What stops 7-Eleven from denying you a coffee unless you also pay for the cup?
This parallel only works if your local 7-Eleven has a deal going with your local government that forbids any other food stores from operating in the area, and it's illegal for you to drive to a remote store for your coffee.
The phone line coming into our house is owned by Verizon, and it's illegal for anyone else (including me;-) to run such lines down the street to our house. A few other kinds of lines are legal, such as electrical or cable, but those are also restricted to just one or two legal suppliers.
I guess we can be glad that food doesn't work this way (yet).
I think it depends on what sort of agreement (legal or otherwise;-) the phone company has with the local political structure. Hereabouts, we just got speakeasy service about a year ago, after years of their telling us that they couldn't supply service to our town. Their first service here was "naked" DSL. They started VoIP a few months ago.
Actually, VoIP worked here fine all the time. That is, I installed Skype on linux and OSX, and it worked fine over the DSL IP service. But, of course, that didn't give access to local phone lines, calling the house next door was "long distance", and calling 911 was right out.
Methinks the phone company is just pulling any strings they can to prevent the inevitable disappearance of traditional POTS. Of course, they use VoIP everywhere internally. They just don't want customers to use it, because it eliminates their century-old copper-wire-and-switch phone service. It's doomed anyway, but they can probably keep charging people for it for some decades, if they play their usual monopolistic game correctly.
Actually, corporations are legal persons and as such have some of the same independent rights that people ('natural persons') have.
In this cased, corporations seem to have a 'right" that no mere human has.
In particular, the phone line to my house is owned by Verizon, and I have no legal rights to it whatsoever. They have a monopoly here, and the FCC seems to be saying that they can force me to pay for services that I don't want. No individual human has the right to do this to me; only a corporation has that right.
Actually, we have DSL through speakeasy now (no port blocking, friendly and competent support for linux and Mac users, yadda, yadda). We recently started VoIP through them, and it works pretty well. I wouldn't be surprised if Verizon is now planning to make us pay them another $30 per month for phone service that we're getting from someone else.
There is an ongoing discussion hereabouts of the prospect that the FCC will decide that Verizon no longer has to lease their lines to companies like speakeasy. Then we'd be stuck with Verizon's crappy DSL, port blocking and all, unless we want to pay double for "business" service. And they'd charge us for a phone line whether or not we use it.
Sure wish that I, as a citizen, had the "rights" of a corporation, and could control the lines to everyone in our neighborhood...
... in favor of projects that will yield short-term military results.
If they can predict beforehand what a project will yield, then it's not research; it's engineering. So they should change their name from DARPA to DAPA.
Would you want to stare at a lit flourescent tube for most of the day? Would would expect your customers to do this? Well, a CRT is a flourescent tube, and a window with a white background is a fully-lit flourescent tube. Guess what this does to your eyes...
The common white background is a visual metaphor for paper. But paper isn't a glowing flourescent tube. And note that publications that are often read outdoors (newspapers, paperback books) are usually printed on an off-white, beige paper. This isn't because the paper is cheap or anything; it's done intentionally to lessen the assault on readers' eyes in sunlight and other bright light. The publishing industry has understood this for decades. The computer industry can't be bothered, though, and continues to deliver systems in which nearly everything defaults to a bright white background.
The solution is to find out how to override this whenever possible. With browsers, you can set the background to a neutral grey or beige or pale yellow, and check the "Always use my colors" setting. This will work for almost all web pages, and will greatly lessen their assault on your eyes.
LCD displays are a lot better than CRTs. So far, they aren't quite as intense, which can cause readability problems in bright light, but it's easier on your eyes. Even there, though, it's still best to override default colors whenever you can, and use a neutral background. Web designers might not like this, but they have no right to assault your eyes by forcing a fully-white background color.
Isn't this really just an argument for greater use of formal methods throughout the software world?
You're right, of course. But try persuading Bill Gates and Steve Ballmer that everyone should have access to not only their code, but also a formal analysis of all of it. A major portion of software development is on their system, and any discussion of formal methods there is pretty much irrelevant, even if software managers would permit it. There's little point of formal methods when all the low-level stuff is firmly and intentionally hidden from developers.
OTOH, in the real-time arena, formal methods are fairly standard already. This is especially important now that we're seeing things like microprocessors controlling major parts of autos and airplanes. And then there's the horror expressed by people faced with the prospect of developing software on the MS Windows that their management has mandated for future models...
Another island nation with similar problems is the Maldives. The mean elevation there is about 2 meters. You can guess what global warming means to them. The conventional estimates are that the oceans will rise 1/2 to 1 meter during this century.
Another interesting case is Scandinavia, which will benefit from the ocean rising. A historic problem there has been that the land is still rebounding from the melted glaciers, and rises of around 1 meter per century are about average. This means that their harbors have always been drying out, and towns on the shore have to gradually migrate downhill. There are ruins of seaports 10 km or more from the current shore. For the next century, their shorelines should be fairly stable. And as the climate warms, they'll have a longer growing season. So overall, the warming is to their benefit.
Maybe they can invite the Tuvaluans to become Norwegian...
Recently I ran across a cute bit of history to illustrate how shoddy builders can be. It seems that Boston was the first known city to pass a building ordinance. Some time in the mid 1600's, they passed a law that made wooden chimneys illegal.
