Toshiba Satellite Pro 6100 with Linux

Email me at or x38x107x191x111 @weller.ws (anti-spam experiment)

(Note: there might be additional detail in my redhat 7.3 instructions, since I'm still in the process of tweaking Mandrake) I'd like to eventually recompile my kernel to include the latest features. I like this exec shield feature (patch). I'm also interested in the kernel's new laptop-mode (more). The kernel batches disk I/O, allowing the disk drive to become idle long enough for the drive's power-saving features to take affect. It looks like there's a nice accumulated patch for all laptop-related patches for 2.4.21.

BIOS Settings
Windows XP
Mandrake Install
Sticky Key Fix
Update Video Driver
Getting X on TV and external monitor
Aesthetics
Applied Mandrake Updates
Network Timeout - too long
Wireless
Iptables (Firewall)
Iptables (Router)
Minimizing Drive Noise
MP3 Player USB drive
Power Suspend/Resume (i dont use this anymore)
Time
Toshiba Utilities
Toshiba Command-line utility
Fan Setup
Modem
Fonts
Applications

Mplayer
LICQ
Java
OpenOffice
Mozilla
Alvaro's MSN
Apache
Real Player
Wine + Kazaa + Nero
VMWare
VMWare - using SAME Windows XP on partition 1
Sniffers
BitTorrent and Python

Over-Heating Problem!
Swap problem
DVD-R read problem
Kernel Upgrade

BIOS Settings

Hold ESC while turning on computer, then press F1.
Set Boot Mode to "resume" to allow resuming after suspending
Set Battery Power to "low", and performance to "optimized" to save on battery power. Set LCD to "bright" instead of "super-bright".

Windows XP
- registered with toshiba + microsoft
- set screen res to 1600x1200 and aurora background
- configured isp for quick internet access
- installed a few "must" apps (ssh, icq, nero, partition magic 7)
- cleaned up desktop
- ran "msconfig" and disabled a lot of unnecessary things like norton; (also more in control panel-> admin tools -> services).
- installed explore2fs so that i can browse my ext3 filesystem from XP.
- actually, since I'm using reiserfs, I now have to use: rfstool. Usage is something like (from DOS prompt): rfstool -p0.16 weller/installs (for /home/weller/installs). "rfstool -p0.16 weller/installs/winamp3_0-full.exe winamp.exe" to extract that file.
- using partition magic 7 (ntfs for XP requires 7 or higher), shrunk 38gig partition to 19gig. it rebooted and ran from DOS, then rebooted again and i was set.
- My wireless network drops connection and reconnects every 3-8 minutes. If I'm in the middle of a download, it aborts. If I'm in SSH, I have to reconnect. I tried the latest toshiba wlan driver. I even tried rolling it back. Is anyone else experiencing this? I hate windows anyways. It's so user-unfriendly and unintuitive. I'm not experiencing any of these problems in linux :)
- control panel -> system -> advanced -> error reporting -> disable. (this gets rid of the annoying "Send Error Report" whenever you get a coredump.
- I dont know what I installed, but my "view wireless network connections" would come up blank (The Advanced is greyed out, and all i have to choose is Cancel). The solution was to add dword "start" with "value" of 2 to HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\W2CVC so that Wireless Zero Configuration Service would show up in services.
- this driver seems to work. i uninstalled and installed this one.
Mandrake Install
- i hooked up optical USB mouse since i hate the laptop mouse
- booted off CD by holding down "c" when turning machine on
- did basic mandrake install; chose all common sense defaults
- when it rebooted, it hung on "initializing usb controller". i now realize that the system has to be powered off to get past this. you'll never be able to reboot. let me know if you resolve this.
- Used standard defaults during install... set security to "standard"
- Custom disk partitioning: linux swap (352M), set up remaining partitions are REISERFS and not EXT3. I suspected that EXT3 was responsible for the drive clicks. /var - 1.2 gig, / - 5 gig, /home - remainder.
- set root pass; added myself as a normal user;
- selected option to automatically login
- installed bootloader on MBR
- graphical interface -> selected "configure". set to 1600x1200 flat panel; geforce 4; xfree86 4.3; 24-bit colour ; start on boot
- network -> configure -> wLAN -> eth0 = wireless: selected "track network card id", start at boot, managed, any
- network -> configure -> LAN -> eth1 -> automatic IP
- zeroconf hostname and hostname -> i set these to the same thing (nickname for my laptop)
- the internet update did not work. i selected it, but it rebooted right away. i manually did it later.
Sticky Key Fix

emacs /etc/X11/XF86Config-4
under Section "InputDevice":
Option "XkbDisable"

