There were some points in there that were a little undefended, but I didn't see any that were wrong. All in all, I'd say he hit the nail on the head. He even pointed out that he intends to switch back to Linux, when it is ready. I think this article really lays out soemthing that seems to be lacking in many Linux circles: Pragmatism. Eveyone is talking about "linux on the desktop" on how it's this big goal right now, but they seem to be missing the point. It doesn't matter how stable or configurable something is, if you want it "on the desktop" Joe User has to be able to _use_ it. And it's not ready for that. Granted, I use Linux as my desktop OS at home and for my independant work, but I'm not "Joe User", and even that is likely to change soon. OSX is exactly what I want and need in a work machine. But even then, I'll still keep my Wintendo, since that is practically the only platform for decent games, with a few notable exceptions.
At my day job, I use Win2k, because it works easily and I can do my job with it. That's the very reason I'm taking the server farm to Linux, away from MS server products. With Linux there, it works easily, and I can do my job with it.
If the Linux community wants Linux to become a serious force in the desktop world, we are well on their way, but we would do well to heed the points that were brought up here. Especially about X, it really is a pretty clunky system for desktop work. Apple seems to have the right idea, IMHO.
Packages are a nightmare right now, and it seems to be a real sore subject with a lot of people. I read somewhere recently about a guy who wanted to remove sendmail and use a differnet mailer system, but couldn't get the package to install. The general response was "who cares that it didn't work? that system sucks anyway, just stick with sendmail". They totally missed the point, it doesn't matter that the other system sucked. What matters is that he wanted to use it but couldn't, because the package system is so clunky. On other OSes he would have simply installed it, played with it, then _decided for himself_ that is sucked, and then switched back. Probably in less time than he wasted with the RPMs. Apt is a step in the right direction, but it's still not there yet.
This is getting too long and I'm rambling. I'm stopping now. Have fun.
Seriously here. A buyer who is paying to get a custom piece of software made for them will be very very reluctant to let the rest of the world have that software for free once they have it. Especially if they have competitors. Especially if that software is mission critical at all.
This is a big assumption. I have done a fair amount of custom software work in the past. While I have not particularly "licensed" any of the code I produced, I did make the choice (mistake?) of leaving the resulting IP in the hands of of my clients. In more than one occasion, I got calls months later from someone I had never met who wanted to hire me to do more work on the program I had produced for the other client. When I ask these third parties how they got the program, the answer is almost invariably something like "Well, I met *so-and-so* at *such-a-place* and we got to talking about work. He said you made this great custom application for him that did exactly what he needed. It sounded like just the piece of software I was looking for! I asked if he would give me a copy..."
You see, you are assuming that the client is in the software business, which is most likely not true. They want to have a custom app made because they can't find one that does what they want it to do. They (often) don't care if someone else uses it, or even if they pay for it. To them, once they have the software and it is working for them, that is what they paid for. Whether or not someone else has it and uses it for free doesn't matter to them. They don't see it as an asset the same way they would see, say, a copy machine or a computer. It's no skin off their back if they give someone else a copy.
When I was getting started with Linux and I didn't have any high-speed Internet access, the discs in the books I got really got me on the road nicely. However, now that I have been using Linux for a couple years, I usually just tear 'em out and never even look at them.
PXE works quite well, check out the LTSP project and the K12LTSP project for examples of working PXE solutions. I use the K12LTSP distribution to run a lab of Linux diskless terminals in my public school district and they work quite well. I turn them on, and they go.
You obviously have quite a bit of experience in influencing legislation and policy in the US. In your opinion, what is the most effecive way for "John Q Public" to help influence public opinion and legislators?
Another reason DSL often doesn't work is because the fiber they talk about often is already running another kind of DSL (Same name, different tech, leave it to the telecom industry...)that is used only for muxing voice signals, and is limited to ~64k/sec speed. These DSL segments are also part of the reason some people on modems get such crappy speeds...
I could swear that I've seen a mochine in a Japanese trade mag that had a MD drive in it. Something like 230 MB capacity if I remember right, but it was some time ago.
