Achem... That weird "disklabel" method is actually the standard UNIX way to do partitions. The idea is that "a =/", "b = swap", "c = whole disk", and the rest of the letters to "/usr", "/var", etc.
The other *BSDs do it that way, Solaris does it, I think IRIX does it, AIX doesn't though (but then again, AIX uses it's fancy volume manager by default).
I've used the Super T-Comp before myself. It's a nice board. I've even build my own I/O expansion circuitry for it. Ray also provides full schematics, which were great at learning how such things are designed, and came in handy when I went to expand upon it.
Some of the commercial UNIX compilers are actually pretty good. I've been happy in my experiences with MipsPro lately. I've also heard SunPro is good, but I don't know about the "standards" side... same with IBM's xlc. On the other hand, "xlc" is one of the few compilers that actually DOES CATCH that missing semicolon without farting on itself.:)
Well, this game truely is a test of hardware. I'm running it at high detail on a 1.2GHz Athlon T-bird with 256MB of RAM and a GeForce 2 GTS 32MB DDR, and it still bumps occasionally.
On the flip side, the graphics look so friggin incredible that I don't care:)
Actually, GAIM sucks. Why? Because it's nothing but a huge memory leak. I would be using it myself, if it weren't for that fact. On one hand, you could argue that it could be run from a machine with gobs of RAM and X-forwarded to the machine you want to chat with. However, it leaks into the X server's memory, so that won't help. I've been looking for a decent AIM client for a while, and my current solution is to run AOL's Linux client on my FreeBSD server and X-forward it to my SPARCstation 5 (machine I chat from). It's a shame, but getting any AIM/ICQ clients to build on non-Linux is damn near impossible. (Tik works because it's Tcl/tk, but is also damn slow and sluggish)
Achem! What's to say you can't run gdm on a Sun? That login manager you're complaining about is just "dtlogin", which is part of CDE. If you just hunt down the right config file, you can change it to whatever you want. X is a standard, and XFree86 is nothing more than an implementation of that standard. "Xsun" is also an implementation that basically works the same way.
Now I'm waiting for it to be put in distros. I guess I find it annoying how some distros (Mandrake, RedHat) go crazy if you change their kernel. However, they are better for getting a desktop to JUST WORK out of the box.
Did you suggest using gcc on UltraSparc and POWER processors for benchmark comparisons? You've got to be mad. Those are CPUs where the commercial compilers, such as Sun Workshop (SunPro) and xlc (or Visual Age C or whatever IBM calls it these days), frankly kick the crap out of gcc. Gcc is only a good compiler when you use it to build x86 code, and "maybe" Alpha code.
Then again the Itanium might look like a joke once the 21364 Alpha (if it ever comes) and the POWER4 hit the market.
I'm about to upgrade my computer (next week, when the parts finally get here) to a setup that will be using a UDMA100 card for the hard drive. I'm gonna want to reinstall Linux from scratch on the thing. (old setup has a badly hacked up Mandrake 6.1 install on it) Is there any distro that will boot on my new hardware? Or do I have to first install differently, then build a new kernel, then swap hardware and hope it works?
I find this interesting, as I saw things moving in that direction over the two years I participated.
When I first participated in '98, everyone was involved. The team was basically a joint effort between the engineers and the students. This works especially well when you have young engineers (recently out of college). Everyone felt a sense of pride in the project, and felt like they were part of it. It also took lots of time, as I basically spend every evening and saturday for 6 weeks down there in the workshop. I mainly worked on the code, but I also helped with the rest as needed. I even got to be one of the driver/operator people that year, which was pretty cool.:)
In the '99 competition, however, things changed. It seemed like the engineers (mostly the same ppl) decided to cut the students out of the loop. We were given a sense of participation in the begining, but it degraded over time. We only met a few days a week. The engineers tried to do most of it themselves, and seemed to procrastinate about it. I still worked on the code, so I was involved. However, most of the students (a mostly new group) didn't seem involved at all. I felt a lot of bitterness between myself and the engineers that year. The students were generally less self-motivated, which could have partially led to it, but the lack of veterans kept most of them from noticing it.
