Then you'll still use MS word, like 99.9% of the scientific and engineering world.
I'd estimate usage at more like "1.5 nines" rather than 3 nines.
I wrote my thesis using LaTeX. After it's handling of equations, there's no migrating to something inferior.
There is simply no joy in using Word.
I use Word under duress, try to avoid learning anything more about it than just what I need to do my job. Also, I will deliberately avoid doing equations, cross-referencing, etc. in Word.
As far as I'm concerned, if I must use a bean-counter communications tool for business, then I may as well use it to express only things that bean counters will appeciate.
A real opt-out list would be maintained by a 3rd party with contractual and legal penalties for distributing your personal info.
I've often maintained the most powerful tool would be for ordinary citizens to claim copyrights on correlated collections of personal information. Then, individuals would be permitted to pursue all the penalties that the *AA organizations do for unauthorized distribution of Britney Spears songs and Owen Wilson movies.
But as you implement it new versions will come out and you'll be in constant catch-up. On top of that, DirectX is used for games so you need to have it perform well.
We focus on just how difficult it is for FOSS programmers and 3rd party commercial would-be competitors of Microsoft to try to match bleeding edge performance and retain backward compatibility at the same time.
The interesting issue is that Microsoft itself is constantly fighting a very similar battle - trying to bring out bleeding edge performance and retain some measure of backwards compatibility.
At some point, though, the sheer complexity of the API accretion becomes expensive to extend and to maintain. Unless a Must Have new application comes out to justify why customers should buy into the new bleeding edge new features, I can see where just plain good backward compatibility can become more important. It's good enough. The same principle of being available and of being Good Enough might allow competition to MS to catch up simply because no single entity can make the whole bundle progress at a reasonable pace for a reasonable cost as the complexity of the overall beast gains weight.
Cleanly-divided modules are technically superior. It's not that MS is technically incapable of doing it, either. If we had the HTML engine, DirectX, etc. built the proper way on top of something like Windows XP embedded the world as a whole would be progressing more easily. It's just that monopoly preservation has genuine shareholder value - if I were an MS shareholder or mananger I, too, would not mind if complicated levers and pulleys prevented my competitors from rapidly developing new markets and would allow me a turbo-boost into those markets whenever I wished. And so it is.
Put a cement block on the floor in front of you. Now stand on it.
Ta da! Instant local gravity increase, because there is now more mass underneath you.
Really?
What I mean is this: does the extra attractive force between you and the cement block more than compensate for the extra distance you've put between yourself and the earth's center of mass (which would tend to decrease the gravitational force you feel from earth)?
that's a lot of hard cash that will be freed up on the consumer side.
It's already been spent to fill up the tank. Enterprising laid-off AOL employees have already converted themselves into barrels of oil^W^W^Wprivate security contractors in Iraq to take advantage of this economic opportunity.
I've recently been using Graphviz with gprof and gprof2dot.awk to map out some really crufty old code. It's really illuminating.
However, what I'd really like to be able to do is have it animated, showing nodes and edges in the graph as they become activated (slowed down from real time, of course and reducing super-repetitive actions logarithmically).
A web application for this kind of graph layout and animation using SVG would be widely applicable, not just to visualizing call trees for computer programs, but also for understanding other phenomena that are visualized well with graphs, such as I/O flow between applications, industrial process equipment, organizational hierarchies, network traffic with hosts appearing and disappearing, etc.
Unfortunately, everything I've seen so far is either cobbled together for a one-off spectacular presentation, specific to some particular application domain, or else locked into a much less general display mechanism (platform X graphics API).
If anyone knows anything approximately close to dynamic web-based graph visualization framework I'd like to know about it.
I'm not privileged to be part of the targeted demographic, so my opinion that it seems unlikely to catch on might be wrong. I experienced a moment of cognitive dissonance when I read
...consider how Wal-Mart helps support their personal style and self-expression through the depth and breadth of products Wal-Mart offers.
What I really wonder is what the real IT folks at Walmart think of this scheme?
to be wrong in 50% of the cases when predicting weather a week ahead?
Faster hardware will only make a slight dent in this problem.
Since you mention continuum physics, you're probably aware of the nonlinearities inherent to such simulations. Given Lorenz exponents that cause infinitesimal differences in initial conditions to diverge exponentially over time, we'll never achieve certainty in weather forecasting. The best we can hope for is to improve our odds by simultaneously increasing computing power to enable us to run finer grids, increase the number of monitoring stations, increase the accuracy and precision of those monitoring stations to improve the 50% probability to say 85% probability of a successful forecast.