Think about this. They found a need to pass that law. This means that builders were building houses with wooden chimneys. Those builders knew exactly what they were doing. They did it anyway, and managed to sell the houses to gullible buyers.
Few pieces of software get as bad as this.
Actually, these days Boston is facing a similar situation. The "Big Dig" project to put the main north-south expressway underground is nearly complete. But it's springing lots of leaks. It's mostly below the water table, and the walls in a lot of places have very sub-standard construction. In some cases, they apparently filled what should have been concrete-filled spaces with construction rubble, and poured concrete around it. Joints and seams weren't properly sealed. And so on. Warnings from people involved in the construction were casually buried and never read by the right people. The managers responsible have long since moved on to new jobs, and the new managers are seeing a lot of finger pointing their way while they try to recover from the mess left by their predecessors. Sound familiar, anyone?
And there was a nice new bridge, which did have several design flaws. They were all minor, and could be fixed, for a small extra charge. But, strictly speaking, the bridge wasn't built correctly the first time. The engineering was good enough that the problems could be fixed before they became serious.
... bridge-building as a discipline is thousands of years old; software engineering is a few decades old. And that bridges typically serve a single, simple purpose, whereas software depends on the complex interaction of thousands of elements that are not well-understood.
All true. And there's another important reason that bridges and buildings don't need as much testing as software: With bridges and buildings, you can order all the components to spec and (aside from the occasional corruption), you can rely on the components matching the specs. You order steel beams of a given size, alloy and strength, that's what you get. Bolts have the composition and shear strength that you ordered. You can use the numbers in your models to calculate the strength of the resulting structures.
With software, for everyone (except Open Source programmers), you are required to use components (OS, compilers, runtime libraries) that are proprietary, can't be inspected, and very often fail to correctly implement what's in the sketchy documentation that's available. You must test everything, so you can discover the underlying failures of the mandated components. If you find that something fails in a critical function, you can beg and plead to have it replaced with something that works, but the vendor can take their good time, or deny your request entirely. Sometimes you can spend extra time rewriting the component, but if it's sufficiently low-level (i.e., inside the OS), this isn't even possible.
So, if you're a programmer, most of your structures are build on unstable, shifting sand. Your tools are inadequate for the job, break at inopportune times. The pieces of your structure fail in unpredictable ways. You are forced to do extensive testing, because engineering calculations just aren't possible when all your materials have unknown defects.
And the jobs are usually so complex that exhaustive testing would take centuries on the fastest machines. So you deliver something that you know will fail. But you shrug, because you understand that it's what the customer wanted. Otherwise, they wouldn't be buying computers that were built to hide the details from the programmers. Everyone understands that if a foundation is unstable, the whole structure is unstable. So if they are buying unstable computers, they must want the software that runs on it to be unstable, too. Right?
I've done a few jobs where I could do a proper engineering job, because I had precise specs (i.e., access to the code) at all levels below my code. This is possible on a few real-time systems. But I don't even dream about it with most jobs, because I know up front that I'm not permitted to see the inner details of some of the things that I'm required to use. I know very well what the end result will be. But it's what the client wants.
Heh. Maybe they were counting Al Gore as the other president, since he was the one who got the most votes in 2000.;-)
Too bad he was such a lazy bum, and delegated the entire job to Dubya.
Oh, well; they say that one part of being a good manager is knowing how to delegate authority and responsibility to others. At least Al doesn't have to take the blame for Iraq.
What will happen when countries like Tuvalu (.tv) reach technological savyness and find that their entire TLD has been used up by TV networks,...
Well, by then global warming will have put all but a few square meters of Tuvalu underwater.
Seriously; here's a story about it from back in 2001. Tuvalu's government is asking other countries to accept the Tuvaluans as citizens, without much luck so far.
I found this and other reports on the topic fairly quickly by googling for "nation flooded rising ocean". I do remember reading about it when the story first came out, but it didn't get a lot of coverage in the MSM (mainstream media).
When their islands are finally abandoned, it's anyone's guess what will happen legally to their TLD. I'd guess that it'll be a "land grab" by various corporations, since the.tv domain has obvious commercial value.
That hadn't occurred to me (and I've never seen it mentioned anywhere that I recall). It did work, with a rather awkward hand position that's reminiscent of some guitar chords. Then another reply suggested a Preferences tweak that moved the key mappings around to something that's a lot easier to use.
I'll do a bit more experimenting. It sure takes a while to stumble onto all these tricks.
Hey, thanks! That really improved the behavior. Now the F keys function as function keys.
I must say that the description of that checkbox in the Preferences window leaves a bit to be desired. There's no way I'd ever guess what it does after reading the description.
Lots of eye candy, but the Mac is rather deficient in the way it documents all this stuff. The main way to learn seems to be to ask others, until you find someone who has stumbled onto an answer. Much like other computer systems that I could name, of course.
I used to think that raw processor clock rate was everything, but I've seen the Macs run circles around Intel machines with half-again as much clock rate.
That's a standard misconception that marketers routinely take advantage of. Clock rate comparisons are only meaningful if the hardware (cpu, bus, etc) are very closely related. It takes more cycles to do something on an Intel-type cpu than on a PPC (but fewer cycles than on a RISC cpu). A wider bus means fewer cycles are needed. This is neither good nor bad; it's just different design.
