Lawful authority means lawful in terms of jurisdiction. USA != Australia, therefore if the law doesn't overlap (as it doesn't in this case), it's not lawful. No-one will read this post unfortunately, but the courts here in Oz do not always follow governmental preference. Witness Mabo for an example. If they don't like the law, think it's unjust, or think it's encroaching on their territory, they get pretty pissed.
True... the only BSODs I've had on win2k were very strange... not repeatable, and I was doing a lot at once (running many applications, not all stable). Plus, the KX-133 had AGP problems under Win2k originally, which is where I think they might have come from. It's since been fixed (in SP2). I figure something got probed that didn't like it...
Scary shit... you just described my system exactly. The only different is I've got a 600mhz Athlon, a 400MX, and I'm running Win2K. Otherwise, exactly the same. I've had probably under 5 BSODs in the last 3 three years. The only crashes and reinstalls I've had are because of dodgy programs overwriting stuff like winsock.dll and god knows what else (I got hit with some spyware).
A second after I hit post, I knew someone was going to point this out. Point taken, and you're right. I really meant only specifically in my case, but stupidly didn't point that out.
BTW, thanks for your suggestions about ECMAscript etc from ages ago. Didn't end up using it, but it was helpful in providing a different viewpoint.
Small point - did you know that running a DVD at 1600x1200 won't show any quality increase, as the video is only encoded at 720x480 in NTSC (720x576 in PAL)? It's like zooming in on a picture - you can try to mitigate the artifacts, but you're not actually getting any improvement in quality. I understand what you mean though (comparing the TV to the computer based on your setup). Anything over aprox 1/2 of the screen at 1600x1200 won't show any increase in quality (and will likely make things blurry in comparison due to the rescaling). If you wanted the best possible quality, drop the resolution to 720x480 (or 800x600 if your card can't do it). Try it and see if there's a difference - things should actually look marginally sharper.
This is true, but if I remember correctly a lot of machines from the first generation were really flaky (skipping frames, not playing some movies, RCE problems, etc). I'm not specifically bagging the chinese machines - the original Pioneers had horrendous audio sync problems.
Besides, I was only responding to when you said the "big players" don't do MP3s, SVCDs, etc. The latest lot of models do. You're right that they don't do the hidden menus though, but I believe most of the newer chinese brands don't do that anymore either. That it's cheaper is a different matter.
The big boys do support MP3s - my Pioneer 533K plays MP3s (same as the slim-line US Pioneer model - can't remember the model number at the moment). I believe there's a Denon that does MP3s as well. The Pioneer also reads recordable DVDs (multiple formats), SVCDs, CDRs, and VCDs. And, it's modded to be region free (but not macrovision free - couldn't be bothered, since what's the point of VHS if you've got DVD quality). The only thing it doesn't play is DivX. Not bad!
Out of interest, do you work longer hours now? Is your quality of living the same (go out as much, save as much, have an equivalent sized place for the same cost)? Do you have as much time off? Also, where are you now?
People do leave their jobs here - in many positions, it's hard to get a reasonable raise unless you change roles or companies / departments. However, there are plenty of people who have been with the same company for 15+ years.
The 3 months off you're talking about would be long service leave, and I believe it's a statutory requirement. I think it might be longer than 3 months though, and I'm not sure it takes 20 years to get it (I think it might be less). I don't know though, as I've changed jobs a few times b/c of changing focus and interest in a pay-rise, so long-service leave doesn't really figure in my decision making process.
4 weeks per year is the legal requirement, plus public holidays (around 10 a year), plus sick days. Figure we only really work 44 weeks a year normally (8 weeks off between public holidays, sick days, and vacation time). You may get paid more in the US, but you *sure* work for it in comparison. In many ways I think the US is quite backwards - Oz and Europe have similar policies towards work (except I think Europe in general is even better for the employee than Oz).
Here's a novel idea - don't know if it's been tried before though. Nationalise the network and split the service delivery arm into a private company. Then, the network (which is the public good) can be run by the government, alongside all social contracts, and anyone can compete in offering telecom services at an equivalent base cost (eliminating the metro / regional cost differential, as the govt is paying for it). The key is recognising that the market (a large scale network) is a natural monopoly.
