Who cares? If an open source project gets taken to court, it's dead, unless it's something big like Linux. Once a patent is *issued*, you can commit extortion with it, or keep open source competitors out of your market, regardless of whether it's *valid*. Look at the whole OpenType font hinting thing - the policy was for a long time (may still be) that we shouldn't use hinting in fonts *just in case* Apple gets upset about it. Even though Apple has never *said* they would get upset about it.
Depending on litigation to create fairness is completely hopeless.
It's just as bogus to rewrite history to conform to your conspiracy theories as it is to rewrite it to conform to someone's miracle theories.
In fact, all we have are the texts and the lineage, which is over 2000 years old at this point. I think it would be very difficult to determine who, if anyone, still holds a pure lineage from 2000 years ago.
So the relevence that these books have for us now is whatever we are able to make of them - disputing the events described in them in the sense that you are trying to do is essentially nonsensical. At this juncture, it doesn't *matter* what happened 2000 years ago, except in the sense that that was the beginning of a chain of events the ongoing culmination of which we are now observing.
If you think there might be something of value in the books, you can try to put it into practice and see if it works; if it does, then you've found something true. If it doesn't, you probably missed the mark. And if you enjoy speculating about history, then you can validly take it that way, but remember what you're doing: speculating.
Actually, the Senhedron (the ruling council) was, rather famously, against him. Thus the whole thing with the pieces of gold, and the arrest and crucifixion.
Give me a break. A document is just data. You can use a program to render it as printable output. The output is just as much covered by the GPL as an image on your screen of a GPL'd program that you're running. The data in the file is just as much covered by the GPL as some source code that you ran through the GNU C compiler is. It's just like saying "you generated this output using groff, so the source from which the output was generated is now covered by the GPL." It's completely ridiculous.
I don't know who came up with this insane idea, but it sounds like FUD from someone whose profit margins are threatened by Open Source software.
If that's the case, you might want to at least have his son be a local boy, literally part of your tribe.
Er, Jesus was a Jew, not a Christian, and the Jews do not generally consider him to have been a legitimate Messiah. Furthermore, the Church, up until Pope John Paul II, held to the principle of supercession, meaning, that the Christian church believed that Christianity superceded Judiasm, which led to the persecution and murder of Jews as heretics, arguably culminating in the Nazi imprisonment and slaughter of the majority of all Jews living in Europe at the time. So this particular argument is completely bogus - having the Son of God (who refused that name, by the way) be a member of the local tribe was a total disaster for his tribe.
Oh, also, the idea of the Jews as the chosen people came much earlier, I think (not a biblical scholar - sorry) when Moses received the ten commandments. The idea isn't that they're chosen in the sense that they get all the goodies, but rather that the burden of bringing God's law to the rest of the world has been placed upon them. History has shown that being the chosen people isn't always a happy thing.
The trouble with stuff like this is that you need it to be usable on an iPod or there's no point in it. Speex doesn't claim to be available on the iPod. I'm not kidding, by the way - half the people who listen to the stuff I produce do so on an iPod, so it has to be MP3 or AAC. I use MP3 because AAC only works on iPod. Sigh.
I use MP3s for audio delivery, just not for archival storage. Dunno where I said in there that you shouldn't use MP3s for anything; just that if you want something that's archival, MP3s aren't the way to go.
You can't edit mp3s. Well, you can, but the results aren't good. You have to decompress the mp3, do your editing, and recompress, and chances are that you're going to get artifacts when you do this. So if you ever intend to use an audio file for any kind of editing, you really want it stored losslessly.
At Diamond Mountain University, we typically record all our classes, at 16bit 44khz mono, which consumes a substantial amount of disk space. When people record directly to MP3 it's a huge hassle because I can't take the audio and do dynamic range compression on it without creating artifacts, which means that you can't listen to it on your car stereo unless you have a luxo-mobile with really good sound baffling.
Right now I store all this stuff as AIFF files, but the idea of converting them to FLAC files is definitely attractive.
The nice thing about an open source codec is that you can keep a copy of the source code alongside the audio files. So at least in theory the codec will last as long as the audio does. For really long-term archiving I'd be tempted to put a minimal implementation of the decompressor at the beginning of each compressed file...