But comparing clock rates is useful as a marketing-fog trick when selling to people who don't understand what clock rates mean.
Re:BOFH = Bastard Operator From Hell
on
Return of the Mac
·
· Score: 1
Heh. How many computer users these days even know what the term "operator" used to signify? It's true that there are still lots of people who work with mainframes, but we probably don't have too many of them hanging out here. For most people, expanding BOFH still leaves them baffled, as they have no referent for the second word.
Actually, we've read a lot of comments here and elsewhere about how one of the signs of a growing success of unix/linux/*x in The Market is that Usenix is more and more populated by people wearing suits. The implication of this is well-understood.
Any real computer geek would understand that what the "top dogs" carry shouldn't mean a whole lot to anyone else. The idea that there's one best computer is as silly as the idea that there's one best car or ice-cream flavor.
It's possible that there is a worst computer, but it's not possible that there's a best computer. That depends greatly on what you want it to do. To some people I know, their Blackberry is the best computer they've ever used, and for their uses, they're right.
Doesn't work at all on my Mac Powerbook. It does the same thing as F2 alone - The screen brightness control appears in the center of the screen, and repeated clicks on F2 (with or without the CTRL key) brightens the screen one notch.
So is there some Preference somewhere that'll make CTRL-F2 give the menu bar? It sure would be useful. I find that moving to the top bar and then back is a major time waster. And it's made even worse by all the times that I'm wrong about what window has the focus. There's little visual difference on the screen, just a slightly darker title bar. It takes extra time to check the name at the upper left, and all too often the app has several windows open anyway, so it's still wrong.
There are nice things about the Mac, but after using one a lot for a couple of years, I still find it frustratingly slow compared to unix/linux with X Windows (after 5 or 10 minutes of configging on a new machine).
I'd half agree with you. The Internet is sorta like a giant public library. But the overlap with the brick-and-mortar libraries isn't all that complete yet, so we need both of them.
As for "relying on the internet as a valid source of information", I'm reminded of a good comment I read some years ago:
The libraries of the world are the repository of all the world's wisdon - and most of its nonsense.
This applies equally to the Internet. In both cases, if you trust them to deliver only wisdom, you're a fool. You have to learn to make your own judgements. If you can do that, both are good tools. And with the advent of search sites, finding obscure information is slowly becoming easier.
Actually, the Internet isn't a material threat to libraries yet. The business that's more threatened is the traditional news media. Nearly all of their nonsense and heir occasional valid information is now available online. If you can find it. And you can easily find stuff that's not provided by your local media, so it's now a lot more difficult for the powerful people to pull the wool over our eyes.
Of course, what really caused the Internet explosion was porn...;-)
Um? I'm pretty sure every filesystem I've used in the past decade or so has been at least case-preserving. Actual case sensitivity is more down to taste.
Um... One lesson I've learned is to never, ever do an rsync between a Mac and any unix or linux system. That produced several major disasters, until we learned about the case insensitivity in the default OSX file system. If a directory has files "foobar" and "FooBar", on OSX you will get only one file, and there's a 50% chance that its contents will be the wrong one. Then, when you rsync again, there's another 50% chance that the OSX file will overwrite the unix/linux file with the wrong contents, giving you two files that are now identical.
The other rsync disaster is that, if any directory contains 8-bit chars in their names, they will be unreadable gibberish on OSX. Then a second rsync will copy the gibberish filename back to the unix/linux system.
Yes, I've asked about solutions to these in several Mac-related fora. The repliest mostly call me an idiot for using such file names (though I'd usually mentioned that I didn't personally name the files in question;-). But this is expected; what's disappointing is that nobody seems to know of a solution to either problem.
Yes, OSX does have a unix-like filesystem. Some Mac apps don't work right on it. And we can't demand that customers use it.
And I do like the idea of having a filesystem that uses UTF-8 for the encoding. But I can't tell all our linux customers that they have to have that. (And I can't tell the MS users anything.;-)
If anyone knows of successful solutions to these problems (that don't require changing customers' file systems), please metion them in a reply. I'd expect that eventually rsync will get options naming the char incodings, but this doesn't seem to exist at present. The caseless filename matching is just a disaster all around.
If you're new to Macs, don't even think of using rsync...
Agreed, there is balkanization in copy/paste methods in KDE. But OSX is not entirely consistent either, if you use any X apps.
I've been using OSX for a couple of years now, and I'm getting more baffled about the claims that it's consistent. In the case of cutting, you can see one of the real problems by opening up a few browser windows, doing a bit of browsing, and then try to edit or copy a URL.
One major problem is selecting what you want selected. A single click sometimes selects the entire URL, sometimes selects nothing (but inserts the cursor), and sometimes selects some "field" that seems to be different at different times. If the click selected the entire URL, you have to slowly click until that goes away and one of the other behaviors happens. Then you can drag the pointer to get a selection. Sometimes. Sometimes dragging the pointer does nothing, and you have to start again.
This behavior happens with all the browsers. It's not just the browsers, though, because there are lots of little popup windows in other apps whose input fields show the same unpredictable selection behavior.
I first thought this was some bug on my machine. But I've used a few other machines, and watched others using them, and it seems to be a problem for everyone. At least for everyone who actually uses copy-and-paste; some users never seem to do this.
