Lexx was supposed to be "bad". That was the whole point. It was cheese - really wierd smelly cheese with psychodelic fungus growing on it. So what? If I want serious drama, I'll watch West Wing. If I want serious sci-fi I'll watch B5. If I want to see some unique and bizarre sci-fi with horror and BDSM overtones that makes me say "WTF was that?", then Lexx fills the bill.
The less tech-savvy of us will of course assume "this browser sucks, it can't render this page correctly", when it is the page itself that can't be rendered properly within standards guidelines.
Broken pages are something browsers are going to have to deal with. The big thing is that when people say "browser x doesn't render this page correctly", what they're really saying is "browser x doesn't render this page the same way browser y does". This shows a weakness in the standards -- they say what proper HTML is supposed to look like but they do not say how to handle all the error conditions. Browsers are improving, but there's still a lot of work needed to define consistant failure modes.
Exactly my experience: Server-side Java works really well in the cross-platform role. Servlets and EJB's are Java's strength and saving grace. As a client-side platform Java leaves much to be desired.
That attitude is the problem -- focusing on the results instead of the cause. Their real job is to make good movies. If they do their real job correctly then they will make money.
most are doing the exact job they are supposed to do in taking only moderate risks to maximize returns
Let me introduce you to the concept of divirsification. That means you balance high-risk investments with low-risk ones. Only an incompetent stockbroker would tell his customers to only invest solely in one kind of stock.
Just to play devil's advocate, let's admit that are many more potential users of Broadband over Powerline than there are HAMs -- by at least one order of magnitude, if not more. A lot more people will benefit from gaining broadband than will be hurt by losing HAM frequencies. Isn't it the FCC's stated duty to allocate the EM spectrum in order to maximize the public's benefit from it? And, realistically, let's follow the money: there's no money to be made from HAM radio; there's a lot to be made from broadband over powerline.
Just pick a throwaway username / password and use it everywhere you need a registration but don't care about security. Instead of N passwords you need to remember N+1. No big deal.
+1, insightful. If I had mod points (and hadn't already posted to this conversation) I'd mod you up.
SW engineering is still in it's infancy: we've only been writing software for 50 years. Look at Civil Engineering for a comparison. As a craft, it dates back to classical antiquity (EG late Bronze age civs like Ancient Greece and Egypt). It didn't become a true engineering discipline as we know it today until (at least) the Renaissance. (One could argue that engineering, in the modern sense of the word, didn't exist until the Industrial Revolution.)
Exactly correct. Also you have to remember that the amount of effort needed to detect a bug scales exponentially over time: if it takes T time to find B bugs, it's going to take T + T^N time to detect 2B bugs (or conversely, if you find B bugs in T time, you'll only find B + B^(1/N) bugs in 2T time. If N > 1 (and it always is), testing rapidly hits a point of deminishing returns.
I like to think I'm an engineer, not a developer. The problem is not that I don't know how to do good SW engineering, it that I'm usually not allowed to do good SW engineering. Good engineering is expensive in terms of time and money. The people who sign the checks aren't usually willing to pay for it and aren't willing to wait. The sad part is that they're often right: if you can't afford to wait, and you can't afford to pay the price, you have to settle for what you can get and hope that it's good enough to keep you moving forward.
You have 4 main variables in the software development equasion: Time, Quality, Functionality, and Efficiency. Notice that we only measure time, not man-hours or monetarycost. As we know from reading The Mythical Man-Month, we cannot reduce time by adding more people or by spending more money. While we list efficiency as a variable, we really have to treat it as a constant within the scope of a single release cycle. Improvements in efficency are generally very gradual and incremental, and for the most part cannot be effectively implemented in the middle of a release cycle.
I postulate that Time is directly proportional to the product of Quality, Functionality, and Efficiency [T = EQF]. Since E is constant within the scope of a single release, we can't use process improvements or similar techniques to improve quality in the short term. Assuming our goal is to improve quality, we either have to decrease functionality or increase time. Since monetary cost is directly proportional to time (time is money!), managers are very reluctant to give you more time. Furthermore, we are frequently under hard time constraints due to contractual obligations or market pressure. If we can't change time, we either have to sacrifice quality or functionality. Missing functionality is very obvious, whereas low quality isn't necessarily noticable in the short term, so it should be no suprise that quality is almost always takes the back seat to functionality.
