There are good reasons to run alternativee operating systems on Macs:
1. You have a Mac that's too old to run a recent MacOS.
2. You have a Mac you want to use as a server. This could be any Mac you might already have, or it could be a shiny new G4. G4s make incredible servers.:-)
3. You need to run software that runs on a unix, but not on MacOS.
4. You want to just install an OS and be done with it, instead of having to drive yourself crazy dealing with flakey hardware components that don't like each other.
There are plenty of others, but those are the highlights.
> to choose an operating sytem that works on both PowerMac and x86 hardware.
Aren't we forgetting NetBSD?
True, you can't buy it off the shelf most places, but it's been running on more different platforms (including PowerPC and Intel) for quite a while now.
Macs are dead? NCs are dead? What rock have you been living under?
None of these ways of doing things have infinite longevity; personally, I'll be glad to see current-day PCs bite the dust. It'll happen sooner than you think.
A number of commercial Playstation and N64 (!) games were written in a mixture of mostly-Lisp and some-C.
There have been small, fast Lisp implementations for a long time; it ran just fine on the earliest PCs. Lisp, Forth, and Smalltalk all have been made to run on tiny hardware (Palm, other PDAs just as examples).
It's always funny to see people write about Java and C++ as "object-oriented" languages, or how object-orientation "doesn't work", etc. If you want to see real OO at work, you have to look at Lisp and/or Smalltalk; even OO dialects of Forth. All of these can be compiled to native code...
A mobile Mac... is called a PowerBook, which until something actually ships with a Transmeta CPU in it, is the best-performing, coolest-running portable with longer battery life than pretty much anything else. A 300 MHz G3 consumes only about 2 or 3 watts of current for typical use... not as low as StrongArm or Transmeta, but a lot less than x86 hardware. As for temperature... the G3 in front of me usually stays around 31 C, as high as 35 C if I really push the system.
Of course, if you want to run a real operating system on your Mac, you could try NetBSD.
It's important to recognize the hard work of the LinuxPPC group in their efforts toward a Mac-operable Linux kernel, but their distribution leaves much to be desired.
I haven't tried Yellow Dog, but almost anything would be better than the LinuxPPC distro. Think RedHat 4.0... ported in a hurry to a different architecture by people who had other things on their minds. At least NetBSD was intended to be portable.
SuSE say their bringing out a beta version of 6.2 or 6.3 (forgot which) for the Mac. You might wait and check that out if you have your heart set on Linux.
It's not "only" the kernel, fortunately. It's also the BSD userland.
Good post... too bad some others need to whine about it not being GPL, etc.
How many of us really ever build big parts of the system from source anyway? As long as you can build a custom kernel, I don't think most people who are going to use the product will care about license or building the GUI.
If you've ever suffered through building X and KDE on your home box, think 5-10x that for a real GUI.
These are the difficulties I've run up against in using KDE 1.1.1 with FreeBSD 3.3-RELEASE:
kdm has some subtle configuration issues, and its error reporting is cryptic beyond the call of duty. I still don't have kdm working, and I've put a lot of time into it. The BSD FAQ makes it sound trivial to do (which it isn't) and the KDE FAQ glosses over it totally. It appears that no one on the KDE mailing lists or in #freebsd or #kde over multiple IRC networks has noticed that this problem exists.
Floppy device support for users other than root. BSD mount and/etc/fstab are different from Linux mount and/etc/fstab; the BSD fstab can't accept the "user" mount option, which KDE depends on to provide mount privileges to regular users. I tried to get around this with a sysctl -w vfs.usermount = 1 and chmod 777/dev/fd0, but still can't mount floppies from KDE. Naturally, if I go to the shell and run mount by hand, BSD mounts the floppy. This is fine for me, but I've been configuring a FreeBSD system from a non-technical friend. I'd have just gone with Linux, but FreeBSD feels a bit more responsive on her old 486 box; even this slight advantage is important enough for running Communicator that I'm reluctant to switch to another OS.