Of course you've got the whole impact that nationalising extremely large private resources would have on business confidence, but hey - that's sovreign risk, isn't it?;) Seriously though, I wonder if it would fly. Any ideas on the viability of it or if it's been tried elsewhere in other industries?
Re:The reason I click on "Read More..." on Katz st
on
Review: Orange County
·
· Score: 2, Insightful
That's really scary... I just realised I do the same!
You've got to wonder where the value is when the replies are more fun to read than the actual article...
But, if you knew what *type* of person kept buying your french loaf, along with what other purchases they made, you could price discriminate between the two groups and make even more profit.
For example, you discovered that the people who kept buying your loaf only bought a french loaf when they came to your store. The people who stopped buying your loaf normally also bought a savory bun (at $0.85). When you increased your price, they substituted a savory bun for the french loaf, buying a total of two savory buns instead of a french loaf and a savory bun.
Now that you've got this data, you set your french loaf price to $1.50 permanently. However, you also offer a deal where if you buy a savory bun for $0.85, you can also pick up a french loaf for $0.79. Your total profit is now higher than it had been previously.
Obviously, this is a simple example. However, if you've got a large consumer base and stock *many* products (as a supermarket does), you can churn the data through an analysis package and figure out a way to maximise your product. If you've got a large enough sample size (think many many stores), you can even test variations in pricing without disturbing your consumer base and scaring them off. Simply identify stores with common consumer characteristics (income, geneology, culture, whatever), and spread the price variations across all of them.
Tracking technologies automate this and make it *much* easier and efficient. Not that this is a bad thing - if you know how it works, you can get a good deal
Good post - excellent application of elasticity theory. My only quibble is that while it holds true in theory, it falls down in practice. Price elasticity is based on personal utility, which cannot be aggregated (in a mathematical sense - this has been proven, but is overlooked in most undergrad economics for simplicity). This has some fairly major impacts on micro and macro in general, but in this case, the assumption that a demand curve can be extrapolated is therefore flawed. You can only therefore predict what you have already observed. If you increase the price from $1.00 to $1.10 and your consumers continue to buy the same quantity, you're lucky. However, that is no guarantee that the same will happen if you increase the price to $1.11.
However, you've hit the nail on the head when you say tracking purchases helps you to know formally whether a price change has had an impact on purchases, and if so, on what type of consumer it has had an impact on. You can then extrapolate this to your broader population and forecast demand / profitability. The supply chain stuff potentially offers huge savings. Good data rocks:)
As I said in my other post above, this isn't correct. If you rely on this, you'll end up with results contrary to those that you expected (especially for proportionally large movements in price). Here's a link to my (brief) explanation. A Bachelor's degree does not an Economist make.
Incidentally, perfect information would allow a monopolistic firm to perfectly price discriminate, which would lead to price variances similar to the one the original poster described.
Tracking would assist this process. And remember, a monopoly is defined by the market. Ergo, if you don't have access to other, competing grocery stores, you are in a monopolistic situation (regardless of whether other groceries compete elsewhere nationally).
Econ 101 doesn't tell you the full story. There isn't a clear linear relationship between price and demand, regardless of what they told you in Econ 101. It's one of the biggest fallacies in modern undergrad economics. See Modigliani for a discussion. The simple linear relationship generally holds true for an individiual, but as utility cannot be compared between individuals, you technically cannot aggregate demand and expect the same simple curve. In reality, it's a higher-order non-linear relationship, leading to demand that may increase when prices go up and fall when prices go down, depending on the price being examined.
Bringing it back on topic, this means that it's not actually as easy as increasing or decreasing prices. You can increase the price and have more people subscribe, simply because you've moved into a different type of consumer who sees the good as now being a luxury (increasing their image, giving them utility). It sounds heretical, but it's true. If you're interested in a fuller explanation about why the foundations of Economics are incorrect, reply to this post.
If you're doing economics, check out Modigliani. If he and his supporters even get enough momentum, a lot of the tenets of micro and macroeconomics are likely to fall (as they're not mathematically sound). Neo-Keynesian economics and the Austrian school of economics are also worth looking at. They normally only teach you basic Keynesian economics and (maybe) monetarism in undergrad at Uni.
The difficulty is that a large part of the final product is the presentation. Presentation not in terms of flashy stuff, but in terms of conceptually mapping things out. This particular model details a process flow and allows users to test various scenarios by changing different elements of the process. Regarding the "links", I was pretty unclear. Basically, there's a dependence flow, where each cell on a spreadsheet is referenced to cells on previous sheets. It's this dependence flow that the model maps out.