Plus, even if 100% of the customers send in the rebates, they get the interest off that money for a couple of months. It's also a nice way to artificially pump up your cash holdings before the next quarterly earnings statement. How much do you want to bet that the rebate checks go out a few days after the quarterly financial statements are filed?
You are 100% correct. In a big server farm, space, heat, and power consumption are major concerns. Assuming your figures are correct and the replacement hardware has similar specs to what they are replacing, they can improve the overall performance of the system by at least 25% at the same power consumption, not counting the additional power savings that would see from the lower heat load. Less heat also translated directly into longer life.
I'll admit that a 6-month replacement cycle is pretty short, but it actually makes sense because they're avoiding the worst of the depreciation. I'm not up on used Mac prices, but x86 server hardware depreciates around 50% per year (refurbished 2 year old x86 servers routinely sell for around 20% - 30% of their original price; refurbished 3 year old gear sells for well under 10% of it's original price. You can get a maxed out Quad processor P-III server for well under $5000 which cost $50K when new.
That said, I don't think that this is a good deal. $200 savings on a $3000 box is only a 6.7% discount for 6 month old hardware; a 20% - 25% discount would be more in line with current market.
Microsoft has a proven track record of rolling roughshod over everyone else's IP rights. Just look at what happened to Stac. They patented a technique for doing transparent real-time compression on a mounted filesystem. Microsoft blatantly and fragrantly violated the patent when they build filesystem compression into thier products. Basically, they said "so sue us" to Stac, and used their legal muscle to keep the case tied up in court until Stac went bankrupt.
If microsoft wouldn't pay licensing fees for a patent which was clearly legitimate, why would they pay out for one as dubious as this one?
Basically a good idea, but what I'd do it using software instead of hardware. Think of a locked-down workstations booting up into a minimal Linux desktop, where all they can do is launch VMWare and load up their own.vmdk image. The only thing the user has write access to is his own virtual disk image. The idea here is that the physical hardware is totally locked down, but the virtual hardware is wide open, giving you the best of both worlds.
Just read your thesis. Very interesting; good work.
I wonder what percentage of web page authors are even actually aware of the fact that there are standards they're supposed to be following? If people are incapable or unwilling to learn to use their own native language correctly, what makes you think they'll be willing or able to use an artificial language correctly? Considering how much web content has basic gramatical errors, I'd say the percentage is pretty low.
[D]o you really believe that 18 year old students are mature enough to [look at the information for what it really is]?
They should be. An 18 y/o is legally an adult; they are supposed mature enough to exercise the rights and duties of citizenship: vote, serve in the military, get married, enter into contracts, and so forth. Furthermore, we're talking about college students, who are a group of people who are supposed to have already mastered basic intellectual skills like reading comprehension.
Call me old fashioned, but I don't think that a student who needs remedial instruction, either in his or her native language or in basic math, should be admitted into a degree program, nor should remedial coursework count towords graduation. The fact that so many high-school graduates are unprepared for college is a pretty scathing indictment of our educational system(*).
(*) Note to parents: you are part of the educational system. Schools and teachers exist to help you educate your children, not to educate them for you. Sending your kids off to school does not relieve you of your personal duty to raise your children.
I recall articles back in the '80's on how to double the RAM in Apples and other microcomputers by stacking more DIP RAM IC's atop the existing ones and running wires for the additional address lines
You recall correctly. I performed the piggyback RAM hack on my Amiga 1000 in '91 or '92. Ahh... those were the days, when upgrading your computer meant you had to break out the soldering iron and not just swap out a card/module.
Well, considering how many clueless gits shop at Best Buy, I'd say that's just a piece of clever marketing. Newbie catches on to the idea that their magic computer box needs a magic surge supressor strip. For somone who thinks an electron is the size and shape of a pea (to borrow a quote from the Grand Master), a quick blurb reassuring them that this magical device will do what they want it to do is not an unreasonable thing.
The absolute ignorance most people have of the technology they rely on for everyday life is staggering. How many drivers have no clue to do simple maintenance jobs like change an air filter or do an oil change, let alone something scary-sounding like putting on new brake pads? A lot of people would consider re-wiring an electric lamp to be as mysterious and complex as brain surgery. People are just willfully stupid.