This being Slashdot and all, I think it's easy to understand where decisions to move from a proprietary solution to an open source implementation can generate a great deal of unwanted heat in the political/religious/monetary venue.
I thought I read somewhere a while back that neural development was affected by environmental factors. That being the case, it's more like decisions are effectively a function of integrated experience.
If you let someone choose between eating a pizza and and a rotten rat
The fundamental problem is that a rotten rat with a sufficiently large advertising budget and good packaging can still be chosen by more people than a pizza.
The free enterprise system is powerful and efficient, but nothing about it guarantees that we will always receive complete and accurate information.
Pardon my cynicism, but I've been eating a steady diet of rotten rats for decades.
Special limited exclusive commercial rights for ideas has been around a few centuries.
By now, we ought to have a good idea of how well it works and under what circumstances it works best.
Compared to the 18th century when patent terms and copyright lifetimes were dreamt up, the pace of innovation today with so many innovators who are able to communicate so quickly, it seems to me that the duration of special patent and copyright protection ought to be much shorter. After a few years, "IP" should drop into the public domain.
The new FCC commissioner should be proud of speaking out on this issue in a way that is likely to garner more world consensus.
The Peoples Republic of China has long had a difficult time with criminals disclosing state secrets. If more of China's state secrets were protected by strong DRM then such damaging leaks could be prevented.
There still isn't a long term way of dealing with the tons of radioactive waste being produced.
I saw a talk recently where this issue was addressed. I'm presuming the safety aspects are conquerable, especially after people can see the implications of Chernobyl.
The biggest problem with nuclear waste is two fold: the quantity produced and the mix of isotopes, including some very long-lived products that persist over thousands of years that need to be kept away from the biosphere until the radioactive decay is sufficient to render them harmless. A lot of stuff for a looong time. Very problematic.
However, it is possible to further "burn" those long half-life isotopes in another reactor (transmute them to shorter half-life isotopes). Along with a separation of the different waste products so that you only have to store the 1 percent or so of the original waste that cannot be further reduced in a repository, those products only have several hundred year half-lives so they wouldn't tie up the planet with poisonous places. A better situation. Not perfect. And, the separation probably still costs too much relative to using alternative energy sources. But it's a step forward, it seems.
Of course, on the political front you need different ambassadors to nuclear power than the current US administration which has a severe credibility and competency deficit. Like Nixon going to China, you need a certifiable pro-environment administration pushing forward on nuclear energy development ( and even then you'll still never convince some fraction of the population for whom even the word nuclear is a bogeyman).
The same concerned, nervous parents that surveil their nervous justifiably paranoid children 24/7 with the latest technology of 2006 will be the same nervous, justifiably paranoid parents in nursing homes being surveiled by their concerned, nervous children 24/7 with the latest technological offerings of 2036.
New features, standards compliance, standards non-compliance, bugs fixed, new bugs exposed in Internet Explorer 7 will have a minimal impact on its rate of market penetration relative to Mozilla Firefox.
More than anything else, the rate of adoption of new users of Internet Explorer 7 is closely correlated to the rate at which consumers buy new PCs with Microsoft operating systems and Internet Explorer pre-installed on them.
That's the signal. Everything else is just noise by comparison.
Business Strength Becomes Technical Weakness
on
Why Windows is Slow
·
· Score: 1
Actually, the feature-laden complexity of Windows and Office has benefitted Microsoft over the years. This is something the Times article doesn't really mention, but which is an important point.
If Microsoft had engineered Windows and its other applications according to widely-accepted technical design principles like modularity, interchangeability, complete and open specification (would have ironed out device driver issues better) it would have made maintenance much easier now. But doing that would also have opened them up to more competition along the way and the possibility of people migrating and branching to competing products and losing market share. Bundling everything together and "leveraging Windows" (BG's mantra) was a successful business strategy for keeping competitors at bay and discouraging users considering a switch to something else.
My sympathies go out to the worker bees in Redmond who have lost out on the big stock appreciation and arrived just in time to re-tie an even larger Gordian Knot. All work and no pay.
Slashdot Mirror
I'd estimate usage at more like "1.5 nines" rather than 3 nines.