I take your point about the spreadsheet being a poor tool for delivering a program, but the problem is it's a strategic model (one of the main value points being it allows users to quickly scan large amounts of numbers and see changes instantly). Without going into too much detail, a DB wouldn't quite fit as well as a spreadsheet for this particular solution. It would in most others I can think of, but not here.
Something you mentioned that I'm curious about is "plugin that has a DOM". What's an example of such a plugin? I'd actually love to move to a browser interface and dump Excel VBA - one of the problems I've identified for the future is one of technological lock-in - they're stuck using Excel for the life of the Model (probably about 5 - 10 years). A browser interface would solve the issue of platform independence while also allowing a much more flexible approach. Have you got any links about plugins under ECMAScript?
You probably didn't miss something stupid - I probably didn't do it the best way possible. I haven't used ECMAScript in a production sense, only played around with it as a minor part of web development. The sandbox issue I agree with - ECMAScript is much easier to sandbox.
The thing I'm not sure about (and here I'm clearly displaying my ignorance) is what capacity ECMAScript has to pull functionality from other applications. Using IE and VB as an example, I could pull Excel functionality and forms from Office 2000 into IE quite easily using ActiveX or another mechanism. But I'd still be using VB to do it. Can ECMAScript give me an Excel spreadsheet within IE? A lot of the modelling work I do needs to have that spreadsheet functionality. I've had a look through MSDN, but can't find anything beyond the same JScript stuff I used before (dates, simple operators, etc).
I can see that it'd be easier with a DB - build the hooks and display selected structured data in IE. No need for a spreadsheet. I can see forms wouldn't be an issue either - just webify. The simulation stuff would be easy to port, it's the links to the spreadsheet that would be difficult (this example is spread out over 30 sheets with somewhere around 2mil calculations).
I may be too late in this discussion, but the value of VBA is that it makes it easy to add additional functionality to documents. I use VB in Excel regularly. Think developing a Monte-Carlo simulation module for a forecasting model. Or a dynamic projection module that automatically selects the best projection method (ARIMA, regression, moving average, etc) based on the characteristics of the data. Or an interface to make handling multiple documents and scenarios easy through meta-data (kludgy, but it works as long as you stay within the rules).
This functionality can be embedded in the spreadsheet or DB for additional value. It would be ideal if VB were a properly designed language, but it's not. It may also have huge security holes, but that's not going to stop me using it. I know it's got problems, but I'll continue to use it. There's some serious business value in VBA.
This value may lock me into using MS products, but what other choices do I have given that all my clients run Office? You're dealing with extensive network externalities created through a defacto standard. Until KOffice or StarOffice can parse or convert VBA, they're pushing uphill.
Damn I wish I could mod this up. Computer science is black and white - things work, or they don't. Social science is grey - it depends on how you approach it. To assess the validity of the results, all you can do is evaluate the methodology chosen while considering the constraints it was done under. It's easy to say "it would have been better to portscan every computer on the Internet three times a day over a period three months", but even ignoring the cost factors in such an approach, servers running Linux may not even be connected to the Internet. His methodology sounds ok to me. There may have been other better ways to do it, but it doesn't look like there were any significant flaws in his approach.
People on/. have asked sampling questions - our normal explanations on how and why we structured our sampling go for 5+ pages of highly technical documentation (that relies on theory most people here probably wouldn't understand or be interested in, such as the benefits of 5 point likert scales vs 7 point, scaling and weighting difficulties, and techniques for accounting for measurement error). You can see why a complete answer to these questions won't be presented here. Our methodology section in our reports will often go for 15+ pages.
As an economist / statistician and part-time programmer / sys-admin, I like the comments about "all things being held equal". The reason we do that is to define the problem in such a way that we can answer it. Interacting with the real-world is a lot messier than coding a program. Stats and research is not easy. It's hard work coming up with something that's defensible and will give you quality outputs.
As has already been stated, Gartner asked end-users what they installed on their computers after they bought them. Not what was pre-installed on their computer. Implication: either Linux doesn't have the marketshare zealots want to belive, or accurately assessing server marketshare is difficult. You decide.