IIRC, Microsoft is already the largest Comcast shareholder, owning approximately 33% of the company last time I checked. BillG owns another substantial chunk of Comcast stock under his own name, too. I remember reading that all the major Microsoft insiders were investing heavily in cable companies.
Insightful. I was just thinking along the lines of a sighted user with a text-only browser. Since in this cases the graphic is basically doing the same thing as a
tag, you'd want the ALT text to produce the same output that you'd get from a bulleted list. How is a bullet handled in text-to-speech?
What is needed is built-in validity checkers in browsers and design tools. If a page does not conform to it's declared DOCTYPE (or doesn't declare a DOCTYPE at all), then the browser should display a visible warning indicator. This would give web "designers" instant feedback that their pages are FUBAR and would encourage users to complain when given buggy web pages.
Don't get me wrong, I totally believe in the idea of open standards and interoperable, browser-neutral web sites. However, it seems to me that nowadays the W3C is more interested in pushing thier political agenda and ideas than they are in codifying and standardizing widely used technology.
Political agendas aside, a standards body has to recognize what technologies and extensions are actually being used "in the wild" and incorporate them into the standard. Whether you like M$ or not, you have to recognize that MSIE is the de-facto standard browser today. A W3C standard means jack shit if MS doesn't implement it, and a W3C standard that doesn't address commonly-used MSIE extensions devalues WC3's credibility and usefulness as a standards body. All de-facto "standard" web languages need to be standardized, just as the competing implementations of JavaScript were standardized into ECMAscript.
While being able to use surveillance footage to corroborate an alibi is definately a good thing, the negative concequences of widespread surveillance, especially when combined with automated identification (EG, facial recognition) and searh & retrieival capibilities.
Imagine the police are trying to find a serial rapist. They pull up the surveillance tapes for the areas surrounding the crime scenes and compile a list of everyone who was in the right places at the right times. Anyone on that list, whether they are actually the rapist or not, is going to be subjected to VERY close scrutiny. Once you get on the cops radar, it's pretty likely that they're going to keep digging until they find SOME dirt on you, even if it's something totally unrelated to the original case. If you think that being mistakenly identified as suspect in a major case isn't traumatic, I'd suggest you review the case of Richard Jewel
Slashdot Mirror
"I don't get it" != "Bad".
Exactly my experience: Server-side Java works really well in the cross-platform role. Servlets and EJB's are Java's strength and saving grace. As a client-side platform Java leaves much to be desired.
Just to play devil's advocate, let's admit that are many more potential users of Broadband over Powerline than there are HAMs -- by at least one order of magnitude, if not more. A lot more people will benefit from gaining broadband than will be hurt by losing HAM frequencies. Isn't it the FCC's stated duty to allocate the EM spectrum in order to maximize the public's benefit from it? And, realistically, let's follow the money: there's no money to be made from HAM radio; there's a lot to be made from broadband over powerline.
Just pick a throwaway username / password and use it everywhere you need a registration but don't care about security. Instead of N passwords you need to remember N+1. No big deal.
SW engineering is still in it's infancy: we've only been writing software for 50 years. Look at Civil Engineering for a comparison. As a craft, it dates back to classical antiquity (EG late Bronze age civs like Ancient Greece and Egypt). It didn't become a true engineering discipline as we know it today until (at least) the Renaissance. (One could argue that engineering, in the modern sense of the word, didn't exist until the Industrial Revolution.)
Exactly correct. Also you have to remember that the amount of effort needed to detect a bug scales exponentially over time: if it takes T time to find B bugs, it's going to take T + T^N time to detect 2B bugs (or conversely, if you find B bugs in T time, you'll only find B + B^(1/N) bugs in 2T time. If N > 1 (and it always is), testing rapidly hits a point of deminishing returns.
You have 4 main variables in the software development equasion: Time, Quality, Functionality, and Efficiency. Notice that we only measure time, not man-hours or monetarycost. As we know from reading The Mythical Man-Month , we cannot reduce time by adding more people or by spending more money. While we list efficiency as a variable, we really have to treat it as a constant within the scope of a single release cycle. Improvements in efficency are generally very gradual and incremental, and for the most part cannot be effectively implemented in the middle of a release cycle.
