputes that a TCP/IP stack, a web browser or a media player is not a useful thing to have on a computer.
If "utility" to the consumer were the only criterion upon which to evaluate improvements to Windows, then there would be no problem in bundling other "useful" items along with Windows, such as MSN, ....(a few years go by)..., and MS House [was a Ted Rall cartoon about this a few years back], which "works best with Windows".
You might rather have a choice at who sells you your house, but given a choice between a "house with computer" or a "computerless house", you'd probably cave-in and buy the Service Agreement from Microsoft that gives upgrades to the latest house features. (There's no other way to buy MS House except on their terms.) There is absolutely no question that the house represents a added value to the Windows experience; consumers want them and Microsoft is providing what the consumer finds useful.
Microsoft has been found guilty of using it's monopoly position in one market to compete unfairly in another. That's the issue - not whether or not the add-on gizmo's have provided some non-zero utility to the consumer.