Glitz

Perhaps a make clean to remove the .o files and force a recompilation? Glitz.

Okay, the problem is that when you do the fopen, it returns a NULL pointer since the file doesn't exist. Then the cor=getc(arq) command segfaults because it is being passed a NULL pointer. Glitz.

I tried running your program. If there is no "test" file for it to open you get a segmentation fault. I created a "test" file and the program ran fine. Glitz.

getch() Can be found in curses.h Otherwise, use standard io functions such as: i=getchar() or i=getc(stdin) Glitz

Thanks, lots of interesting stuff there! Glitz.

I'm no legal beagle but I think a provision like that is required in a license. I don't think you can sell someone something before they have a chance to read and accept the license and then when they do read the license deny them any means of returning the product for a refund if they don't accept the license. Can you imagine the marketing schemes possible if this were legal? Here's my EULA: You agree to pay XYZ company a yearly user's fee, to be established by XYZ company on a yearly basis, for a minimum of 10 years. . . . If you do not agree to the terms of this EULA then you may not use the software. How do you like that? Either you use the software and I get a user's fee and the original purchase price of the software or you don't get to use the software, I don't get the yearly user's fee, but I still get the original purchase price of the software. In other words, if you don't want to pay me the user's fee then you have just paid me for the priviledge of holding in your hands a software package that you can not use. Pretty sweet huh? Somehow I don't think that kind of situation would fly in a court of law. However, Toshiba may change their EULA to indicate that the whole computer must be returned. That would be fair, however, I think they would be required to refund any shipping unless they get the customer to agree to the EULA before shipping. Glitz. PS. If you do get something like that to work then please let me know!

Apparently the new patch makes it so that it can run without using the advanced texture compression. This allows a wider range of graphics cards to be used with the game. Glitz.

Java is overkill considering the nature of the problem and the because there is no operating system and it would have to be completely taylored for the hardware. With the BASIC interpreter I hope to produce a small stand alone program that only requires hardware specific modifications in a few places. This does have to be bootable from a 32KB serial eeprom. I have looked for any type of configuration settings that allow me to interpose a script that does some work before calling gcc. No luck. I will probably add the script to the rc.local startup script. That should be fine. I rarely have to recompile 176 times in one day :wink: I really have to look into learning bash. Glitz.

The problem is a little vague but I think that windows is set to rescan the network every 30 minutes or so. This causes problems if a new computer is booted up and windows misses its announcement that it has connected to the network. This also happens amongst windows machines. Therefore, it wont pick up the new machine until it rechecks the network. I don't know if there is a way to force it to rescan the network or shorten the rescanning period. Linux allows you to force a rescan any time. I don't know if this is actually the problem you're describing but that's my $0.02 worth (not adjusted for inflation :wink: ). Glitz.

I suspect they got the motherboards dirt cheap due to the speaker connector problem. There's no way ASUS would be caught dead selling motherboards with a problem like that. If this doesn't bother you then I wouldn't consider it a big problem. The refurbished hard drive sets off red flags though. As usual the graphics chipset may be a problem if you want 3D. SCSI shouldn't be a problem. Glitz.

I don't know why they are disappearing. It may have something to do with the IDE I'm using. I use C-Forge and it has a built in terminal window that it spits compiler messages to. The window probably links to a TTY when it is active. Maybe there is a bug that deletes the TTY after. There are 176 ttys available so it takes a while to use them all up (ie. compile 176 times). It is an early MacMillan version of the program and I got it doubly cheap (~$10) since the store had it on sale. I really like the environment though. This version of C-Forge is taylored to C/C++, python, and tck/tk. I'm currently developing a BASIC interpreter optimized for use in larger embedded systems (~128KB memory) that use a watchdog timer. I'm writing it in C for portability and easy customization to different platforms. It is designed for non-preemptive multitasking that is transparent to the BASIC program (ie. The program does not have to worry about multitasking at all, especially allowing other parts of the system to run) and that deterministically returns to the C main loop before the watchdog timer times out. I intend to use it for writing all the fiddly user interface code which I currently write as huge state machines in C (since in order to deterministically return to the main loop before the watchdog timer expires requires that I cannot wait in a loop or perform long computations all in one go). I could use pre-emptive multitasking which would relieve all my headaches but then there is no way to quickly and accurately determine if any given task has entered an infinite loop or is executing runaway code. And tickling the watchdog timer at more than one location in the program is bad practice. Glitz.

Thanks guys!!! I'm using LM8.1 and makedev needs to be capitalized (ie. MAKEDEV). MAKEDEV also doesn't work properly (I had this problem before). It will create a device with a - after the name and then fail saying that the device could not be created do to permission problems. Anyway, to make a long story short I modified the script to use mknod as follows: #!/bin/bash n=0 for i in p q r s t u v w x y z; do for j in 0 1 2 3 4 5 6 7 8 9 a b c d e f; do mknod -m 666 /dev/tty$i$j c 3 $n chown root /dev/tty$i$j chgrp tty /dev/tty$i$j n=$((n+1)) done done It works like a charm. Thanks again. Glitz. PS. This scripting language has some powerful for loops.