Avail yourself of the online resources.
on
Modern Mac Development?
·
· Score: 4, Informative
My three main bookmarks for Cocoa help right now are:
The main challenges with Cocoa are learning how to navigate project builder so as to connect things up, and figuring out all the undocumented interdependencies between Cocoa objects.
One of the things that really screwed me up for a long time was not realizing that the UI objects aren't re-entrant, so if you call a UI object from a thread other than the main thread, you're likely to get random crashes.
The biggest thing for me though was figuring out project builder - figuring out how outlets and actions work. This is particularly a problem because there's no file you can edit to do outlets and connections - you have to do it with the GUI, and it's extremely counterintuitive. Once you figure it out it's easy, though. So it's good to just type your way through a couple of tutorials just to get the hang of it.
If you want to look at some example source code, I have something up on sourceforge that's (a) not trivial, but (b) not heinously complex, so it might be worth looking at. http://www.sourceforge.net/projects/gofer
Prior art research is complicated by the fact that a lot of patents are for things that are "obvious", and so nobody's ever published anything about them because what would be the point? People who are actually doing good work in the field don't have time to exhaustively list all the possible uses for every tool they invent, even if the uses are obvious, so the result is that some obvious, workmanlike extension of an existing piece of technology suddenly becomes patentable. And because nobody published anything about it, there's no prior art, so the patent would be granted even if it were researched properly.
The real problem is that the standard for obviousness is way, way too low for software. This makes a certain amount of sense coming from the physical world, because incremental improvements to existing processes are actually at least in theory expensive to develop. But when you can come up with an idea in two minutes and have a working implementation the next day, I'm sorry, but that's an obvious idea. If software patents are ever justified, they should be for really subtle thinking, not straightforward extensions of existing work. To me, the test of obviousness is, if you present the same problem to a skilled practitioner, how long does it take them to come up with a solution? If they can tell you how to solve the problem two minutes later, or even an hour later, the process is obvious and therefore not patentable.
The way things work right now, what's really being patented is the question, not the answer, and while unfortunately there's a good deal of precedent at this point for this way of thinking about software patents, I think this is an extremely pernicious standard that really benefits nobody.
It's unfortunate that IBM has taken this halfway position, but I think they've done so somewhat strategically - it's not at all clear to me that their stated position is their actual position. It may be that they're trying to start a process that will blow up the existing patent system for software, but they can't say that because it would weaken their defense in court while they're waiting for the result they've targeted. It also may be that they Just Don't Get it - I don't know.
Violating terms of service is a contract violation. So no, it's not theft. It's lying.
BTW, in communist China, copyright law, such as it is, is only enforced when it's convenient, which means that by and large any money you make producing music is made by means other than sales of music. So your "comrade" remark is backwards - you are the one who's advocating the communist way; the person that you're criticizing is taking a more capitalist position.
Generally speaking, if there is a way to do something in the language standard, it should be done that way, and not in some other way. If there is no way to do the thing you want in the standard, then I guess you have to do it a different way. I would tend to consider the FIPS standard to be the most current, but I'm no expert. The main place I've run into problems with compatibility is in the way that UNIQUE is specified, and also differences between VARCHAR and special types that MySQL provides that, frankly, aren't very useful and have equivalents in standard SQL.
One of the things that's a real problem for me with MySQL is the places where MySQL doesn't follow the SQL language standard. This means that MySQL scripts typically only run against MySQL. This is probably just ignorance on my part - perhaps they fixed this long ago, and people are just coding to the old standard - but does anybody know anything about this? I wasn't able to find anything about it in the press release.
The bottom line is this: can the end user get the functionality they need easily? The answer is no. I see a trend toward it getting better, but the user experience on Linux for a regular person, who is not a geek, is still pretty awful. If they don't have someone like me to get their system running, they are SOL.
What Marcus doesn't mention, that I think is the reason why things aren't quite as gloomy as he makes them out to be, is that Windows isn't designed for users. It's designed for corporations. It *sucks* for users. Windows has a really fantastic security model, if what you are trying to secure is a corporate infrastructure. I'm not sure how well it works, but it's pretty cool from that perspective.