Getting glory for the controls stuff is indeed much more difficult than expected. On my team (#179, Swamp Thing) we put a lot of work into the control system and had lots of different features. Multiple speed modes, adjustable torque/gearing control, auto-torque limiting, closed-loop positioning, and automation of many things that other teams left to manual control. We even had a fancy manual describing our whole system and it's usage.
In the 1998 competition, we lost the controls award, but felt no grief. Why? The team who won it had done everything we did, and then some. They deserved to win.
In 1999, however, we were really angry that we lost. Why? The awary was totally misjudged. The winner was cited for a creative robot idea (being able to climb over other robots), and not an innovative controls system.
Another dissappointment is that the nature of the 1999 competition did not allow us to "show off" our nifty new features that would have blown people away on the '98 robot, but were developed too late to be of competition use that year. (such as complex closed-loop arm control)
I was active in the '98 and '99 competitions, and what you say is 100% true. This competition is mostly about the mechanical engineering stuff. However, I was the person who worked on the coding. It was rather interesting what you can do when you program the interface between joysticks/buttons, and a mechanical device. With my team, myself and one other guy put a lot of work into the code, and came up with some pretty neat stuff. However, our hands were tied beyond a point, as it is mainly a mechanical competition.
You can code some pretty neat features that ease operator burden, but you can only go so far. The provided microcontroller is based on the Parallax Basic STAMP II, and you run into it's limitations (32 bytes for variables, no interrupts, slow execution, no negative numbers) very fast. 2/3 the effort in coding routines was spend figuring out how to make them work on the thing.
I can hear that CRT scanning noise too. It really annoys me. I discovered after spending time at college only behind high quality Sony Trinitron computer monitors, that the scanning noise of a normal TV drives me nuts and can give me a headache.
The fan noise doesn't bother me at all. Well, when I tried running a SPARCserver 670MP in my dorm room, it did. However, nothing else there makes an annoying fan noise. Actually, the most audible noise from my rack of machines is a hard drive. And it's only a soft high-frequency whine that isn't annoying at all.
Finding old Sparc machines is actually very easy, and a lot of hobbyists use them. Just look at sites like "www.sunhelp.org". Second in line from that list would probably be SGI. But people have to work around the company to get them, as IRIX costs an arm/leg.
However, as a collector, I also have an IBM RS/6000. I've looked far and wide, and I'm convinced that there are only 15 hobbist RS6k users in the world. (which includes myself and the guy I got my POWERstation 350 from) On the other hand, old RS6ks kick the crap out of old SPARCs for performance. The "POWER" processor (well, chipset, actually) kicks ass.
I've got a bunch of Sparcs sitting around, and I didn't pay an arm/leg for them.
SPARCstation 5 (Solaris 8, secondary desktop)
SPARCstation IPX (OpenBSD 2.8, firewall)
SPARC Xterminal 1 (stole RAM for SS5, Xt1 to appear on eBay)
SPARCserver 670MP (makes too much noise. parting out on eBay)
Of course I collect my machines for variety. If they were all running one OS, I wouldn't see the point. If I wanted to use Linux, I would just use a PC. If I have a Sparc, normally that means I want to play with Solaris. Likewise, my SGI runs IRIX, and my RS/6000 runs AIX.
I really do collect "UNIX Diversity". Look at my "systems" page:
http://www.logicprobe.org/systems/
When I got my first Sun, a SPARCstation IPX, I had an interesting OS adventure with it. In the begining, all I had was serial console. So, I fooled around. I think I tried RedHat, but Linux distros are total crap when I comes to network installation. So, I went with NetBSD. It worked pretty nicely. Once I got a second monitor for it (got a 17" Sony for my PC, and gave it my 15" Sony), I started trying more things. I think I first tried Solaris 2.4. Ran just fine, but it was so old that finding stuff for it wasn't easy. Then I tried Solaris 7. Sol 7 has some big problems with it that caused it to run at 1/4 the speed on my IPX as it did on a friend's SS2 (equivalent machine). So, I dug up a copy of Solaris 2.6, and the machine's been running that just fine for quite a while.