Yes, I too had noticed the info module not working, and it was clearly a/proc matter. Not a big deal, mostly.
I'm not sure what to make of it all. I am a very seasoned BSD user (since 4.2 running on a VAX) who dabbled in Linux for a while before realizing that beloved BSD runs on microcomputers these days.
I'm still far from knowing all the answers... BSD has changed a lot since 4.2 and running it on microcomputers is in many ways more challenging than running it on a VAX.;-) I've probably been around BSD longer than a good many of the #freebsd or #netbsd regulars.
I've found that asking for help in #linux or #freebsd is rarely helpful. The reasons for this seem to be:
1. Everyone assumes you're a newbie if they've never seen you before. Instant prejudice.
2. Most participants don't know any useful answers to (what should be) simple questions.
3. Some things don't work well on Linux because they came from BSD. Some things don't work well on BSD because they came from Linux. This is a much bigger problem than people customarily admit. The two operating systems are different enough to create real problems, despite binary compatibility. KDE, for example, is extremely Linux-centric and some of its features simply will not work on BSD even if you use the Linux compatibility API.
The parent post reads as if someone has had a very brief and bitter experience with BSD. That's sad. If that's how things are, at least know that it's not how they used to be. I think it is more than possible to have the same sort of experiences in the Linux world, and I think that there is some truth to the notion that most of the people active on Linux and BSD newsgroups and IRC channels simply don't have a clue to answer technical questions. It takes a great deal of perserverance to find useful information in these forums.
I *ran* a BBS on such a system back then, coded from scratch in a mess of BASIC and machine code. The whole thing, including messages, had to reside in memory (a whopping 16K).
There was something incredibly satisfying about the whole experience which I've never managed to find again. Perhaps running a MUD or something would be as good, but it would be prohibitively expensive.
Even LambdaMOO suffers from too many users, resulting in many people attempting to be completely outrageous or inane in order to be noticed. The old-timers who don't want anything to do with such nonsense keep to themselves for the most part, and have developed their own culture in the opposite direction; radically exclusive of the flaming masses, yet totally boring and staid.
There are other such places around which are more rewarding.
I was also one of Huffman's students. He was certainly a brilliant individual, there's no doubt about that. However, he was dismal as a teacher and students generally hated his classes. It was clear that he also hated having to teach us. You could learn a lot from him, but it was like fighting a war.
I just remembered -- it seemed to help me some -- try a vitamin B6 supplement. For me it was not a permanent solution, but it seems to help some people with the inflammation.
I am amazed that no one has mentioned acupuncture. It helped me more than drugs, more than chiropractic (though chiropractic can be a very good thing). Surgery was not an option for me (I'm also a musician). When looking for an acupuncturist, ask if they are planning to needle the carpal tunnel area itself. If the answer is yes, find another acupuncturist; needling the local area will just make you feel worse.
It helps to get ergonomic input devices (I like Datadesk, Kinesis). It helps to make sure your workstation is set up properly, too.
I think the most important thing is taking breaks every 20-30 minutes. Repetitive stress injuries only happen when you stay frozen in the same position for long periods of time. If you get up, stretch and move periodically, you are much less likely to suffer from this type of problem.
Macs do lose as a Linux platform, but not for the reasons you describe.
One of the biggest obstacles to Macs as a good Linux platform is that the kernel source is *still* not integrated into the main source tree. Every time one asks when this will happen, the answer is always Real Soon Now -- Maybe. Also, there are no good distributions that will run on a Mac. While the LinuxPPC group is to be commended for their kernel efforts, the LinuxPPC distribution can't compare to anything on Intel. Yellow Dog isn't much different; nor is Turbolinux; and Debian is still in the distant future.
The oft-cited non-expandability argument against Mac hardware is ancient history. When was the last time you saw a Mac that didn't have several PCI slots, several DIMM slots, and easy CPU upgrades? It's been years, folks.