Garter is a research agency that has its value locked in its reputation. If it produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports. Short term gains (ie fabricating truths in reports for specific companies) will lead to long term losses (ie loss of reputation leading to loss of credibility and revenue). Are you arrogant enough to believe they haven't thought of this? The reality is, they analysed the information available to them and made the best, least biased forecast as they could.
Garter is not in the business of assessing obscure technical facts. They provide a strategic business perspective on technology. Tech-heads are not their market. They don't care about the operational aspects of the technology. The people who run business (which, most commonly, are not tech-heads and have different skills) are their market. They care about the strategic implications of the technology and longer-term market trends.
You think you could do a better job and produce higher quality research? Don't whinge - go out and do it. If you produce research of such a high calibre, you'll drive Gartner out of business in months. I'm constantly amazed at how many people on Slashdot are willing to spout off on things they know nothing about. I don't tell people how to fly a plane - I realise I don't know how. But, the village idiots on slashdot are always willing to provide legal advice, assume everyone in management is a PHB, that companies never ever know what they're doing, that everything is part of a conspiracy and that anyone who doesn't know how to write a sound compression script using bash is an idiot. On the other hand, some people here have some seriously big clues. You my friend, are not one.
In summary, read the bloody report and get some perspective before spouting off.
Heh... here I was trying to figure out what you were talking about, claiming that liberals were socialists. Then I remembered:
Being Liberal or having right political leanings is the same as conservatism here in Australia. Supporting Labor or having left political leanings here in Australia is the same as being an American liberal (which is broadly the same as socialism).
Don't mind me, I'll just be standing in the corner looking confused...
Anecdotal evidence, but I must be really lucky - both of my computers with integrated sound (purchased 1996 and 2000) work under linux. I don't think integrated sound cards are such a big deal now under linux. If you're using a *really* old motherboard with an obscure chipset, I guess you may have difficulties. I didn't though.
But your other point is right - the quality sucks. Stick with a real sound card, unless you just don't care at all. As in, you don't mind having disney sound source quality to play your MP3s.
Slashdot Mirror
I've seen them for around $1 at JB-HIFI. Bought 'em too ...
So you're telling my I'll have to spend an additional 50 UKP for no benefit (using my existing system)? Gee, that sounds good, where do I sign up?
Lawful authority means lawful in terms of jurisdiction. USA != Australia, therefore if the law doesn't overlap (as it doesn't in this case), it's not lawful. No-one will read this post unfortunately, but the courts here in Oz do not always follow governmental preference. Witness Mabo for an example. If they don't like the law, think it's unjust, or think it's encroaching on their territory, they get pretty pissed.
True ... the only BSODs I've had on win2k were very strange ... not repeatable, and I was doing a lot at once (running many applications, not all stable). Plus, the KX-133 had AGP problems under Win2k originally, which is where I think they might have come from. It's since been fixed (in SP2). I figure something got probed that didn't like it ...
Scary shit ... you just described my system exactly. The only different is I've got a 600mhz Athlon, a 400MX, and I'm running Win2K. Otherwise, exactly the same. I've had probably under 5 BSODs in the last 3 three years. The only crashes and reinstalls I've had are because of dodgy programs overwriting stuff like winsock.dll and god knows what else (I got hit with some spyware).
Not much to say really, but if anyone is interested, here's the link to the article about Spyro - it's a great read. Here it is.
A second after I hit post, I knew someone was going to point this out. Point taken, and you're right. I really meant only specifically in my case, but stupidly didn't point that out.
BTW, thanks for your suggestions about ECMAscript etc from ages ago. Didn't end up using it, but it was helpful in providing a different viewpoint.
Small point - did you know that running a DVD at 1600x1200 won't show any quality increase, as the video is only encoded at 720x480 in NTSC (720x576 in PAL)? It's like zooming in on a picture - you can try to mitigate the artifacts, but you're not actually getting any improvement in quality. I understand what you mean though (comparing the TV to the computer based on your setup). Anything over aprox 1/2 of the screen at 1600x1200 won't show any increase in quality (and will likely make things blurry in comparison due to the rescaling). If you wanted the best possible quality, drop the resolution to 720x480 (or 800x600 if your card can't do it). Try it and see if there's a difference - things should actually look marginally sharper.