Unfortunately, from the perspective of someone who just bought a laptop with WinXP Pro installed on it, it's completely wrong. Its purpose in life is to do something that is inimical to the well-being of the end-user - it's there to prevent the misappropriation of your data, and of Microsoft's licenses. But misappropriation of data includes things like backups and restores.
And unfortunately, MS WIndows is extremely vulnerable to scenarios that require complete system rebuilds. For an end user who is not a trained MCSE or knowledgable windows geek, getting back to where they were after an event like this is virtually impossible, mostly because of the cumbersome Windows security model, which is designed to protect someone other than them.
So there is hope for Linux, particularly in the land of the end user. The way the hope will surface is that one of the way-too-many distributions will finally figure out how to deliver a good experience to an end-user who is not a geek. And then, one by one, all the end users who are the target market for that product will install it over their broken windows installs.
Microsoft will still probably own the corporate market. Let them have it - who cares? But the platform I'm talking about will take over the end user space.
In case anybody's wondering, this platform doesn't yet exist. Everybody gets it wrong. Ubuntu is supposed to be the distribution for the people. I like it. It's Debian with an end-user-friendly face.
But a fresh Ubuntu install can't find my windows or rendezvous printers. Why, because there's no support for browsing printers? No, because they've turned it off because it's supposed to be a security hole. Sigh.
If you want your distro to be the one that finally breaks the barrier, you have to put the end user experience in the forefront, and make your decisions based on that. You're worried about the security of the browser? Don't get rid of the browser. Do something about the security issue, and keep the browser.
I think it depends on what you're buying. For example, I can buy a new laptop computer that's as fast and powerful as my old desktop computer, and that definitely draws less power. Or I can buy a whizzy new desktop computer that draws more power than the old one.
Just as an example of this, consider a Soekris 486 box with a 2.5" 7200 RPM drive. My main server for fugue.com used to be a 486 that ran slower than the Soekris 486, had a huge power supply, and drew probably at least one order of magnitude more power.
I am hoping that as the mad dash to find the tail end of Moore's law starts to chill out, it'll be more possible to purchase competent but not blazingly-fast machines that do the job and don't draw a lot of juice. I think the boxes Soekris makes are a good example of this, and I've seen other examples as well.
Right now I'm in the early phase of experimenting with using a powerbook with a broken screen as a server; it'll be interesting to see how that plays out.
On the topic of flourescents, there's a really wide range of products out there. You can get full-spectrum compact flourescents, and they do save you money - the bulk costs more, but it uses sufficiently less energy to produce the same light that it's still cheaper. Full-spectrum CFs are not the same as incandescents, BTW - they try to be close to the same spectrum you'd get from direct sunlight, which is a little less warm than incandescent light.
If you have flourescent tube fixtures, you can also get full spectrum lights for these. They're more expensive than the standard sickly-green tubes, but quite available. There are a wide variety of ballasts available - you don't have to use 60Hz ballasts. Electronic ballasts typically oscillate at 10KHz, which doesn't heterodyne with your TV tube, and doesn't cause eyestrain. I think they are also more energy-efficient, but I'm not positive anymore - it's been a while since I did all the research.
Not an easy thing to do, and not everyone can do it, but one thing that really surprises me about Tucson, and this is true for many cities, I think, is how incredibly inappropriate the building materials are that people use. Most of the houses in Tucson are made of wood, with minimal insulation and huge black roofs to collect maximal solar energy during the long summers. Many houses are cooled with swamp coolers, which typically go through 100 gallons of water a day, even though this is a desert climate. Many people have lawns, which of course have to be watered daily throughout most of the year.
When we moved to Tucson, we knew we weren't going to buy a wood house, because wood is simply a lousy material for this climate. Back in the days before massive cookie-cutter developments, houses in Tucson were made of 12-18" thick adobe or 6-8" burnt adobe/slump block masonry. This material is good for the climate, because it has a lot of thermal mass. This means that if you get the thickness right, the outside temperature at night will be coming through the walls during the day, and the outside temperature during the day will be coming through the walls at night. So if you open the windows at night, and close them during the day, you can be fairly comfortable even in the heat of summer, without using any heating at all, and in the winter you might want to put on a sweater, but you'll basically be warm enough, again without any heating.