I have an old 7500 chassis which has been upgraded through several generations of processors (now running a fast G3). I could put a swell G4 card in here if I cared to spend the money. Can you take a n old Pentium 75 box and put a 550 MHz Pentium III in it without a new motherboard? I hardly think so.
It's great to see someone acknowledge that Apple's hardware is in many ways superior to Wintel hardware.
I don't think Apple is more controlling than MS, however. Apple clearly follows standards more than MS does, and doesn't try to subvert standards for their own ends. Apple Java vs. MS Java is one good example.
There is no *official* Debian distribution for PowerPC; and there are no guarantees that there will be when Potato gets released, either. The list is so active because there have been many difficulties with the PowerPC port.
Heaven forbid a CPU being benchmarked should have the benchmark code optimized for it. Do you think Intel hasn't been hard at work optimizing the piss out of their own benchmarking code for decades?
Shame on Apple for trying to sell their products by making them look good.
I am, at this moment, working on a box that formerly had a 180 MHz 604e in it. It now has a 266 MHz G3 in it. The current configuration is about 70 percent faster for integer calculations and 50 percent faster for floating point than the previous one.
I don't know how you were testing, but you made a mistake.
We've all heard from wise Linux aficionados by now how decrepit and lame the MacOS is. Fine, it's nothing like a state-of-the-art OS. However, it is capable of getting out of the way of applications quite nicely; it doesn't consume more CPU at idle than Linux does, and it doesn't steal more than a bare minimum of cycles from the foreground application. Slashdaughters are constantly painting this grim picture of MacOS as a resource-hog whenever the subject of Macs comes up. Get over it! You're thinking of NT.
Slashdot Mirror
There are good reasons to run alternativee operating systems on Macs:
:-)
1. You have a Mac that's too old to run a recent MacOS.
2. You have a Mac you want to use as a server. This could be any Mac you might already have, or it could be a shiny new G4. G4s make incredible servers.
3. You need to run software that runs on a unix, but not on MacOS.
4. You want to just install an OS and be done with it, instead of having to drive yourself crazy dealing with flakey hardware components that don't like each other.
There are plenty of others, but those are the highlights.
> to choose an operating sytem that works on both PowerMac and x86 hardware.
Aren't we forgetting NetBSD?
True, you can't buy it off the shelf most places, but it's been running on more different platforms (including PowerPC and Intel) for quite a while now.
You're impressed...? ...that someone ported RedHat 6 to the PowerPC kernel? Big deal.
Yellow Dog is directly derived from LinuxPPC... big deal again.
If you want a real Linux on PowerPC, Debian will be ready soon.
Of course, you could bite down and try NetBSD.
gcc (egcs anyway) has the best optimizing compiler for the PowerPC architecture.
What in the world does this have to do with the poster's question about whether gcc is optimized for PowerPC?
Macs are dead? NCs are dead? What rock have you been living under?
None of these ways of doing things have infinite longevity; personally, I'll be glad to see current-day PCs bite the dust. It'll happen sooner than you think.
Indeed. Thanks for making these points.
A number of commercial Playstation and N64 (!) games were written in a mixture of mostly-Lisp and some-C.
There have been small, fast Lisp implementations for a long time; it ran just fine on the earliest PCs. Lisp, Forth, and Smalltalk all have been made to run on tiny hardware (Palm, other PDAs just as examples).
It's always funny to see people write about Java and C++ as "object-oriented" languages, or how object-orientation "doesn't work", etc. If you want to see real OO at work, you have to look at Lisp and/or Smalltalk; even OO dialects of Forth. All of these can be compiled to native code...
A mobile Mac... is called a PowerBook, which until something actually ships with a Transmeta CPU in it, is the best-performing, coolest-running portable with longer battery life than pretty much anything else. A 300 MHz G3 consumes only about 2 or 3 watts of current for typical use... not as low as StrongArm or Transmeta, but a lot less than x86 hardware. As for temperature... the G3 in front of me usually stays around 31 C, as high as 35 C if I really push the system.
