The lowest thing I booted Linux on was a Toshiba laptop 386sx, 16 Mhz with only 3 Mb of RAM. I had to create a special boot diskette to be able to partition it first and add 5 Mb swap, before I could install Slackware on it.
The lowest thing I ran X on was another 386sx with 16 Mb RAM and a Hercules card (yeah , b/w), 80 Meg HDD. I used it as an Xserver for my large system.
The only thing with these systems is that even on the command line they are slow. I suspect it is the kernels, because in 1993 I already had installed Linux on such systems, and they did not seem slow at that time.
I still can't believe that people have problems running Mozilla and OOo on systems that are more powerful than mine.
I run OOo 1.0 on the following systems :
200 MHz PPro, 128 Mb RAM
233 MHz PII, 96 Mb RAM
I use these to write documents which I must export to.doc format and to create slideshows for courses on Linux. Yes, startup time is a little long, but I am now used to it, but working with it is no problem.
Mozilla is a little bit slow in the tooth, and I would certainly not recommend it for slower systems than mine, but it performs fine for me.
This is something that I had already thought over in the past.
If you have an organisation that is large enough, you could pull of grabbing the codebase, modifying it and releasing the program without sources.
However, you will have do one of two things : saying that your program is compatible with the original, or saying that it is something totally new.
In the first case, you have locked yourself in, because if you claim compatibility, you must make this true, and you will need to follow that original organisation from which the codebase grew, and which is still developing. This means that the best you can do is just take their code everytime and provide new releases as fast as the original group.
If you say that it is something completely new, then you will have to make an effort in convincing people that it is better, but you cannot say that it is better than the original. If you really fork your codebase, than you will need a real large group of people with the same capacities as the original group from which the code forked.
Smaller projects can easily be taken over, but for large projects like the Linux kernel, I think it is not really practical, because of the investments needed. Investments in equally competent people, but also extra investments in marketing and so on.
You can base a product, a package, a distribution on it, but really forking Linux and maintaining a complete separate codebase ? It would be just an enormous waste of time and money.
As far as I am a computer geek, I also do like things simple. My oven only has two knobs, one for temperature, one for type of warming. The cooking plates each have one button (and I like cooking very much).
Take remote controls. I do not have a problem with them. For my father in law however, there should be a thing with only volume control, stations up and down and one button for remote powercontrol of his TV set.
Of course, it is also a question of philosophy. KISS is something that is not only good for design and maintenance, but also for usage. But what most people do not understand is that SIMPLE != EASY. It is much more difficult to come up with something simple and elegant, than it is to come up with something contrived. I think that simplicity is lost because of too short times-to-market. It results from the fact that designers do not have to time to look and search for several solutions to a problem, and pick the best. Instead, they use whatever solution they come up with.
I still have my Casio FX-8000G, though. My father bought it for me, for my 21st birthday, on the airport of Abu-Dhabi. It is 15 years old, it still serves me well, although maybe it could benefit from some cleaning. It has a binary and hex mode, and I think (but I am not sure) that you just put it in this mode, and that the A-F keys are usable without special shifting.
I have seen this in reality. On my previous job the systems were migrated to HP systems, which claimed 99.999%. One of the first things which broke was the redundant Fibre Channel controller. It took two days to fix it.
I do not think you have read 'Chthon' from Piers Anthony.
Btw. I haven't found any references to his works of the seventies yet, Chthon, Sos The Rope, Var The Stick...
Slashdot Mirror
The problem is that Intel succumbed for installing DRM, after AMD had first said to implement DRM in response to a request of MS.
Need one secondhand ? 25 EUR!
The lowest thing I booted Linux on was a Toshiba laptop 386sx, 16 Mhz with only 3 Mb of RAM. I had to create a special boot diskette to be able to partition it first and add 5 Mb swap, before I could install Slackware on it.
The lowest thing I ran X on was another 386sx with 16 Mb RAM and a Hercules card (yeah , b/w), 80 Meg HDD. I used it as an Xserver for my large system.
The only thing with these systems is that even on the command line they are slow. I suspect it is the kernels, because in 1993 I already had installed Linux on such systems, and they did not seem slow at that time.
I still can't believe that people have problems running Mozilla and OOo on systems that are more powerful than mine.
I run OOo 1.0 on the following systems :
I use these to write documents which I must export to .doc format and to create slideshows for courses on Linux. Yes, startup time is a little long, but I am now used to it, but working with it is no problem.
Mozilla is a little bit slow in the tooth, and I would certainly not recommend it for slower systems than mine, but it performs fine for me.
No problem running Linux and X on such a machine. I once installed it on my father's 100 Mhz Pentium.
These days I am experimenting with low-latency and preemptive kernels, and they do a real good job on such machines.
I think that the printer driver probably should make the decision.
The GPL also allows commercial use !!
This is something that I had already thought over in the past.
If you have an organisation that is large enough, you could pull of grabbing the codebase, modifying it and releasing the program without sources.
However, you will have do one of two things : saying that your program is compatible with the original, or saying that it is something totally new.
In the first case, you have locked yourself in, because if you claim compatibility, you must make this true, and you will need to follow that original organisation from which the codebase grew, and which is still developing. This means that the best you can do is just take their code everytime and provide new releases as fast as the original group.
If you say that it is something completely new, then you will have to make an effort in convincing people that it is better, but you cannot say that it is better than the original. If you really fork your codebase, than you will need a real large group of people with the same capacities as the original group from which the code forked.
Smaller projects can easily be taken over, but for large projects like the Linux kernel, I think it is not really practical, because of the investments needed. Investments in equally competent people, but also extra investments in marketing and so on.
You can base a product, a package, a distribution on it, but really forking Linux and maintaining a complete separate codebase ? It would be just an enormous waste of time and money.
Jurgen
As far as I am a computer geek, I also do like things simple. My oven only has two knobs, one for temperature, one for type of warming. The cooking plates each have one button (and I like cooking very much).
Take remote controls. I do not have a problem with them. For my father in law however, there should be a thing with only volume control, stations up and down and one button for remote powercontrol of his TV set.
Of course, it is also a question of philosophy. KISS is something that is not only good for design and maintenance, but also for usage. But what most people do not understand is that SIMPLE != EASY. It is much more difficult to come up with something simple and elegant, than it is to come up with something contrived. I think that simplicity is lost because of too short times-to-market. It results from the fact that designers do not have to time to look and search for several solutions to a problem, and pick the best. Instead, they use whatever solution they come up with.
I still have my Casio FX-8000G, though. My father bought it for me, for my 21st birthday, on the airport of Abu-Dhabi. It is 15 years old, it still serves me well, although maybe it could benefit from some cleaning. It has a binary and hex mode, and I think (but I am not sure) that you just put it in this mode, and that the A-F keys are usable without special shifting.
Jurgen
I have seen this in reality. On my previous job the systems were migrated to HP systems, which claimed 99.999%. One of the first things which broke was the redundant Fibre Channel controller. It took two days to fix it.
Most businesses not only have to account for daytime interactive processing, but also for processing at night and even in the weekends.
I do not think you have read 'Chthon' from Piers Anthony. Btw. I haven't found any references to his works of the seventies yet, Chthon, Sos The Rope, Var The Stick...