Recently my best Sun has been a SPARCstation 5 running Solaris 8, so I found a new use for my IPX. I threw an s-bus NIC into it (hard to find, since Suns all come with on-board NICs, but I had it laying around), and installed OpenBSD. Makes a great firewall. Fits under the network switch just nicely. (and works a lot better than the crappy crash-happy PC I was trying to do it with before. it was hardware, as I was using OpenBSD in both cases)
I'm using an SGI Indigo2. While the display quality is very good, I didn't notice that X did any anti-aliasing at all. If it did on your O2, how do you enable it?
Pretty funny how you put it like that. People have their Windows machines, and their pet Linux box in the corner. Cool to tinker with, but not much beyond that.
Well, I'm practically the opposite of you. Before college, I had Linux on my machine but used Windows most of the time. When I got here, I started using Linux nearly all the time.
Now, I too have "moved on" from playing with Linux.
I'm now running FreeBSD, Solaris, IRIX, and AIX.
In fact, I've been using an SGI Indigo2 as my main desktop since the summer, and I'm looking at Sun for my next machine.
I've got my pet Windows box, and that's dual boot.
I'd try it, but I've only got a POWERstation 350. I don't have all week;) Besides, sticking to CDE is fine for the casual xterm usage I get out of the machine.
On the otherhand, KDE2 (well the RC, anyways) builds rather cleanly under IRIX. I'm back to 4Dwm now, though, as it's better tweaked for the system. I'm gonna try building Qt and KDE2 with exception handling off now (for performance) that I figured out the MIPSpro compiler flag "-LANG:exceptions=OFF". (as opposed to "-fno-exceptions" or whatever gcc uses)
Slashdot Mirror
Achem... That weird "disklabel" method is actually the standard UNIX way to do partitions. The idea is that "a = /", "b = swap", "c = whole disk", and the rest of the letters to "/usr", "/var", etc.
The other *BSDs do it that way, Solaris does it, I think IRIX does it, AIX doesn't though (but then again, AIX uses it's fancy volume manager by default).
I've used the Super T-Comp before myself. It's a nice board. I've even build my own I/O expansion circuitry for it. Ray also provides full schematics, which were great at learning how such things are designed, and came in handy when I went to expand upon it.
See the Super T-COMP's site:
http://www.teleport.com/~raybutts/index.htm
I read all about this Trinity competition back when I was building my own robot project. I would have entered, had I not lived in Florida.
See my project, Chip II:
http://www.logicprobe.org/~octo/robot/
Some of the commercial UNIX compilers are actually pretty good. I've been happy in my experiences with MipsPro lately. I've also heard SunPro is good, but I don't know about the "standards" side... same with IBM's xlc. On the other hand, "xlc" is one of the few compilers that actually DOES CATCH that missing semicolon without farting on itself. :)
Well, this game truely is a test of hardware. I'm running it at high detail on a 1.2GHz Athlon T-bird with 256MB of RAM and a GeForce 2 GTS 32MB DDR, and it still bumps occasionally. On the flip side, the graphics look so friggin incredible that I don't care :)
Actually, GAIM sucks. Why? Because it's nothing but a huge memory leak. I would be using it myself, if it weren't for that fact. On one hand, you could argue that it could be run from a machine with gobs of RAM and X-forwarded to the machine you want to chat with. However, it leaks into the X server's memory, so that won't help. I've been looking for a decent AIM client for a while, and my current solution is to run AOL's Linux client on my FreeBSD server and X-forward it to my SPARCstation 5 (machine I chat from). It's a shame, but getting any AIM/ICQ clients to build on non-Linux is damn near impossible. (Tik works because it's Tcl/tk, but is also damn slow and sluggish)
Achem! What's to say you can't run gdm on a Sun? That login manager you're complaining about is just "dtlogin", which is part of CDE. If you just hunt down the right config file, you can change it to whatever you want. X is a standard, and XFree86 is nothing more than an implementation of that standard. "Xsun" is also an implementation that basically works the same way.
Now I'm waiting for it to be put in distros. I guess I find it annoying how some distros (Mandrake, RedHat) go crazy if you change their kernel. However, they are better for getting a desktop to JUST WORK out of the box.