Not that I'm especially criticising OSX software, of course; you can produce similarly frustrating selection behavior on MS Windows very easily. I've seen in with X Windows, but there it seems an occasional bug rather than normal behavior. A bigger problem there is all the text that can't be selected at all. This happens on all the windowing systems, of course. For example, where can you select text from a title bar? But statistically it's worse on X Windows, with a lot of ordinary-looking text in windows that can't be selected.
Overall, I've found X Windows the most predictable, in the sense that when I learn some new trick, I quickly push it into my subconscious, because after a dozen times, I can use it without thought. MS Windows is the least predictable, with lots of picky details slightly different between apps so that I can't ever quite stop thinking and frequently backtracking when I clicked something wrong. It's not just me, because when I watch experienced MS Windows users, they stumble the same way, and you hear a lot of quite cursing as they try to get it to work right. OSX is somewhere in between; often it "just works", but sometimes it gets really random.
Now let's see if this gets a "flamebait" rating, because it's not sufficiently praising of either the Mac or linux.;-)
Now, linux-heads love choice and more power to them for that. BUT such up-front confusion with linux is not the way to win over the general public.
Oh, nonsense. There are a zillion examples (;-) of similar choices presented to the public, and you rarely if ever hear anyone but a few marketers complaining.
In most of the Western world, especially the US, if you want to but an auto, you have at least a hundred choices, many of them very similar. People don't complain about this; they discuss the differences between models at the drop of a hat.
It's common to observe that supermarkets have an insane variety of many products, frm bread and cereals to apples to soft drinks to.... This does not give an advantage to the small neighborhood street-corner store; people go to supermarkets precisely because of the variety. No matter what your family likes, you'll find it there.
Similarly, people overwhelmingly go to the large clothing stores in malls, because that's where you'll find a huge choice in what chothes you can buy. And they'll let you try things on before you buy.
Imagine a restaurant with just one fixed meal. Not even McDonalds does that.
There's very little real evidence anywhere that people, even stupid people, are bothered by a variety of choices. They're only bothered if they don't have a way to choose. (This is a problem for ethnic restaurants with words that customers don't recognize. That's why they'll usually provide explanations.)
We do hear this sort of claim a lot in the computer biz. But I think it's totally bogus. It's a way of saying "I don't want to buy your system, and I'm not going to admit why (because it'd make me sound really stupid;-). So instead I'll pick what in all other areas would be a great advantage, the variety you offer, and I'll criticize you for that. I'm smart enough to know that if you're stupid enough to believe me and offer a system without choices, I'll just tell you that I don't like the single choice you offered me, because it's not 101% identical with my old system that I'm used to."
You can't win such an argument no matter what you do.
But you're a lot more likely to win if you can offer people what they really want. And other consumer products show clearly that there is never a single version that everyone wants. You have to offer choices. And you have to make it easy for people to browse a bit and pick one that they'll like. Just like they do in clothing stores and food stores.
Heh. Actually, you should have done a topical version of the very old joke:
Q: Was the Sept 11 attack due to the American public's ignorance of the history of its government's actions in other parts of the world, or was it due to the public's apathetic response to reports on the topic?
A: We don't know, and we don't care.
During the last election, it became clear that this wasn't just a joke, when several surveys reported that around 80% of the Americans who voted for Bush were unable to give correct answers to any questions about Bush's policies. About 25% of the Kerry voters were equally ignorant. And this ignorance was mostly because they couldn't be bothered to read any of the easily available information on the topic; they only got their election info from television. So ignorance and apathy were indeed the order of the day.
What I've found effective is to reply to a Word doc by pointing out that it is encoded via a proprietary, patented scheme, and here in the US, the DMCA says that I can be fined $500,000 and sent to a federal prison for 5 years for decoding it with unauthorized software. Granted, nobody has yet been fined like this for decoding a Word doc, but I don't want to be the test case.
This always gets me a copy in a format that I can legally read.
(Actually, it's more likely that the courts would toss this part of the law. After all, if you sent me the doc, presumably you want me to read it. But again, I don't want to be the test case. I'd like to save my money for something more valuable.)
Look a few miles north, in Hanover. That's where you'll find Dartmouth College. Lebanon is effectively a suburb of Hanover, inasmuch as "suburb" has meaning in such a rural area. But this isn't some remote redneck area. It's a rich-private-school-in-the-country area.
Well, IBM has done just fine with this business model.;-)
Of course, IBM has historically made a lot of money selling hardware. They give away a lot of software as a way of encouraging people to buy more hardware.
But they have also sold software support contracts. Details are hard to come by, but I've read a number of claims that IBM makes a lot more money from support than from sales. This is rarely broken down between hardware support and software support, so it's hard to know just how profitable their software support is. But they clearly consider "support" a profit center.
This has been their approach with AIX all along, and it's fairly clear that this is why they're getting behind linux. They've listened to the FUD saying that businesses won't use linux because they demand supported software. IBM's sales people look at this and think "Hey, we don't have to pay the programmers, we just need to train the support staff, and we can rake in the money."
This isn't a gamble for them, because they've been doing this sort of thing for decades. They know how to make it profitable. Especially when they don't have to pay the programmers.