I postulate that Time is directly proportional to the product of Quality, Functionality, and Efficiency [T = EQF]. Since E is constant within the scope of a single release, we can't use process improvements or similar techniques to improve quality in the short term. Assuming our goal is to improve quality, we either have to decrease functionality or increase time. Since monetary cost is directly proportional to time (time is money!), managers are very reluctant to give you more time. Furthermore, we are frequently under hard time constraints due to contractual obligations or market pressure. If we can't change time, we either have to sacrifice quality or functionality. Missing functionality is very obvious, whereas low quality isn't necessarily noticable in the short term, so it should be no suprise that quality is almost always takes the back seat to functionality.
Plus, even if 100% of the customers send in the rebates, they get the interest off that money for a couple of months. It's also a nice way to artificially pump up your cash holdings before the next quarterly earnings statement. How much do you want to bet that the rebate checks go out a few days after the quarterly financial statements are filed?
I'll admit that a 6-month replacement cycle is pretty short, but it actually makes sense because they're avoiding the worst of the depreciation. I'm not up on used Mac prices, but x86 server hardware depreciates around 50% per year (refurbished 2 year old x86 servers routinely sell for around 20% - 30% of their original price; refurbished 3 year old gear sells for well under 10% of it's original price. You can get a maxed out Quad processor P-III server for well under $5000 which cost $50K when new.
That said, I don't think that this is a good deal. $200 savings on a $3000 box is only a 6.7% discount for 6 month old hardware; a 20% - 25% discount would be more in line with current market.
If microsoft wouldn't pay licensing fees for a patent which was clearly legitimate, why would they pay out for one as dubious as this one?
Basically a good idea, but what I'd do it using software instead of hardware. Think of a locked-down workstations booting up into a minimal Linux desktop, where all they can do is launch VMWare and load up their own .vmdk image. The only thing the user has write access to is his own virtual disk image. The idea here is that the physical hardware is totally locked down, but the virtual hardware is wide open, giving you the best of both worlds.
I wonder what percentage of web page authors are even actually aware of the fact that there are standards they're supposed to be following? If people are incapable or unwilling to learn to use their own native language correctly, what makes you think they'll be willing or able to use an artificial language correctly? Considering how much web content has basic gramatical errors, I'd say the percentage is pretty low.
The absolute ignorance most people have of the technology they rely on for everyday life is staggering. How many drivers have no clue to do simple maintenance jobs like change an air filter or do an oil change, let alone something scary-sounding like putting on new brake pads? A lot of people would consider re-wiring an electric lamp to be as mysterious and complex as brain surgery. People are just willfully stupid.
Surfing Pr0n on your employer's computer isn't a good idea, so you use your laptop and someone else's WiFi net to get your lunchtime b00bie fix :-)
IIRC, Microsoft is already the largest Comcast shareholder, owning approximately 33% of the company last time I checked. BillG owns another substantial chunk of Comcast stock under his own name, too. I remember reading that all the major Microsoft insiders were investing heavily in cable companies.
tag, you'd want the ALT text to produce the same output that you'd get from a bulleted list. How is a bullet handled in text-to-speech?
What is needed is built-in validity checkers in browsers and design tools. If a page does not conform to it's declared DOCTYPE (or doesn't declare a DOCTYPE at all), then the browser should display a visible warning indicator. This would give web "designers" instant feedback that their pages are FUBAR and would encourage users to complain when given buggy web pages.
If you want to be pedantic, a more appropriate alt text for the triangle icon would be ">" or maybe "|>" or even "*".
Political agendas aside, a standards body has to recognize what technologies and extensions are actually being used "in the wild" and incorporate them into the standard. Whether you like M$ or not, you have to recognize that MSIE is the de-facto standard browser today. A W3C standard means jack shit if MS doesn't implement it, and a W3C standard that doesn't address commonly-used MSIE extensions devalues WC3's credibility and usefulness as a standards body. All de-facto "standard" web languages need to be standardized, just as the competing implementations of JavaScript were standardized into ECMAscript.
Imagine the police are trying to find a serial rapist. They pull up the surveillance tapes for the areas surrounding the crime scenes and compile a list of everyone who was in the right places at the right times. Anyone on that list, whether they are actually the rapist or not, is going to be subjected to VERY close scrutiny. Once you get on the cops radar, it's pretty likely that they're going to keep digging until they find SOME dirt on you, even if it's something totally unrelated to the original case. If you think that being mistakenly identified as suspect in a major case isn't traumatic, I'd suggest you review the case of Richard Jewel