Did you suggest using gcc on UltraSparc and POWER processors for benchmark comparisons? You've got to be mad. Those are CPUs where the commercial compilers, such as Sun Workshop (SunPro) and xlc (or Visual Age C or whatever IBM calls it these days), frankly kick the crap out of gcc. Gcc is only a good compiler when you use it to build x86 code, and "maybe" Alpha code. Then again the Itanium might look like a joke once the 21364 Alpha (if it ever comes) and the POWER4 hit the market.
I'm about to upgrade my computer (next week, when the parts finally get here) to a setup that will be using a UDMA100 card for the hard drive. I'm gonna want to reinstall Linux from scratch on the thing. (old setup has a badly hacked up Mandrake 6.1 install on it) Is there any distro that will boot on my new hardware? Or do I have to first install differently, then build a new kernel, then swap hardware and hope it works?
I find this interesting, as I saw things moving in that direction over the two years I participated.
:)
When I first participated in '98, everyone was involved. The team was basically a joint effort between the engineers and the students. This works especially well when you have young engineers (recently out of college). Everyone felt a sense of pride in the project, and felt like they were part of it. It also took lots of time, as I basically spend every evening and saturday for 6 weeks down there in the workshop. I mainly worked on the code, but I also helped with the rest as needed. I even got to be one of the driver/operator people that year, which was pretty cool.
In the '99 competition, however, things changed. It seemed like the engineers (mostly the same ppl) decided to cut the students out of the loop. We were given a sense of participation in the begining, but it degraded over time. We only met a few days a week. The engineers tried to do most of it themselves, and seemed to procrastinate about it. I still worked on the code, so I was involved. However, most of the students (a mostly new group) didn't seem involved at all. I felt a lot of bitterness between myself and the engineers that year. The students were generally less self-motivated, which could have partially led to it, but the lack of veterans kept most of them from noticing it.
Getting glory for the controls stuff is indeed much more difficult than expected. On my team (#179, Swamp Thing) we put a lot of work into the control system and had lots of different features. Multiple speed modes, adjustable torque/gearing control, auto-torque limiting, closed-loop positioning, and automation of many things that other teams left to manual control. We even had a fancy manual describing our whole system and it's usage.
In the 1998 competition, we lost the controls award, but felt no grief. Why? The team who won it had done everything we did, and then some. They deserved to win.
In 1999, however, we were really angry that we lost. Why? The awary was totally misjudged. The winner was cited for a creative robot idea (being able to climb over other robots), and not an innovative controls system.
Another dissappointment is that the nature of the 1999 competition did not allow us to "show off" our nifty new features that would have blown people away on the '98 robot, but were developed too late to be of competition use that year. (such as complex closed-loop arm control)
I was active in the '98 and '99 competitions, and what you say is 100% true. This competition is mostly about the mechanical engineering stuff. However, I was the person who worked on the coding. It was rather interesting what you can do when you program the interface between joysticks/buttons, and a mechanical device. With my team, myself and one other guy put a lot of work into the code, and came up with some pretty neat stuff. However, our hands were tied beyond a point, as it is mainly a mechanical competition.
You can code some pretty neat features that ease operator burden, but you can only go so far. The provided microcontroller is based on the Parallax Basic STAMP II, and you run into it's limitations (32 bytes for variables, no interrupts, slow execution, no negative numbers) very fast. 2/3 the effort in coding routines was spend figuring out how to make them work on the thing.
I can hear that CRT scanning noise too. It really annoys me. I discovered after spending time at college only behind high quality Sony Trinitron computer monitors, that the scanning noise of a normal TV drives me nuts and can give me a headache.
The fan noise doesn't bother me at all. Well, when I tried running a SPARCserver 670MP in my dorm room, it did. However, nothing else there makes an annoying fan noise. Actually, the most audible noise from my rack of machines is a hard drive. And it's only a soft high-frequency whine that isn't annoying at all.
Today, if you've got 10k to burn, you can buy a new Sun Blade 1000, which has an UltraSparc-III processor.