;-)
Paving the way for a mobile Mac, indeed.
Of course, if you want to run a real operating system on your Mac, you could try NetBSD.
It's important to recognize the hard work of the LinuxPPC group in their efforts toward a Mac-operable Linux kernel, but their distribution leaves much to be desired.
I haven't tried Yellow Dog, but almost anything would be better than the LinuxPPC distro. Think RedHat 4.0... ported in a hurry to a different architecture by people who had other things on their minds. At least NetBSD was intended to be portable.
SuSE say their bringing out a beta version of 6.2 or 6.3 (forgot which) for the Mac. You might wait and check that out if you have your heart set on Linux.
It's not "only" the kernel, fortunately. It's also the BSD userland.
Good post... too bad some others need to whine about it not being GPL, etc.
How many of us really ever build big parts of the system from source anyway? As long as you can build a custom kernel, I don't think most people who are going to use the product will care about license or building the GUI.
If you've ever suffered through building X and KDE on your home box, think 5-10x that for a real GUI.
These are the difficulties I've run up against in using KDE 1.1.1 with FreeBSD 3.3-RELEASE:
/etc/fstab are different from Linux mount and /etc/fstab; the BSD fstab can't accept the "user" mount option, which KDE depends on to provide mount privileges to regular users. I tried to get around this with a sysctl -w vfs.usermount = 1 and chmod 777 /dev/fd0, but still can't mount floppies from KDE. Naturally, if I go to the shell and run mount by hand, BSD mounts the floppy. This is fine for me, but I've been configuring a FreeBSD system from a non-technical friend. I'd have just gone with Linux, but FreeBSD feels a bit more responsive on her old 486 box; even this slight advantage is important enough for running Communicator that I'm reluctant to switch to another OS.
/proc matter. Not a big deal, mostly.
kdm has some subtle configuration issues, and its error reporting is cryptic beyond the call of duty. I still don't have kdm working, and I've put a lot of time into it. The BSD FAQ makes it sound trivial to do (which it isn't) and the KDE FAQ glosses over it totally. It appears that no one on the KDE mailing lists or in #freebsd or #kde over multiple IRC networks has noticed that this problem exists.
Floppy device support for users other than root. BSD mount and
Yes, I too had noticed the info module not working, and it was clearly a
Thanks for reading.
I'm not sure what to make of it all. I am a very seasoned BSD user (since 4.2 running on a VAX) who dabbled in Linux for a while before realizing that beloved BSD runs on microcomputers these days.
;-) I've probably been around BSD longer than a good many of the #freebsd or #netbsd regulars.
I'm still far from knowing all the answers... BSD has changed a lot since 4.2 and running it on microcomputers is in many ways more challenging than running it on a VAX.
I've found that asking for help in #linux or #freebsd is rarely helpful. The reasons for this seem to be:
1. Everyone assumes you're a newbie if they've never seen you before. Instant prejudice.
2. Most participants don't know any useful answers to (what should be) simple questions.
3. Some things don't work well on Linux because they came from BSD. Some things don't work well on BSD because they came from Linux. This is a much bigger problem than people customarily admit. The two operating systems are different enough to create real problems, despite binary compatibility. KDE, for example, is extremely Linux-centric and some of its features simply will not work on BSD even if you use the Linux compatibility API.
The parent post reads as if someone has had a very brief and bitter experience with BSD. That's sad. If that's how things are, at least know that it's not how they used to be. I think it is more than possible to have the same sort of experiences in the Linux world, and I think that there is some truth to the notion that most of the people active on Linux and BSD newsgroups and IRC channels simply don't have a clue to answer technical questions. It takes a great deal of perserverance to find useful information in these forums.
I *ran* a BBS on such a system back then, coded from scratch in a mess of BASIC and machine code. The whole thing, including messages, had to reside in memory (a whopping 16K).
There was something incredibly satisfying about the whole experience which I've never managed to find again. Perhaps running a MUD or something would be as good, but it would be prohibitively expensive.