I wrote my thesis using LaTeX. After it's handling of equations, there's no migrating to something inferior.
There is simply no joy in using Word.
I use Word under duress, try to avoid learning anything more about it than just what I need to do my job. Also, I will deliberately avoid doing equations, cross-referencing, etc. in Word.
As far as I'm concerned, if I must use a bean-counter communications tool for business, then I may as well use it to express only things that bean counters will appeciate.
I've often maintained the most powerful tool would be for ordinary citizens to claim copyrights on correlated collections of personal information. Then, individuals would be permitted to pursue all the penalties that the *AA organizations do for unauthorized distribution of Britney Spears songs and Owen Wilson movies.
We focus on just how difficult it is for FOSS programmers and 3rd party commercial would-be competitors of Microsoft to try to match bleeding edge performance and retain backward compatibility at the same time.
The interesting issue is that Microsoft itself is constantly fighting a very similar battle - trying to bring out bleeding edge performance and retain some measure of backwards compatibility.
At some point, though, the sheer complexity of the API accretion becomes expensive to extend and to maintain. Unless a Must Have new application comes out to justify why customers should buy into the new bleeding edge new features, I can see where just plain good backward compatibility can become more important. It's good enough. The same principle of being available and of being Good Enough might allow competition to MS to catch up simply because no single entity can make the whole bundle progress at a reasonable pace for a reasonable cost as the complexity of the overall beast gains weight.
Cleanly-divided modules are technically superior. It's not that MS is technically incapable of doing it, either. If we had the HTML engine, DirectX, etc. built the proper way on top of something like Windows XP embedded the world as a whole would be progressing more easily. It's just that monopoly preservation has genuine shareholder value - if I were an MS shareholder or mananger I, too, would not mind if complicated levers and pulleys prevented my competitors from rapidly developing new markets and would allow me a turbo-boost into those markets whenever I wished. And so it is.
Media PCs are meeting the same fate as DAT.
Really?
What I mean is this: does the extra attractive force between you and the cement block more than compensate for the extra distance you've put between yourself and the earth's center of mass (which would tend to decrease the gravitational force you feel from earth)?
It's already been spent to fill up the tank. Enterprising laid-off AOL employees have already converted themselves into barrels of oil^W^W^Wprivate security contractors in Iraq to take advantage of this economic opportunity.
I've recently been using Graphviz with gprof and gprof2dot.awk to map out some really crufty old code. It's really illuminating.
However, what I'd really like to be able to do is have it animated, showing nodes and edges in the graph as they become activated (slowed down from real time, of course and reducing super-repetitive actions logarithmically).
A web application for this kind of graph layout and animation using SVG would be widely applicable, not just to visualizing call trees for computer programs, but also for understanding other phenomena that are visualized well with graphs, such as I/O flow between applications, industrial process equipment, organizational hierarchies, network traffic with hosts appearing and disappearing, etc.
Unfortunately, everything I've seen so far is either cobbled together for a one-off spectacular presentation, specific to some particular application domain, or else locked into a much less general display mechanism (platform X graphics API).
If anyone knows anything approximately close to dynamic web-based graph visualization framework I'd like to know about it.
Yeah, my company's health care plan does this to me, too.
I think that must be why my doctor has to encrypt the results of my checkup exams.
I'm not privileged to be part of the targeted demographic, so my opinion that it seems unlikely to catch on might be wrong. I experienced a moment of cognitive dissonance when I read
What I really wonder is what the real IT folks at Walmart think of this scheme?
Recall that the company maintains one of the largest IT shops in the world that keeps timely up-to-date information on exactly what is selling where and how suppliers need to adjust (SCM). Non-trivial stuff that serves as casebook exercises in lessons for why Walmart is Changin the Landscape.
They've got to be trying hard not to LOL at the ludicrosity.
Faster hardware will only make a slight dent in this problem.
Since you mention continuum physics, you're probably aware of the nonlinearities inherent to such simulations. Given Lorenz exponents that cause infinitesimal differences in initial conditions to diverge exponentially over time, we'll never achieve certainty in weather forecasting. The best we can hope for is to improve our odds by simultaneously increasing computing power to enable us to run finer grids, increase the number of monitoring stations, increase the accuracy and precision of those monitoring stations to improve the 50% probability to say 85% probability of a successful forecast.
My favorite: the odometer/slot machine password cracking software, whirring the last few places as you hear the Bad Guy® coming down the hall...