Hi guys! I have a problem with disappearing terminal devices (the TTYxx) devices. Everytime I compile a program with my IDE one more device disappears. When they are all gone the compiler isn't invoked anymore from the IDE. Right now I have to reboot and hit the i option during boot to get the device nodes to be created again. What I am looking for is a quick little script to run from root that will create the devices (so I don't have to reboot). The devices to be created are as follows: crw-rw-rw- root tty 3, 0 ttyp0 crw-rw-rw- root tty 3, 1 ttyp1 crw-rw-rw- root tty 3, 2 ttyp2 . . . crw-rw-rw- root tty 3, 15 ttypf crw-rw-rw- root tty 3, 16 ttyq0 . . . crw-rw-rw- root tty 3, 31 ttyqf crw-rw-rw- root tty 3, 32 ttyr0 . . . crw-rw-rw- root tty 3, 175 ttyzf I don't know enough about the bash shell to create an iterative process to create the devices. Any help would be appreciated. Glitz.

You have to set up your CDRW as a scsi device in order for any CD burning software to recognize it. What does it say when you try to mount the CDROM? Glitz.

Does that mean that the modem is working too? Glitz.

Okay, from what I can gather from your post you have two problems: 1) You can't play CDs from your CDRW. 2) You can't mount a data CD in either your CDROM or CDRW. Please correct me if I'm wrong. Well, the answer to 1) is probably that you do not have an audio cable connecting the CDRW to the sound card. There are some programs that will play CDs by extracting the digital information right off the CD but most programs just use the audio cable. I'm not sure about 2). Could you provide some more information like what commands you tried and what error messages you got back. Glitz.

Is supermount really so buggy that it is unusable for everyone? I've used Mandrake 8.2 and recently 9.0, have always had supermount on, and I've never had a problem with it. Am I in the minority? I haven't done anything special to make sure it works. I think it's a "combination of hardware" thing. It seems to work for some people and not for others. The symptoms are even different amoung people who have problems with it. That's what makes it so maddening. You're never quite sure if it is supermount or a hardware problem until you actually solve the problem. Ah, well, whada ya gonna do? Glitz.

I still use LM8.1 as my workhorse. I didn't bother with 8.2 because of the missing RPM fiasco. LM9.0 is also still marked experimental on my system (at least until I get everything I use under 8.1 running on it). Anyway, your best bet for 8.1 is to install the nvidia binaries by hand. If you use XFDrake you will have to edit the XF86Config-4 file by hand and get rid of the illegal 32bit screen modes it puts in there. Glitz.

1) The user configuration tool only needs to be modified to add a link from the /home/user/device directory to the /mnt/ directory. 2) Unfortunately supermount still doesn't work. But I think that a small tutorial supplied with linux showing linux users coming from windows how to do these sorts of things would solve all those sorts of problems. Unfortunately no one has ever thought of that. 3) This is a licencing issue. All boxed sets should include a separate CD with non-GPL software that clearly states that the contents may not be distributed. 4) Yes but most new users will stick to one distro until they are either familiar with linux or totally frustrated with it. The tutorial mentioned earlier would go a long way in this respect. 5) That's up to the person creating the package. Some kind of standard for how to put icons on desktops and configure menus would help tremendously. Glitz.

You have to make sure that the hosts.allow and hosts.deny files are set up properly to allow connections from your intranet. Glitz.

eeep now thats something I didn't expect to happen. Does it do that if I ssh from a windows box also? Now theres something I"m just going to HAVE to try Monday from work. I don't know. It tunnels to the host system's X window server. Since X windows is not running on the windows machine I don't think it will work. It can't hurt to try though... Glitz.

If you really want to get blown away, try logging on through ssh and then run a graphical application... :wink: Glitz.

I use a dual 433MHz celeron motherboard. It runs great. The amount of usage you get out of 2 CPUs increases with the number of processes and/or threads you run at the same time. Most software only use one thread so you wont see any increase by just running that per se but if you run other stuff at the same time it helps. Some newer games use multiple threads pretty well. I bought UT2003 and it makes pretty good use of both processors (on average about 1.5 times the power of 1 CPU). It really works for me when I run programs that I write that just do plain computations flat out. The program will use one CPU at 100% constantly and I can still get good mouse and X windows response when using other programs on the second CPU. However, a single CPU at twice the dual CPU speed will always be as good or better. Glitz.

Your problem is with the automount daemon (AMD). The reason it doesn't work properly seems to be that it is called too late in the shutdown process. Either disable it entirely or get it to shutdown sooner. There is a thread on this board describing how to do this. Glitz.

Interoperability standards are a good thing. The linux kernel is based on the POSIX standard which is why a lot of UNiX software can be easily ported to it. You could concevably write an operating system that looks and behaves exactly like linux (from the user's standpoint) yet have an API that is so foreign that it would be very difficult to port even UNIX programs to it. On the other hand, you could also write an operating system that behaves like windows (again, from the user's standpoint) but uses APIs that comply with the POSIX standard and could use UNIX programs. It all depends on what aspects you standardize. It would be nice to standardize the directory structure and base libraries that linux distros contain so that software companies wouldn't always have to hit multiple moving targets when porting their software to linux. Glitz. PS. The only aspect of the GUI that needs to be standardized is cut and paste. As long as information can be passed from one program to another that's fine. The look and feel of the window manager is irrelevant to the actual software.