Even LambdaMOO suffers from too many users, resulting in many people attempting to be completely outrageous or inane in order to be noticed. The old-timers who don't want anything to do with such nonsense keep to themselves for the most part, and have developed their own culture in the opposite direction; radically exclusive of the flaming masses, yet totally boring and staid.
There are other such places around which are more rewarding.
I was also one of Huffman's students. He was certainly a brilliant individual, there's no doubt about that. However, he was dismal as a teacher and students generally hated his classes. It was clear that he also hated having to teach us. You could learn a lot from him, but it was like fighting a war.
Nevertheless, he'll be missed.
I just remembered -- it seemed to help me some -- try a vitamin B6 supplement. For me it was not a permanent solution, but it seems to help some people with the inflammation.
I am amazed that no one has mentioned acupuncture. It helped me more than drugs, more than chiropractic (though chiropractic can be a very good thing). Surgery was not an option for me (I'm also a musician). When looking for an acupuncturist, ask if they are planning to needle the carpal tunnel area itself. If the answer is yes, find another acupuncturist; needling the local area will just make you feel worse.
It helps to get ergonomic input devices (I like Datadesk, Kinesis). It helps to make sure your workstation is set up properly, too.
I think the most important thing is taking breaks every 20-30 minutes. Repetitive stress injuries only happen when you stay frozen in the same position for long periods of time. If you get up, stretch and move periodically, you are much less likely to suffer from this type of problem.
Macs do lose as a Linux platform, but not for the reasons you describe.
One of the biggest obstacles to Macs as a good Linux platform is that the kernel source is *still* not integrated into the main source tree. Every time one asks when this will happen, the answer is always Real Soon Now -- Maybe. Also, there are no good distributions that will run on a Mac. While the LinuxPPC group is to be commended for their kernel efforts, the LinuxPPC distribution can't compare to anything on Intel. Yellow Dog isn't much different; nor is Turbolinux; and Debian is still in the distant future.
The oft-cited non-expandability argument against Mac hardware is ancient history. When was the last time you saw a Mac that didn't have several PCI slots, several DIMM slots, and easy CPU upgrades? It's been years, folks.
I have an old 7500 chassis which has been upgraded through several generations of processors (now running a fast G3). I could put a swell G4 card in here if I cared to spend the money. Can you take a n old Pentium 75 box and put a 550 MHz Pentium III in it without a new motherboard? I hardly think so.
It's great to see someone acknowledge that Apple's hardware is in many ways superior to Wintel hardware.
I don't think Apple is more controlling than MS, however. Apple clearly follows standards more than MS does, and doesn't try to subvert standards for their own ends. Apple Java vs. MS Java is one good example.
There is no *official* Debian distribution for PowerPC; and there are no guarantees that there will be when Potato gets released, either. The list is so active because there have been many difficulties with the PowerPC port.
You can't run anything on Alpha hardware besides Linux. Unless you like running old bootleg copies of Windows NT.
Heaven forbid a CPU being benchmarked should have the benchmark code optimized for it. Do you think Intel hasn't been hard at work optimizing the piss out of their own benchmarking code for decades?
Shame on Apple for trying to sell their products by making them look good.
400 megabytes is *huge* for a unix-like OS.
Um, this is totally wrong.
I am, at this moment, working on a box that formerly had a 180 MHz 604e in it. It now has a 266 MHz G3 in it. The current configuration is about 70 percent faster for integer calculations and 50 percent faster for floating point than the previous one.
I don't know how you were testing, but you made a mistake.
We've all heard from wise Linux aficionados by now how decrepit and lame the MacOS is. Fine, it's nothing like a state-of-the-art OS. However, it is capable of getting out of the way of applications quite nicely; it doesn't consume more CPU at idle than Linux does, and it doesn't steal more than a bare minimum of cycles from the foreground application. Slashdaughters are constantly painting this grim picture of MacOS as a resource-hog whenever the subject of Macs comes up. Get over it! You're thinking of NT.