Finding old Sparc machines is actually very easy, and a lot of hobbyists use them. Just look at sites like "www.sunhelp.org". Second in line from that list would probably be SGI. But people have to work around the company to get them, as IRIX costs an arm/leg.
However, as a collector, I also have an IBM RS/6000. I've looked far and wide, and I'm convinced that there are only 15 hobbist RS6k users in the world. (which includes myself and the guy I got my POWERstation 350 from) On the other hand, old RS6ks kick the crap out of old SPARCs for performance. The "POWER" processor (well, chipset, actually) kicks ass.
I've got a bunch of Sparcs sitting around, and I didn't pay an arm/leg for them.
SPARCstation 5 (Solaris 8, secondary desktop)
SPARCstation IPX (OpenBSD 2.8, firewall)
SPARC Xterminal 1 (stole RAM for SS5, Xt1 to appear on eBay)
SPARCserver 670MP (makes too much noise. parting out on eBay)
Of course I collect my machines for variety. If they were all running one OS, I wouldn't see the point. If I wanted to use Linux, I would just use a PC. If I have a Sparc, normally that means I want to play with Solaris. Likewise, my SGI runs IRIX, and my RS/6000 runs AIX.
I really do collect "UNIX Diversity". Look at my "systems" page:
http://www.logicprobe.org/systems/
When I got my first Sun, a SPARCstation IPX, I had an interesting OS adventure with it. In the begining, all I had was serial console. So, I fooled around. I think I tried RedHat, but Linux distros are total crap when I comes to network installation. So, I went with NetBSD. It worked pretty nicely. Once I got a second monitor for it (got a 17" Sony for my PC, and gave it my 15" Sony), I started trying more things. I think I first tried Solaris 2.4. Ran just fine, but it was so old that finding stuff for it wasn't easy. Then I tried Solaris 7. Sol 7 has some big problems with it that caused it to run at 1/4 the speed on my IPX as it did on a friend's SS2 (equivalent machine). So, I dug up a copy of Solaris 2.6, and the machine's been running that just fine for quite a while.
Recently my best Sun has been a SPARCstation 5 running Solaris 8, so I found a new use for my IPX. I threw an s-bus NIC into it (hard to find, since Suns all come with on-board NICs, but I had it laying around), and installed OpenBSD. Makes a great firewall. Fits under the network switch just nicely. (and works a lot better than the crappy crash-happy PC I was trying to do it with before. it was hardware, as I was using OpenBSD in both cases)
It's also got a place on my desktop. Then again, so does IRIX, Solaris, and AIX. (FreeBSD on the server)
I'm using an SGI Indigo2. While the display quality is very good, I didn't notice that X did any anti-aliasing at all. If it did on your O2, how do you enable it?
Pretty funny how you put it like that. People have their Windows machines, and their pet Linux box in the corner. Cool to tinker with, but not much beyond that. Well, I'm practically the opposite of you. Before college, I had Linux on my machine but used Windows most of the time. When I got here, I started using Linux nearly all the time. Now, I too have "moved on" from playing with Linux. I'm now running FreeBSD, Solaris, IRIX, and AIX. In fact, I've been using an SGI Indigo2 as my main desktop since the summer, and I'm looking at Sun for my next machine. I've got my pet Windows box, and that's dual boot.
or IRIX, AIX, Solaris, HP-UX, Tru64, etc....
(I tend to diversify my UNIX experience. I've got most of the bases covered.)
http://www.logicprobe.org/systems
No, EBCDIC!
(ASCII is too compatable for IBM)
j/k, of course
Many more of us have professors who ALWAYS give lectures in powerpoint.
I'd try it, but I've only got a POWERstation 350. I don't have all week ;) Besides, sticking to CDE is fine for the casual xterm usage I get out of the machine.
On the otherhand, KDE2 (well the RC, anyways) builds rather cleanly under IRIX. I'm back to 4Dwm now, though, as it's better tweaked for the system. I'm gonna try building Qt and KDE2 with exception handling off now (for performance) that I figured out the MIPSpro compiler flag "-LANG:exceptions=OFF". (as opposed to "-fno-exceptions" or whatever gcc uses)
4Dwm
It's got that lightweight characteristic, and the "desktop environment" thingie!