This is true, but if I remember correctly a lot of machines from the first generation were really flaky (skipping frames, not playing some movies, RCE problems, etc). I'm not specifically bagging the chinese machines - the original Pioneers had horrendous audio sync problems.
Besides, I was only responding to when you said the "big players" don't do MP3s, SVCDs, etc. The latest lot of models do. You're right that they don't do the hidden menus though, but I believe most of the newer chinese brands don't do that anymore either. That it's cheaper is a different matter.
The big boys do support MP3s - my Pioneer 533K plays MP3s (same as the slim-line US Pioneer model - can't remember the model number at the moment). I believe there's a Denon that does MP3s as well. The Pioneer also reads recordable DVDs (multiple formats), SVCDs, CDRs, and VCDs. And, it's modded to be region free (but not macrovision free - couldn't be bothered, since what's the point of VHS if you've got DVD quality). The only thing it doesn't play is DivX. Not bad!
Out of interest, do you work longer hours now? Is your quality of living the same (go out as much, save as much, have an equivalent sized place for the same cost)? Do you have as much time off? Also, where are you now?
People do leave their jobs here - in many positions, it's hard to get a reasonable raise unless you change roles or companies / departments. However, there are plenty of people who have been with the same company for 15+ years.
The 3 months off you're talking about would be long service leave, and I believe it's a statutory requirement. I think it might be longer than 3 months though, and I'm not sure it takes 20 years to get it (I think it might be less). I don't know though, as I've changed jobs a few times b/c of changing focus and interest in a pay-rise, so long-service leave doesn't really figure in my decision making process.
4 weeks per year is the legal requirement, plus public holidays (around 10 a year), plus sick days. Figure we only really work 44 weeks a year normally (8 weeks off between public holidays, sick days, and vacation time). You may get paid more in the US, but you *sure* work for it in comparison. In many ways I think the US is quite backwards - Oz and Europe have similar policies towards work (except I think Europe in general is even better for the employee than Oz).
Here's a novel idea - don't know if it's been tried before though. Nationalise the network and split the service delivery arm into a private company. Then, the network (which is the public good) can be run by the government, alongside all social contracts, and anyone can compete in offering telecom services at an equivalent base cost (eliminating the metro / regional cost differential, as the govt is paying for it). The key is recognising that the market (a large scale network) is a natural monopoly.
Of course you've got the whole impact that nationalising extremely large private resources would have on business confidence, but hey - that's sovreign risk, isn't it? ;) Seriously though, I wonder if it would fly. Any ideas on the viability of it or if it's been tried elsewhere in other industries?
That's really scary ... I just realised I do the same!
...
You've got to wonder where the value is when the replies are more fun to read than the actual article
But, if you knew what *type* of person kept buying your french loaf, along with what other purchases they made, you could price discriminate between the two groups and make even more profit.
For example, you discovered that the people who kept buying your loaf only bought a french loaf when they came to your store. The people who stopped buying your loaf normally also bought a savory bun (at $0.85). When you increased your price, they substituted a savory bun for the french loaf, buying a total of two savory buns instead of a french loaf and a savory bun.
Now that you've got this data, you set your french loaf price to $1.50 permanently. However, you also offer a deal where if you buy a savory bun for $0.85, you can also pick up a french loaf for $0.79. Your total profit is now higher than it had been previously.
Obviously, this is a simple example. However, if you've got a large consumer base and stock *many* products (as a supermarket does), you can churn the data through an analysis package and figure out a way to maximise your product. If you've got a large enough sample size (think many many stores), you can even test variations in pricing without disturbing your consumer base and scaring them off. Simply identify stores with common consumer characteristics (income, geneology, culture, whatever), and spread the price variations across all of them.
Tracking technologies automate this and make it *much* easier and efficient. Not that this is a bad thing - if you know how it works, you can get a good deal
Good post - excellent application of elasticity theory. My only quibble is that while it holds true in theory, it falls down in practice. Price elasticity is based on personal utility, which cannot be aggregated (in a mathematical sense - this has been proven, but is overlooked in most undergrad economics for simplicity). This has some fairly major impacts on micro and macro in general, but in this case, the assumption that a demand curve can be extrapolated is therefore flawed. You can only therefore predict what you have already observed. If you increase the price from $1.00 to $1.10 and your consumers continue to buy the same quantity, you're lucky. However, that is no guarantee that the same will happen if you increase the price to $1.11.