However, it turns out that buying a house built this way nowadays is quite expensive, unless you buy an older house, and older houses have the problem that most of them have wood floors, meaning that you're very vulnerable to termites.
We were very fortunate to find a builder who is working on renewable-energy housing right in downtown Tucson - our house is made of concrete masonry, but is insulated on the outside, so rather than depending on the diurnal cycle, it is isolated from the outside temperature swings. The thermal mass of the all-masonry construction and the 10" thick concrete floor mean that once you get it to a certain temperature, it tends to want to stay there. So it's quite cheap to keep cool, even in the dead of summer.
On top of this, there's a solar hot water heater and 1500WDC solar panel on the roof, so that although we still draw energy from the grid, we draw a lot less of it, and our air conditioning can run mostly off the solar panel during the day, when energy is in most demand. To back up the solar hot water heater we have an electric instant-hot-water heater from Seisco that works really well - the hot water out of the tap has a really consistent temperature with no pulsing.
That's really our big way of saving energy. We'd like to have a hybrid car, because unfortunately we aren't quite able to go cold turkey on automotive transportation, but for now we're making do with our Honda Civic, which gets pretty good milage.
As for computers, unfortunately I think the best solution is to always buy newer ones, but it costs energy to make them, so this isn't perfect. Newer computers do seem to use less energy as long as you're not pushing them to extreme clock speeds. Probably using just a laptop would help, but for work it's really handy to have a faster disk and processor.
Slashdot Mirror
Who cares? If an open source project gets taken to court, it's dead, unless it's something big like Linux. Once a patent is *issued*, you can commit extortion with it, or keep open source competitors out of your market, regardless of whether it's *valid*. Look at the whole OpenType font hinting thing - the policy was for a long time (may still be) that we shouldn't use hinting in fonts *just in case* Apple gets upset about it. Even though Apple has never *said* they would get upset about it.
Depending on litigation to create fairness is completely hopeless.
It's just as bogus to rewrite history to conform to your conspiracy theories as it is to rewrite it to conform to someone's miracle theories.
In fact, all we have are the texts and the lineage, which is over 2000 years old at this point. I think it would be very difficult to determine who, if anyone, still holds a pure lineage from 2000 years ago.
So the relevence that these books have for us now is whatever we are able to make of them - disputing the events described in them in the sense that you are trying to do is essentially nonsensical. At this juncture, it doesn't *matter* what happened 2000 years ago, except in the sense that that was the beginning of a chain of events the ongoing culmination of which we are now observing.
If you think there might be something of value in the books, you can try to put it into practice and see if it works; if it does, then you've found something true. If it doesn't, you probably missed the mark. And if you enjoy speculating about history, then you can validly take it that way, but remember what you're doing: speculating.
Actually, the Senhedron (the ruling council) was, rather famously, against him. Thus the whole thing with the pieces of gold, and the arrest and crucifixion.
Give me a break. A document is just data. You can use a program to render it as printable output. The output is just as much covered by the GPL as an image on your screen of a GPL'd program that you're running. The data in the file is just as much covered by the GPL as some source code that you ran through the GNU C compiler is. It's just like saying "you generated this output using groff, so the source from which the output was generated is now covered by the GPL." It's completely ridiculous.
I don't know who came up with this insane idea, but it sounds like FUD from someone whose profit margins are threatened by Open Source software.
Longhorn is actually a picture in Mac OS X's closet...
If that's the case, you might want to at least have his son be a local boy, literally part of your tribe.
Er, Jesus was a Jew, not a Christian, and the Jews do not generally consider him to have been a legitimate Messiah. Furthermore, the Church, up until Pope John Paul II, held to the principle of supercession, meaning, that the Christian church believed that Christianity superceded Judiasm, which led to the persecution and murder of Jews as heretics, arguably culminating in the Nazi imprisonment and slaughter of the majority of all Jews living in Europe at the time. So this particular argument is completely bogus - having the Son of God (who refused that name, by the way) be a member of the local tribe was a total disaster for his tribe.