Slashdot Mirror
What stops 7-Eleven from denying you a coffee unless you also pay for the cup?
;-) to run such lines down the street to our house. A few other kinds of lines are legal, such as electrical or cable, but those are also restricted to just one or two legal suppliers.
This parallel only works if your local 7-Eleven has a deal going with your local government that forbids any other food stores from operating in the area, and it's illegal for you to drive to a remote store for your coffee.
The phone line coming into our house is owned by Verizon, and it's illegal for anyone else (including me
I guess we can be glad that food doesn't work this way (yet).
I think it depends on what sort of agreement (legal or otherwise ;-) the phone company has with the local political structure. Hereabouts, we just got speakeasy service about a year ago, after years of their telling us that they couldn't supply service to our town. Their first service here was "naked" DSL. They started VoIP a few months ago.
Actually, VoIP worked here fine all the time. That is, I installed Skype on linux and OSX, and it worked fine over the DSL IP service. But, of course, that didn't give access to local phone lines, calling the house next door was "long distance", and calling 911 was right out.
Methinks the phone company is just pulling any strings they can to prevent the inevitable disappearance of traditional POTS. Of course, they use VoIP everywhere internally. They just don't want customers to use it, because it eliminates their century-old copper-wire-and-switch phone service. It's doomed anyway, but they can probably keep charging people for it for some decades, if they play their usual monopolistic game correctly.
Actually, corporations are legal persons and as such have some of the same independent rights that people ('natural persons') have.
...
In this cased, corporations seem to have a 'right" that no mere human has.
In particular, the phone line to my house is owned by Verizon, and I have no legal rights to it whatsoever. They have a monopoly here, and the FCC seems to be saying that they can force me to pay for services that I don't want. No individual human has the right to do this to me; only a corporation has that right.
Actually, we have DSL through speakeasy now (no port blocking, friendly and competent support for linux and Mac users, yadda, yadda). We recently started VoIP through them, and it works pretty well. I wouldn't be surprised if Verizon is now planning to make us pay them another $30 per month for phone service that we're getting from someone else.
There is an ongoing discussion hereabouts of the prospect that the FCC will decide that Verizon no longer has to lease their lines to companies like speakeasy. Then we'd be stuck with Verizon's crappy DSL, port blocking and all, unless we want to pay double for "business" service. And they'd charge us for a phone line whether or not we use it.
Sure wish that I, as a citizen, had the "rights" of a corporation, and could control the lines to everyone in our neighborhood
... in favor of projects that will yield short-term military results.
If they can predict beforehand what a project will yield, then it's not research; it's engineering. So they should change their name from DARPA to DAPA.
One of the better comments I've seen about this:
...
Would you want to stare at a lit flourescent tube for most of the day? Would would expect your customers to do this? Well, a CRT is a flourescent tube, and a window with a white background is a fully-lit flourescent tube. Guess what this does to your eyes
The common white background is a visual metaphor for paper. But paper isn't a glowing flourescent tube. And note that publications that are often read outdoors (newspapers, paperback books) are usually printed on an off-white, beige paper. This isn't because the paper is cheap or anything; it's done intentionally to lessen the assault on readers' eyes in sunlight and other bright light. The publishing industry has understood this for decades. The computer industry can't be bothered, though, and continues to deliver systems in which nearly everything defaults to a bright white background.
The solution is to find out how to override this whenever possible. With browsers, you can set the background to a neutral grey or beige or pale yellow, and check the "Always use my colors" setting. This will work for almost all web pages, and will greatly lessen their assault on your eyes.
LCD displays are a lot better than CRTs. So far, they aren't quite as intense, which can cause readability problems in bright light, but it's easier on your eyes. Even there, though, it's still best to override default colors whenever you can, and use a neutral background. Web designers might not like this, but they have no right to assault your eyes by forcing a fully-white background color.
Isn't this really just an argument for greater use of formal methods throughout the software world?
...
You're right, of course. But try persuading Bill Gates and Steve Ballmer that everyone should have access to not only their code, but also a formal analysis of all of it. A major portion of software development is on their system, and any discussion of formal methods there is pretty much irrelevant, even if software managers would permit it. There's little point of formal methods when all the low-level stuff is firmly and intentionally hidden from developers.
OTOH, in the real-time arena, formal methods are fairly standard already. This is especially important now that we're seeing things like microprocessors controlling major parts of autos and airplanes. And then there's the horror expressed by people faced with the prospect of developing software on the MS Windows that their management has mandated for future models
Another island nation with similar problems is the Maldives. The mean elevation there is about 2 meters. You can guess what global warming means to them. The conventional estimates are that the oceans will rise 1/2 to 1 meter during this century.
...
Another interesting case is Scandinavia, which will benefit from the ocean rising. A historic problem there has been that the land is still rebounding from the melted glaciers, and rises of around 1 meter per century are about average. This means that their harbors have always been drying out, and towns on the shore have to gradually migrate downhill. There are ruins of seaports 10 km or more from the current shore. For the next century, their shorelines should be fairly stable. And as the climate warms, they'll have a longer growing season. So overall, the warming is to their benefit.
Maybe they can invite the Tuvaluans to become Norwegian
Recently I ran across a cute bit of history to illustrate how shoddy builders can be. It seems that Boston was the first known city to pass a building ordinance. Some time in the mid 1600's, they passed a law that made wooden chimneys illegal.