However, you've hit the nail on the head when you say tracking purchases helps you to know formally whether a price change has had an impact on purchases, and if so, on what type of consumer it has had an impact on. You can then extrapolate this to your broader population and forecast demand / profitability. The supply chain stuff potentially offers huge savings. Good data rocks :)
As I said in my other post above, this isn't correct. If you rely on this, you'll end up with results contrary to those that you expected (especially for proportionally large movements in price). Here's a link to my (brief) explanation. A Bachelor's degree does not an Economist make.
Incidentally, perfect information would allow a monopolistic firm to perfectly price discriminate, which would lead to price variances similar to the one the original poster described. Tracking would assist this process. And remember, a monopoly is defined by the market. Ergo, if you don't have access to other, competing grocery stores, you are in a monopolistic situation (regardless of whether other groceries compete elsewhere nationally).
Econ 101 doesn't tell you the full story. There isn't a clear linear relationship between price and demand, regardless of what they told you in Econ 101. It's one of the biggest fallacies in modern undergrad economics. See Modigliani for a discussion. The simple linear relationship generally holds true for an individiual, but as utility cannot be compared between individuals, you technically cannot aggregate demand and expect the same simple curve. In reality, it's a higher-order non-linear relationship, leading to demand that may increase when prices go up and fall when prices go down, depending on the price being examined.
Bringing it back on topic, this means that it's not actually as easy as increasing or decreasing prices. You can increase the price and have more people subscribe, simply because you've moved into a different type of consumer who sees the good as now being a luxury (increasing their image, giving them utility). It sounds heretical, but it's true. If you're interested in a fuller explanation about why the foundations of Economics are incorrect, reply to this post.
If you're doing economics, check out Modigliani. If he and his supporters even get enough momentum, a lot of the tenets of micro and macroeconomics are likely to fall (as they're not mathematically sound). Neo-Keynesian economics and the Austrian school of economics are also worth looking at. They normally only teach you basic Keynesian economics and (maybe) monetarism in undergrad at Uni.
The difficulty is that a large part of the final product is the presentation. Presentation not in terms of flashy stuff, but in terms of conceptually mapping things out. This particular model details a process flow and allows users to test various scenarios by changing different elements of the process. Regarding the "links", I was pretty unclear. Basically, there's a dependence flow, where each cell on a spreadsheet is referenced to cells on previous sheets. It's this dependence flow that the model maps out.
I take your point about the spreadsheet being a poor tool for delivering a program, but the problem is it's a strategic model (one of the main value points being it allows users to quickly scan large amounts of numbers and see changes instantly). Without going into too much detail, a DB wouldn't quite fit as well as a spreadsheet for this particular solution. It would in most others I can think of, but not here.
Something you mentioned that I'm curious about is "plugin that has a DOM". What's an example of such a plugin? I'd actually love to move to a browser interface and dump Excel VBA - one of the problems I've identified for the future is one of technological lock-in - they're stuck using Excel for the life of the Model (probably about 5 - 10 years). A browser interface would solve the issue of platform independence while also allowing a much more flexible approach. Have you got any links about plugins under ECMAScript?
You probably didn't miss something stupid - I probably didn't do it the best way possible. I haven't used ECMAScript in a production sense, only played around with it as a minor part of web development. The sandbox issue I agree with - ECMAScript is much easier to sandbox.
The thing I'm not sure about (and here I'm clearly displaying my ignorance) is what capacity ECMAScript has to pull functionality from other applications. Using IE and VB as an example, I could pull Excel functionality and forms from Office 2000 into IE quite easily using ActiveX or another mechanism. But I'd still be using VB to do it. Can ECMAScript give me an Excel spreadsheet within IE? A lot of the modelling work I do needs to have that spreadsheet functionality. I've had a look through MSDN, but can't find anything beyond the same JScript stuff I used before (dates, simple operators, etc).
I can see that it'd be easier with a DB - build the hooks and display selected structured data in IE. No need for a spreadsheet. I can see forms wouldn't be an issue either - just webify. The simulation stuff would be easy to port, it's the links to the spreadsheet that would be difficult (this example is spread out over 30 sheets with somewhere around 2mil calculations).
Any suggestions? Or links?