Oh, also, the idea of the Jews as the chosen people came much earlier, I think (not a biblical scholar - sorry) when Moses received the ten commandments. The idea isn't that they're chosen in the sense that they get all the goodies, but rather that the burden of bringing God's law to the rest of the world has been placed upon them. History has shown that being the chosen people isn't always a happy thing.
The trouble with stuff like this is that you need it to be usable on an iPod or there's no point in it. Speex doesn't claim to be available on the iPod. I'm not kidding, by the way - half the people who listen to the stuff I produce do so on an iPod, so it has to be MP3 or AAC. I use MP3 because AAC only works on iPod. Sigh.
I use MP3s for audio delivery, just not for archival storage. Dunno where I said in there that you shouldn't use MP3s for anything; just that if you want something that's archival, MP3s aren't the way to go.
You can't edit mp3s. Well, you can, but the results aren't good. You have to decompress the mp3, do your editing, and recompress, and chances are that you're going to get artifacts when you do this. So if you ever intend to use an audio file for any kind of editing, you really want it stored losslessly.
At Diamond Mountain University, we typically record all our classes, at 16bit 44khz mono, which consumes a substantial amount of disk space. When people record directly to MP3 it's a huge hassle because I can't take the audio and do dynamic range compression on it without creating artifacts, which means that you can't listen to it on your car stereo unless you have a luxo-mobile with really good sound baffling.
Right now I store all this stuff as AIFF files, but the idea of converting them to FLAC files is definitely attractive.
The nice thing about an open source codec is that you can keep a copy of the source code alongside the audio files. So at least in theory the codec will last as long as the audio does. For really long-term archiving I'd be tempted to put a minimal implementation of the decompressor at the beginning of each compressed file...
My three main bookmarks for Cocoa help right now are:
o rum/. htm
http://rentzsch.com/
http://www.idevapps.com/f
http://www.zathras.de/angelweb/x2004-12-05b
The main challenges with Cocoa are learning how to navigate project builder so as to connect things up, and figuring out all the undocumented interdependencies between Cocoa objects.
One of the things that really screwed me up for a long time was not realizing that the UI objects aren't re-entrant, so if you call a UI object from a thread other than the main thread, you're likely to get random crashes.
The biggest thing for me though was figuring out project builder - figuring out how outlets and actions work. This is particularly a problem because there's no file you can edit to do outlets and connections - you have to do it with the GUI, and it's extremely counterintuitive. Once you figure it out it's easy, though. So it's good to just type your way through a couple of tutorials just to get the hang of it.
If you want to look at some example source code, I have something up on sourceforge that's (a) not trivial, but (b) not heinously complex, so it might be worth looking at. http://www.sourceforge.net/projects/gofer
Right, you're effectively blowing up the patent system if you reduce their lives to less than the amount of time that it takes to get one.
Prior art research is complicated by the fact that a lot of patents are for things that are "obvious", and so nobody's ever published anything about them because what would be the point? People who are actually doing good work in the field don't have time to exhaustively list all the possible uses for every tool they invent, even if the uses are obvious, so the result is that some obvious, workmanlike extension of an existing piece of technology suddenly becomes patentable. And because nobody published anything about it, there's no prior art, so the patent would be granted even if it were researched properly.
The real problem is that the standard for obviousness is way, way too low for software. This makes a certain amount of sense coming from the physical world, because incremental improvements to existing processes are actually at least in theory expensive to develop. But when you can come up with an idea in two minutes and have a working implementation the next day, I'm sorry, but that's an obvious idea. If software patents are ever justified, they should be for really subtle thinking, not straightforward extensions of existing work. To me, the test of obviousness is, if you present the same problem to a skilled practitioner, how long does it take them to come up with a solution? If they can tell you how to solve the problem two minutes later, or even an hour later, the process is obvious and therefore not patentable.
The way things work right now, what's really being patented is the question, not the answer, and while unfortunately there's a good deal of precedent at this point for this way of thinking about software patents, I think this is an extremely pernicious standard that really benefits nobody.