Think about this. They found a need to pass that law. This means that builders were building houses with wooden chimneys. Those builders knew exactly what they were doing. They did it anyway, and managed to sell the houses to gullible buyers.
Few pieces of software get as bad as this.
Actually, these days Boston is facing a similar situation. The "Big Dig" project to put the main north-south expressway underground is nearly complete. But it's springing lots of leaks. It's mostly below the water table, and the walls in a lot of places have very sub-standard construction. In some cases, they apparently filled what should have been concrete-filled spaces with construction rubble, and poured concrete around it. Joints and seams weren't properly sealed. And so on. Warnings from people involved in the construction were casually buried and never read by the right people. The managers responsible have long since moved on to new jobs, and the new managers are seeing a lot of finger pointing their way while they try to recover from the mess left by their predecessors. Sound familiar, anyone?
And there was a nice new bridge, which did have several design flaws. They were all minor, and could be fixed, for a small extra charge. But, strictly speaking, the bridge wasn't built correctly the first time. The engineering was good enough that the problems could be fixed before they became serious.
... bridge-building as a discipline is thousands of years old; software engineering is a few decades old. And that bridges typically serve a single, simple purpose, whereas software depends on the complex interaction of thousands of elements that are not well-understood.
All true. And there's another important reason that bridges and buildings don't need as much testing as software: With bridges and buildings, you can order all the components to spec and (aside from the occasional corruption), you can rely on the components matching the specs. You order steel beams of a given size, alloy and strength, that's what you get. Bolts have the composition and shear strength that you ordered. You can use the numbers in your models to calculate the strength of the resulting structures.
With software, for everyone (except Open Source programmers), you are required to use components (OS, compilers, runtime libraries) that are proprietary, can't be inspected, and very often fail to correctly implement what's in the sketchy documentation that's available. You must test everything, so you can discover the underlying failures of the mandated components. If you find that something fails in a critical function, you can beg and plead to have it replaced with something that works, but the vendor can take their good time, or deny your request entirely. Sometimes you can spend extra time rewriting the component, but if it's sufficiently low-level (i.e., inside the OS), this isn't even possible.
So, if you're a programmer, most of your structures are build on unstable, shifting sand. Your tools are inadequate for the job, break at inopportune times. The pieces of your structure fail in unpredictable ways. You are forced to do extensive testing, because engineering calculations just aren't possible when all your materials have unknown defects.
And the jobs are usually so complex that exhaustive testing would take centuries on the fastest machines. So you deliver something that you know will fail. But you shrug, because you understand that it's what the customer wanted. Otherwise, they wouldn't be buying computers that were built to hide the details from the programmers. Everyone understands that if a foundation is unstable, the whole structure is unstable. So if they are buying unstable computers, they must want the software that runs on it to be unstable, too. Right?
I've done a few jobs where I could do a proper engineering job, because I had precise specs (i.e., access to the code) at all levels below my code. This is possible on a few real-time systems. But I don't even dream about it with most jobs, because I know up front that I'm not permitted to see the inner details of some of the things that I'm required to use. I know very well what the end result will be. But it's what the client wants.
Heh. Maybe they were counting Al Gore as the other president, since he was the one who got the most votes in 2000. ;-)
Too bad he was such a lazy bum, and delegated the entire job to Dubya.
Oh, well; they say that one part of being a good manager is knowing how to delegate authority and responsibility to others. At least Al doesn't have to take the blame for Iraq.
What will happen when countries like Tuvalu (.tv) reach technological savyness and find that their entire TLD has been used up by TV networks, ...
.tv domain has obvious commercial value.
Well, by then global warming will have put all but a few square meters of Tuvalu underwater.
Seriously; here's a story about it from back in 2001. Tuvalu's government is asking other countries to accept the Tuvaluans as citizens, without much luck so far.
I found this and other reports on the topic fairly quickly by googling for "nation flooded rising ocean". I do remember reading about it when the story first came out, but it didn't get a lot of coverage in the MSM (mainstream media).
When their islands are finally abandoned, it's anyone's guess what will happen legally to their TLD. I'd guess that it'll be a "land grab" by various corporations, since the
That hadn't occurred to me (and I've never seen it mentioned anywhere that I recall). It did work, with a rather awkward hand position that's reminiscent of some guitar chords. Then another reply suggested a Preferences tweak that moved the key mappings around to something that's a lot easier to use.
I'll do a bit more experimenting. It sure takes a while to stumble onto all these tricks.
Hey, thanks! That really improved the behavior. Now the F keys function as function keys.
I must say that the description of that checkbox in the Preferences window leaves a bit to be desired. There's no way I'd ever guess what it does after reading the description.
Lots of eye candy, but the Mac is rather deficient in the way it documents all this stuff. The main way to learn seems to be to ask others, until you find someone who has stumbled onto an answer. Much like other computer systems that I could name, of course.
I used to think that raw processor clock rate was everything, but I've seen the Macs run circles around Intel machines with half-again as much clock rate.
That's a standard misconception that marketers routinely take advantage of. Clock rate comparisons are only meaningful if the hardware (cpu, bus, etc) are very closely related. It takes more cycles to do something on an Intel-type cpu than on a PPC (but fewer cycles than on a RISC cpu). A wider bus means fewer cycles are needed. This is neither good nor bad; it's just different design.