I may be too late in this discussion, but the value of VBA is that it makes it easy to add additional functionality to documents. I use VB in Excel regularly. Think developing a Monte-Carlo simulation module for a forecasting model. Or a dynamic projection module that automatically selects the best projection method (ARIMA, regression, moving average, etc) based on the characteristics of the data. Or an interface to make handling multiple documents and scenarios easy through meta-data (kludgy, but it works as long as you stay within the rules).
This functionality can be embedded in the spreadsheet or DB for additional value. It would be ideal if VB were a properly designed language, but it's not. It may also have huge security holes, but that's not going to stop me using it. I know it's got problems, but I'll continue to use it. There's some serious business value in VBA.
This value may lock me into using MS products, but what other choices do I have given that all my clients run Office? You're dealing with extensive network externalities created through a defacto standard. Until KOffice or StarOffice can parse or convert VBA, they're pushing uphill.
Damn I wish I could mod this up. Computer science is black and white - things work, or they don't. Social science is grey - it depends on how you approach it. To assess the validity of the results, all you can do is evaluate the methodology chosen while considering the constraints it was done under. It's easy to say "it would have been better to portscan every computer on the Internet three times a day over a period three months", but even ignoring the cost factors in such an approach, servers running Linux may not even be connected to the Internet. His methodology sounds ok to me. There may have been other better ways to do it, but it doesn't look like there were any significant flaws in his approach.
People on /. have asked sampling questions - our normal explanations on how and why we structured our sampling go for 5+ pages of highly technical documentation (that relies on theory most people here probably wouldn't understand or be interested in, such as the benefits of 5 point likert scales vs 7 point, scaling and weighting difficulties, and techniques for accounting for measurement error). You can see why a complete answer to these questions won't be presented here. Our methodology section in our reports will often go for 15+ pages.
As an economist / statistician and part-time programmer / sys-admin, I like the comments about "all things being held equal". The reason we do that is to define the problem in such a way that we can answer it. Interacting with the real-world is a lot messier than coding a program. Stats and research is not easy. It's hard work coming up with something that's defensible and will give you quality outputs.
Three points.
As has already been stated, Gartner asked end-users what they installed on their computers after they bought them. Not what was pre-installed on their computer. Implication: either Linux doesn't have the marketshare zealots want to belive, or accurately assessing server marketshare is difficult. You decide.
Garter is a research agency that has its value locked in its reputation. If it produces poor quality research on a regular basis, people will stop subscribing or purchasing its reports. Short term gains (ie fabricating truths in reports for specific companies) will lead to long term losses (ie loss of reputation leading to loss of credibility and revenue). Are you arrogant enough to believe they haven't thought of this? The reality is, they analysed the information available to them and made the best, least biased forecast as they could.
Garter is not in the business of assessing obscure technical facts. They provide a strategic business perspective on technology. Tech-heads are not their market. They don't care about the operational aspects of the technology. The people who run business (which, most commonly, are not tech-heads and have different skills) are their market. They care about the strategic implications of the technology and longer-term market trends.
You think you could do a better job and produce higher quality research? Don't whinge - go out and do it. If you produce research of such a high calibre, you'll drive Gartner out of business in months. I'm constantly amazed at how many people on Slashdot are willing to spout off on things they know nothing about. I don't tell people how to fly a plane - I realise I don't know how. But, the village idiots on slashdot are always willing to provide legal advice, assume everyone in management is a PHB, that companies never ever know what they're doing, that everything is part of a conspiracy and that anyone who doesn't know how to write a sound compression script using bash is an idiot. On the other hand, some people here have some seriously big clues. You my friend, are not one.
In summary, read the bloody report and get some perspective before spouting off.
Heh ... here I was trying to figure out what you were talking about, claiming that liberals were socialists. Then I remembered:
Being Liberal or having right political leanings is the same as conservatism here in Australia. Supporting Labor or having left political leanings here in Australia is the same as being an American liberal (which is broadly the same as socialism).
Don't mind me, I'll just be standing in the corner looking confused ...
Anecdotal evidence, but I must be really lucky - both of my computers with integrated sound (purchased 1996 and 2000) work under linux. I don't think integrated sound cards are such a big deal now under linux. If you're using a *really* old motherboard with an obscure chipset, I guess you may have difficulties. I didn't though.
But your other point is right - the quality sucks. Stick with a real sound card, unless you just don't care at all. As in, you don't mind having disney sound source quality to play your MP3s.