It's unfortunate that IBM has taken this halfway position, but I think they've done so somewhat strategically - it's not at all clear to me that their stated position is their actual position. It may be that they're trying to start a process that will blow up the existing patent system for software, but they can't say that because it would weaken their defense in court while they're waiting for the result they've targeted. It also may be that they Just Don't Get it - I don't know.
Violating terms of service is a contract violation. So no, it's not theft. It's lying.
BTW, in communist China, copyright law, such as it is, is only enforced when it's convenient, which means that by and large any money you make producing music is made by means other than sales of music. So your "comrade" remark is backwards - you are the one who's advocating the communist way; the person that you're criticizing is taking a more capitalist position.
This is looking pretty workable to me - might make a good reason to get a Prius sooner rather than later.
Generally speaking, if there is a way to do something in the language standard, it should be done that way, and not in some other way. If there is no way to do the thing you want in the standard, then I guess you have to do it a different way. I would tend to consider the FIPS standard to be the most current, but I'm no expert. The main place I've run into problems with compatibility is in the way that UNIQUE is specified, and also differences between VARCHAR and special types that MySQL provides that, frankly, aren't very useful and have equivalents in standard SQL.
One of the things that's a real problem for me with MySQL is the places where MySQL doesn't follow the SQL language standard. This means that MySQL scripts typically only run against MySQL. This is probably just ignorance on my part - perhaps they fixed this long ago, and people are just coding to the old standard - but does anybody know anything about this? I wasn't able to find anything about it in the press release.
That's probably true, and I would consider it inappropriate to vote without reading all the books. Dunno why you feel the need to belabor this point.
I really recommend it. I haven't read any of the others, so my vote for Mr. Stross' book is a bit biased, but I definitely think it deserves a Hugo.
...work to get rid of them. Seriously. There is absolutely nothing you can do to protect yourself from software patents, other than this.
You certainly wouldn't hear me complaining if it turned out that way. But they have a ways to go yet.
The bottom line is this: can the end user get the functionality they need easily? The answer is no. I see a trend toward it getting better, but the user experience on Linux for a regular person, who is not a geek, is still pretty awful. If they don't have someone like me to get their system running, they are SOL.
What Marcus doesn't mention, that I think is the reason why things aren't quite as gloomy as he makes them out to be, is that Windows isn't designed for users. It's designed for corporations. It *sucks* for users. Windows has a really fantastic security model, if what you are trying to secure is a corporate infrastructure. I'm not sure how well it works, but it's pretty cool from that perspective.
Unfortunately, from the perspective of someone who just bought a laptop with WinXP Pro installed on it, it's completely wrong. Its purpose in life is to do something that is inimical to the well-being of the end-user - it's there to prevent the misappropriation of your data, and of Microsoft's licenses. But misappropriation of data includes things like backups and restores.
And unfortunately, MS WIndows is extremely vulnerable to scenarios that require complete system rebuilds. For an end user who is not a trained MCSE or knowledgable windows geek, getting back to where they were after an event like this is virtually impossible, mostly because of the cumbersome Windows security model, which is designed to protect someone other than them.
So there is hope for Linux, particularly in the land of the end user. The way the hope will surface is that one of the way-too-many distributions will finally figure out how to deliver a good experience to an end-user who is not a geek. And then, one by one, all the end users who are the target market for that product will install it over their broken windows installs.
Microsoft will still probably own the corporate market. Let them have it - who cares? But the platform I'm talking about will take over the end user space.
In case anybody's wondering, this platform doesn't yet exist. Everybody gets it wrong. Ubuntu is supposed to be the distribution for the people. I like it. It's Debian with an end-user-friendly face.
But a fresh Ubuntu install can't find my windows or rendezvous printers. Why, because there's no support for browsing printers? No, because they've turned it off because it's supposed to be a security hole. Sigh.
If you want your distro to be the one that finally breaks the barrier, you have to put the end user experience in the forefront, and make your decisions based on that. You're worried about the security of the browser? Don't get rid of the browser. Do something about the security issue, and keep the browser.