But comparing clock rates is useful as a marketing-fog trick when selling to people who don't understand what clock rates mean.
Heh. How many computer users these days even know what the term "operator" used to signify? It's true that there are still lots of people who work with mainframes, but we probably don't have too many of them hanging out here. For most people, expanding BOFH still leaves them baffled, as they have no referent for the second word.
Actually, we've read a lot of comments here and elsewhere about how one of the signs of a growing success of unix/linux/*x in The Market is that Usenix is more and more populated by people wearing suits. The implication of this is well-understood.
Any real computer geek would understand that what the "top dogs" carry shouldn't mean a whole lot to anyone else. The idea that there's one best computer is as silly as the idea that there's one best car or ice-cream flavor.
It's possible that there is a worst computer, but it's not possible that there's a best computer. That depends greatly on what you want it to do. To some people I know, their Blackberry is the best computer they've ever used, and for their uses, they're right.
Control+F2 to give the menu bar the input focus,
Doesn't work at all on my Mac Powerbook. It does the same thing as F2 alone - The screen brightness control appears in the center of the screen, and repeated clicks on F2 (with or without the CTRL key) brightens the screen one notch.
So is there some Preference somewhere that'll make CTRL-F2 give the menu bar? It sure would be useful. I find that moving to the top bar and then back is a major time waster. And it's made even worse by all the times that I'm wrong about what window has the focus. There's little visual difference on the screen, just a slightly darker title bar. It takes extra time to check the name at the upper left, and all too often the app has several windows open anyway, so it's still wrong.
There are nice things about the Mac, but after using one a lot for a couple of years, I still find it frustratingly slow compared to unix/linux with X Windows (after 5 or 10 minutes of configging on a new machine).
I'd half agree with you. The Internet is sorta like a giant public library. But the overlap with the brick-and-mortar libraries isn't all that complete yet, so we need both of them.
... ;-)
As for "relying on the internet as a valid source of information", I'm reminded of a good comment I read some years ago:
The libraries of the world are the repository of all the world's wisdon - and most of its nonsense.
This applies equally to the Internet. In both cases, if you trust them to deliver only wisdom, you're a fool. You have to learn to make your own judgements. If you can do that, both are good tools. And with the advent of search sites, finding obscure information is slowly becoming easier.
Actually, the Internet isn't a material threat to libraries yet. The business that's more threatened is the traditional news media. Nearly all of their nonsense and heir occasional valid information is now available online. If you can find it. And you can easily find stuff that's not provided by your local media, so it's now a lot more difficult for the powerful people to pull the wool over our eyes.
Of course, what really caused the Internet explosion was porn
Um? I'm pretty sure every filesystem I've used in the past decade or so has been at least case-preserving. Actual case sensitivity is more down to taste.
... One lesson I've learned is to never, ever do an rsync between a Mac and any unix or linux system. That produced several major disasters, until we learned about the case insensitivity in the default OSX file system. If a directory has files "foobar" and "FooBar", on OSX you will get only one file, and there's a 50% chance that its contents will be the wrong one. Then, when you rsync again, there's another 50% chance that the OSX file will overwrite the unix/linux file with the wrong contents, giving you two files that are now identical.
;-). But this is expected; what's disappointing is that nobody seems to know of a solution to either problem.
;-)
...
Um
The other rsync disaster is that, if any directory contains 8-bit chars in their names, they will be unreadable gibberish on OSX. Then a second rsync will copy the gibberish filename back to the unix/linux system.
Yes, I've asked about solutions to these in several Mac-related fora. The repliest mostly call me an idiot for using such file names (though I'd usually mentioned that I didn't personally name the files in question
Yes, OSX does have a unix-like filesystem. Some Mac apps don't work right on it. And we can't demand that customers use it.
And I do like the idea of having a filesystem that uses UTF-8 for the encoding. But I can't tell all our linux customers that they have to have that. (And I can't tell the MS users anything.
If anyone knows of successful solutions to these problems (that don't require changing customers' file systems), please metion them in a reply. I'd expect that eventually rsync will get options naming the char incodings, but this doesn't seem to exist at present. The caseless filename matching is just a disaster all around.
If you're new to Macs, don't even think of using rsync
Agreed, there is balkanization in copy/paste methods in KDE. But OSX is not entirely consistent either, if you use any X apps.
;-)
I've been using OSX for a couple of years now, and I'm getting more baffled about the claims that it's consistent. In the case of cutting, you can see one of the real problems by opening up a few browser windows, doing a bit of browsing, and then try to edit or copy a URL.
One major problem is selecting what you want selected. A single click sometimes selects the entire URL, sometimes selects nothing (but inserts the cursor), and sometimes selects some "field" that seems to be different at different times. If the click selected the entire URL, you have to slowly click until that goes away and one of the other behaviors happens. Then you can drag the pointer to get a selection. Sometimes. Sometimes dragging the pointer does nothing, and you have to start again.
This behavior happens with all the browsers. It's not just the browsers, though, because there are lots of little popup windows in other apps whose input fields show the same unpredictable selection behavior.
I first thought this was some bug on my machine. But I've used a few other machines, and watched others using them, and it seems to be a problem for everyone. At least for everyone who actually uses copy-and-paste; some users never seem to do this.