Slashdot Mirror
There were some points in there that were a little undefended, but I didn't see any that were wrong. All in all, I'd say he hit the nail on the head. He even pointed out that he intends to switch back to Linux, when it is ready. I think this article really lays out soemthing that seems to be lacking in many Linux circles: Pragmatism. Eveyone is talking about "linux on the desktop" on how it's this big goal right now, but they seem to be missing the point. It doesn't matter how stable or configurable something is, if you want it "on the desktop" Joe User has to be able to _use_ it. And it's not ready for that. Granted, I use Linux as my desktop OS at home and for my independant work, but I'm not "Joe User", and even that is likely to change soon. OSX is exactly what I want and need in a work machine. But even then, I'll still keep my Wintendo, since that is practically the only platform for decent games, with a few notable exceptions.
At my day job, I use Win2k, because it works easily and I can do my job with it. That's the very reason I'm taking the server farm to Linux, away from MS server products. With Linux there, it works easily, and I can do my job with it.
If the Linux community wants Linux to become a serious force in the desktop world, we are well on their way, but we would do well to heed the points that were brought up here. Especially about X, it really is a pretty clunky system for desktop work. Apple seems to have the right idea, IMHO.
Packages are a nightmare right now, and it seems to be a real sore subject with a lot of people. I read somewhere recently about a guy who wanted to remove sendmail and use a differnet mailer system, but couldn't get the package to install. The general response was "who cares that it didn't work? that system sucks anyway, just stick with sendmail". They totally missed the point, it doesn't matter that the other system sucked. What matters is that he wanted to use it but couldn't, because the package system is so clunky. On other OSes he would have simply installed it, played with it, then _decided for himself_ that is sucked, and then switched back. Probably in less time than he wasted with the RPMs. Apt is a step in the right direction, but it's still not there yet.
This is getting too long and I'm rambling. I'm stopping now. Have fun.
Seriously here. A buyer who is paying to get a custom piece of software made for them will be very very reluctant to let the rest of the world have that software for free once they have it. Especially if they have competitors. Especially if that software is mission critical at all.
This is a big assumption. I have done a fair amount of custom software work in the past. While I have not particularly "licensed" any of the code I produced, I did make the choice (mistake?) of leaving the resulting IP in the hands of of my clients. In more than one occasion, I got calls months later from someone I had never met who wanted to hire me to do more work on the program I had produced for the other client. When I ask these third parties how they got the program, the answer is almost invariably something like "Well, I met *so-and-so* at *such-a-place* and we got to talking about work. He said you made this great custom application for him that did exactly what he needed. It sounded like just the piece of software I was looking for! I asked if he would give me a copy..."
You see, you are assuming that the client is in the software business, which is most likely not true. They want to have a custom app made because they can't find one that does what they want it to do. They (often) don't care if someone else uses it, or even if they pay for it. To them, once they have the software and it is working for them, that is what they paid for. Whether or not someone else has it and uses it for free doesn't matter to them. They don't see it as an asset the same way they would see, say, a copy machine or a computer. It's no skin off their back if they give someone else a copy.
But not much to say, other than that they need some video of one of those things spontaneously combusting a table, that would be cool...
When I was getting started with Linux and I didn't have any high-speed Internet access, the discs in the books I got really got me on the road nicely. However, now that I have been using Linux for a couple years, I usually just tear 'em out and never even look at them.
PXE works quite well, check out the LTSP project and the K12LTSP project for examples of working PXE solutions. I use the K12LTSP distribution to run a lab of Linux diskless terminals in my public school district and they work quite well. I turn them on, and they go.
You obviously have quite a bit of experience in influencing legislation and policy in the US. In your opinion, what is the most effecive way for "John Q Public" to help influence public opinion and legislators?
Another reason DSL often doesn't work is because the fiber they talk about often is already running another kind of DSL (Same name, different tech, leave it to the telecom industry...)that is used only for muxing voice signals, and is limited to ~64k/sec speed. These DSL segments are also part of the reason some people on modems get such crappy speeds...
I could swear that I've seen a mochine in a Japanese trade mag that had a MD drive in it. Something like 230 MB capacity if I remember right, but it was some time ago.
-Deque-
And yes, I'm sure it wasn't MO.