I think it depends on what you're buying. For example, I can buy a new laptop computer that's as fast and powerful as my old desktop computer, and that definitely draws less power. Or I can buy a whizzy new desktop computer that draws more power than the old one.
Just as an example of this, consider a Soekris 486 box with a 2.5" 7200 RPM drive. My main server for fugue.com used to be a 486 that ran slower than the Soekris 486, had a huge power supply, and drew probably at least one order of magnitude more power.
I am hoping that as the mad dash to find the tail end of Moore's law starts to chill out, it'll be more possible to purchase competent but not blazingly-fast machines that do the job and don't draw a lot of juice. I think the boxes Soekris makes are a good example of this, and I've seen other examples as well.
Right now I'm in the early phase of experimenting with using a powerbook with a broken screen as a server; it'll be interesting to see how that plays out.
On the topic of flourescents, there's a really wide range of products out there. You can get full-spectrum compact flourescents, and they do save you money - the bulk costs more, but it uses sufficiently less energy to produce the same light that it's still cheaper. Full-spectrum CFs are not the same as incandescents, BTW - they try to be close to the same spectrum you'd get from direct sunlight, which is a little less warm than incandescent light.
If you have flourescent tube fixtures, you can also get full spectrum lights for these. They're more expensive than the standard sickly-green tubes, but quite available. There are a wide variety of ballasts available - you don't have to use 60Hz ballasts. Electronic ballasts typically oscillate at 10KHz, which doesn't heterodyne with your TV tube, and doesn't cause eyestrain. I think they are also more energy-efficient, but I'm not positive anymore - it's been a while since I did all the research.
Not an easy thing to do, and not everyone can do it, but one thing that really surprises me about Tucson, and this is true for many cities, I think, is how incredibly inappropriate the building materials are that people use. Most of the houses in Tucson are made of wood, with minimal insulation and huge black roofs to collect maximal solar energy during the long summers. Many houses are cooled with swamp coolers, which typically go through 100 gallons of water a day, even though this is a desert climate. Many people have lawns, which of course have to be watered daily throughout most of the year.
When we moved to Tucson, we knew we weren't going to buy a wood house, because wood is simply a lousy material for this climate. Back in the days before massive cookie-cutter developments, houses in Tucson were made of 12-18" thick adobe or 6-8" burnt adobe/slump block masonry. This material is good for the climate, because it has a lot of thermal mass. This means that if you get the thickness right, the outside temperature at night will be coming through the walls during the day, and the outside temperature during the day will be coming through the walls at night. So if you open the windows at night, and close them during the day, you can be fairly comfortable even in the heat of summer, without using any heating at all, and in the winter you might want to put on a sweater, but you'll basically be warm enough, again without any heating.
However, it turns out that buying a house built this way nowadays is quite expensive, unless you buy an older house, and older houses have the problem that most of them have wood floors, meaning that you're very vulnerable to termites.
We were very fortunate to find a builder who is working on renewable-energy housing right in downtown Tucson - our house is made of concrete masonry, but is insulated on the outside, so rather than depending on the diurnal cycle, it is isolated from the outside temperature swings. The thermal mass of the all-masonry construction and the 10" thick concrete floor mean that once you get it to a certain temperature, it tends to want to stay there. So it's quite cheap to keep cool, even in the dead of summer.
On top of this, there's a solar hot water heater and 1500WDC solar panel on the roof, so that although we still draw energy from the grid, we draw a lot less of it, and our air conditioning can run mostly off the solar panel during the day, when energy is in most demand. To back up the solar hot water heater we have an electric instant-hot-water heater from Seisco that works really well - the hot water out of the tap has a really consistent temperature with no pulsing.
That's really our big way of saving energy. We'd like to have a hybrid car, because unfortunately we aren't quite able to go cold turkey on automotive transportation, but for now we're making do with our Honda Civic, which gets pretty good milage.
As for computers, unfortunately I think the best solution is to always buy newer ones, but it costs energy to make them, so this isn't perfect. Newer computers do seem to use less energy as long as you're not pushing them to extreme clock speeds. Probably using just a laptop would help, but for work it's really handy to have a faster disk and processor.