You're making a common error.
What you term "civil war on Christmas" is really better known as the "Terrorist Surveillance Program".
What the MSM calls "illegal warrantless spying" is really "isolated insurgency by post-Baathist and foreign fighters and Natalee Holloway".
This being Slashdot and all, I think it's easy to understand where decisions to move from a proprietary solution to an open source implementation can generate a great deal of unwanted heat in the political/religious/monetary venue.
I thought I read somewhere a while back that neural development was affected by environmental factors. That being the case, it's more like decisions are effectively a function of integrated experience.
The fundamental problem is that a rotten rat with a sufficiently large advertising budget and good packaging can still be chosen by more people than a pizza.
The free enterprise system is powerful and efficient, but nothing about it guarantees that we will always receive complete and accurate information.
Pardon my cynicism, but I've been eating a steady diet of rotten rats for decades.
Special limited exclusive commercial rights for ideas has been around a few centuries.
By now, we ought to have a good idea of how well it works and under what circumstances it works best.
Compared to the 18th century when patent terms and copyright lifetimes were dreamt up, the pace of innovation today with so many innovators who are able to communicate so quickly, it seems to me that the duration of special patent and copyright protection ought to be much shorter. After a few years, "IP" should drop into the public domain.
There are plenty of ivy-league graduates who can't freely express their ideas. Q.E.D.
Microsoft will be happy to $upply $oftware that i$ interoperable
The new FCC commissioner should be proud of speaking out on this issue in a way that is likely to garner more world consensus.
The Peoples Republic of China has long had a difficult time with criminals disclosing state secrets. If more of China's state secrets were protected by strong DRM then such damaging leaks could be prevented.
I saw a talk recently where this issue was addressed. I'm presuming the safety aspects are conquerable, especially after people can see the implications of Chernobyl.
The biggest problem with nuclear waste is two fold: the quantity produced and the mix of isotopes, including some very long-lived products that persist over thousands of years that need to be kept away from the biosphere until the radioactive decay is sufficient to render them harmless. A lot of stuff for a looong time. Very problematic.
However, it is possible to further "burn" those long half-life isotopes in another reactor (transmute them to shorter half-life isotopes). Along with a separation of the different waste products so that you only have to store the 1 percent or so of the original waste that cannot be further reduced in a repository, those products only have several hundred year half-lives so they wouldn't tie up the planet with poisonous places. A better situation. Not perfect. And, the separation probably still costs too much relative to using alternative energy sources. But it's a step forward, it seems.
Of course, on the political front you need different ambassadors to nuclear power than the current US administration which has a severe credibility and competency deficit. Like Nixon going to China, you need a certifiable pro-environment administration pushing forward on nuclear energy development ( and even then you'll still never convince some fraction of the population for whom even the word nuclear is a bogeyman).
The same concerned, nervous parents that surveil their nervous justifiably paranoid children 24/7 with the latest technology of 2006 will be the same nervous, justifiably paranoid parents in nursing homes being surveiled by their concerned, nervous children 24/7 with the latest technological offerings of 2036.
If I were Bill Gates I'd enter My Name into my Google News customization search string, just as I do for My Corporation.
I'm not Bill Gates, nor as famous, so I wouldn't get any hits.
New features, standards compliance, standards non-compliance, bugs fixed, new bugs exposed in Internet Explorer 7 will have a minimal impact on its rate of market penetration relative to Mozilla Firefox.
More than anything else, the rate of adoption of new users of Internet Explorer 7 is closely correlated to the rate at which consumers buy new PCs with Microsoft operating systems and Internet Explorer pre-installed on them.
That's the signal. Everything else is just noise by comparison.
Actually, the feature-laden complexity of Windows and Office has benefitted Microsoft over the years. This is something the Times article doesn't really mention, but which is an important point.
If Microsoft had engineered Windows and its other applications according to widely-accepted technical design principles like modularity, interchangeability, complete and open specification (would have ironed out device driver issues better) it would have made maintenance much easier now. But doing that would also have opened them up to more competition along the way and the possibility of people migrating and branching to competing products and losing market share. Bundling everything together and "leveraging Windows" (BG's mantra) was a successful business strategy for keeping competitors at bay and discouraging users considering a switch to something else.
My sympathies go out to the worker bees in Redmond who have lost out on the big stock appreciation and arrived just in time to re-tie an even larger Gordian Knot. All work and no pay.