Not that I'm especially criticising OSX software, of course; you can produce similarly frustrating selection behavior on MS Windows very easily. I've seen in with X Windows, but there it seems an occasional bug rather than normal behavior. A bigger problem there is all the text that can't be selected at all. This happens on all the windowing systems, of course. For example, where can you select text from a title bar? But statistically it's worse on X Windows, with a lot of ordinary-looking text in windows that can't be selected.
Overall, I've found X Windows the most predictable, in the sense that when I learn some new trick, I quickly push it into my subconscious, because after a dozen times, I can use it without thought. MS Windows is the least predictable, with lots of picky details slightly different between apps so that I can't ever quite stop thinking and frequently backtracking when I clicked something wrong. It's not just me, because when I watch experienced MS Windows users, they stumble the same way, and you hear a lot of quite cursing as they try to get it to work right. OSX is somewhere in between; often it "just works", but sometimes it gets really random.
Now let's see if this gets a "flamebait" rating, because it's not sufficiently praising of either the Mac or linux.
Now, linux-heads love choice and more power to them for that. BUT such up-front confusion with linux is not the way to win over the general public.
.... This does not give an advantage to the small neighborhood street-corner store; people go to supermarkets precisely because of the variety. No matter what your family likes, you'll find it there.
;-). So instead I'll pick what in all other areas would be a great advantage, the variety you offer, and I'll criticize you for that. I'm smart enough to know that if you're stupid enough to believe me and offer a system without choices, I'll just tell you that I don't like the single choice you offered me, because it's not 101% identical with my old system that I'm used to."
Oh, nonsense. There are a zillion examples (;-) of similar choices presented to the public, and you rarely if ever hear anyone but a few marketers complaining.
In most of the Western world, especially the US, if you want to but an auto, you have at least a hundred choices, many of them very similar. People don't complain about this; they discuss the differences between models at the drop of a hat.
It's common to observe that supermarkets have an insane variety of many products, frm bread and cereals to apples to soft drinks to
Similarly, people overwhelmingly go to the large clothing stores in malls, because that's where you'll find a huge choice in what chothes you can buy. And they'll let you try things on before you buy.
Imagine a restaurant with just one fixed meal. Not even McDonalds does that.
There's very little real evidence anywhere that people, even stupid people, are bothered by a variety of choices. They're only bothered if they don't have a way to choose. (This is a problem for ethnic restaurants with words that customers don't recognize. That's why they'll usually provide explanations.)
We do hear this sort of claim a lot in the computer biz. But I think it's totally bogus. It's a way of saying "I don't want to buy your system, and I'm not going to admit why (because it'd make me sound really stupid
You can't win such an argument no matter what you do.
But you're a lot more likely to win if you can offer people what they really want. And other consumer products show clearly that there is never a single version that everyone wants. You have to offer choices. And you have to make it easy for people to browse a bit and pick one that they'll like. Just like they do in clothing stores and food stores.
we don't care.
Heh. Actually, you should have done a topical version of the very old joke:
Q: Was the Sept 11 attack due to the American public's ignorance of the history of its government's actions in other parts of the world, or was it due to the public's apathetic response to reports on the topic?
A: We don't know, and we don't care.
During the last election, it became clear that this wasn't just a joke, when several surveys reported that around 80% of the Americans who voted for Bush were unable to give correct answers to any questions about Bush's policies. About 25% of the Kerry voters were equally ignorant. And this ignorance was mostly because they couldn't be bothered to read any of the easily available information on the topic; they only got their election info from television. So ignorance and apathy were indeed the order of the day.
What I've found effective is to reply to a Word doc by pointing out that it is encoded via a proprietary, patented scheme, and here in the US, the DMCA says that I can be fined $500,000 and sent to a federal prison for 5 years for decoding it with unauthorized software. Granted, nobody has yet been fined like this for decoding a Word doc, but I don't want to be the test case.
This always gets me a copy in a format that I can legally read.
(Actually, it's more likely that the courts would toss this part of the law. After all, if you sent me the doc, presumably you want me to read it. But again, I don't want to be the test case. I'd like to save my money for something more valuable.)
Look a few miles north, in Hanover. That's where you'll find Dartmouth College. Lebanon is effectively a suburb of Hanover, inasmuch as "suburb" has meaning in such a rural area. But this isn't some remote redneck area. It's a rich-private-school-in-the-country area.
Well, IBM has done just fine with this business model. ;-)
Of course, IBM has historically made a lot of money selling hardware. They give away a lot of software as a way of encouraging people to buy more hardware.
But they have also sold software support contracts. Details are hard to come by, but I've read a number of claims that IBM makes a lot more money from support than from sales. This is rarely broken down between hardware support and software support, so it's hard to know just how profitable their software support is. But they clearly consider "support" a profit center.
This has been their approach with AIX all along, and it's fairly clear that this is why they're getting behind linux. They've listened to the FUD saying that businesses won't use linux because they demand supported software. IBM's sales people look at this and think "Hey, we don't have to pay the programmers, we just need to train the support staff, and we can rake in the money."
This isn't a gamble for them, because they've been doing this sort of thing for decades. They know how to make it profitable. Especially when they don't have to pay the programmers.