Update Video Driver

Fetch latest NVidia Driver for GeForce4 420 Go NV17
rebooted. hit ESC at lilo splash screen. type "linux single"
"sh NVIDIA-Linux-x86-1.0-4363.run"
Modify /etc/X11/XF86Config-4, and change: Driver "nv" to "nvidia" since that's what the module filename is. Otherwise it will not update!
"exit" then returned it to init level 5.
Test the resolution changes with ctrl-fn-alt-; and ctrl-fn-alt-p.
Wow, now the screensaver is not sluggish! (right-click on desktop -> configure desktop -> screensaver -> solar wind -> 5 minutes.)
for some reason, some security app changed inittab from 5 to 3. I had to rechange it back. (edit /etc/inittab).
in the same section as Driver nvidia, add on a new line:
Option "IgnoreEDID" "True"
this will allow 1600x1200. Otherwise, you'll be stuck at 1400x1050 with the error in /var/log/XFree86.0.log states: "(WW) NVIDIA(0): Not using mode "1600x1200" (width 1600 is larger than EDID-specified maximum 1588)"
In the Monitors section, add the line "DisplaySize 305 229". This allows apps like amsn and openoffice to show things in the exact size (mm as opposed to pixels).
My /etc/X11/XF86Config-4 for anyone that needs to look at it.

Getting X on TV and external monitor
- After the below modifications to /etc/X11/XF86Config-4, all you simply have to do is hit CTRL-ALT-BACKSPACE (not delete) to restart the X server. If the RCA video out is plugged in, output goes to both devices at 1024x768. If video out is not plugged in, X boots into 1600x1200 on the LCD only.
- In text mode, "toshset -video ext|int|tv|both" works. If I start X and return to text, I experience some flakiness and sometimes I can't see anything unless I run "startx"... I just keep my runlevel at 5.
- Here's a usenet thread arguing that you can change resolutions with hotkeys, but the desktop size remains the same (ie, you view the same desktop through a different size viewport). To actually change the desktop resolution, you need to restart the X server :( I find that this isn't going to matter anyways, because the X server needs to be restarted to detect an external tv or monitor.
- In X, you have to configure XF86Config-4.
- Here are my additions to the /Section "Device"/ Driver "nvidia":
```
    Option      "IgnoreEDID"    "True"
    Option "TwinView"
    Option "SecondMonitorHorizSync"   "30-60"
    Option "SecondMonitorVertRefresh" "50-70"
    Option "TwinViewOrientation"      "Clone"
    Option "MetaModes"                "1024x768,1024x768; 800x600,800x600; 640x480,640x480"
```
- TwinView is ONLY activated if a monitor is detected in the TV (RCA) or EXT (VGA) ports. If NO monitor is detected in either port, then XFree86 will use "Modes "1600x1200" "1400x1050" "1280x1024" "1152x864" "1024x768"" in the Section "Screen". Otherwise, the "MetaMode" overrides it and puts the display in 1024x768 (I'll keep it at a lower res since not all monitors can go beyond 1024. Also, the TV out won't work beyond that).
- You can restart X from the KDE login screen. If you are in 1600x1200, and plug in a TV or monitor, restart X, and it will go into 1024x768.
- According to /usr/share/doc/NVIDIA_GLX-1.0/README, TwinViewOrientation can be "RightOf" "LeftOf" "Above" "Below" "Clone". Clone suits my purposes so I made that default.
- You can also set Option "ConnectedMonitor" to be "tv,crt" or "crt, crt", but that would be stupid to set. Let X detect it automatically.
- I initially had MetaModes set to "1024x768 @1600x1200,1600x1200;". This means that the external monitor would see the 1600x1200 desktop through a 1024x768 window. This is gay if you want full-screen DVD on the TV.
- One more fix. Output will show up on TV and INT, but it doesnt show up on INT if EXT is hooked up. Add "/usr/local/bin/toshset -video both" to the start clause of /etc/init.d/fan which starts up on bootup.
Aesthetics
- K (start menu) -> Configuration -> KDE -> Look and Feel -> Colours -> set to Solaris
- K (start menu) -> Configuration -> KDE -> Look and Feel -> Window Decoration -> Redmond
- K (start menu) -> Configuration -> KDE -> Look and Feel -> Screensaver -> Euphoria
- right click on taskbar -> configure panel -> taskbar -> group similar tasks when... "always"
Applied Mandrake Updates

At the time, there were 73 RPMs available to download. I downloaded them all.
clicked on "Mandrake Control Center" -> software management -> Mandrake Update

Network Timeout - too long

This will solve the delay problem on bootup (if you have nothing plugged into eth0):
- /sbin/ifup: added "DHCP_TIMEOUT=5"
- create /etc/pump.conf, and put: "device eth0 { timeout 5 }" in it.
This will solve the 60-s delay problem when opening windows in KDE:
- edit /etc/hosts, and add your hostname to "127.0.0.1 localhost ". mandrake will take your hostname:0 as your DISPLAY and try to resolve it via /etc/hosts or DNS
- if you use wireless, sometimes, your hostname will change to whatever your given IP# reverses to. I just reset it manually to my hostname by adding this to the end of /etc/sysconfig/network-scripts/ifup-post (but before "exit 0").
- hostname weller (change hostname to your permanent hostname)
- Update: I changed the following lines in /etc/init.d/network:
```
# I inserted "tac" so that it does eth1 first, THEN eth0... but only if
# necessary...
interfaces=`ls ifcfg* | tac| LANG=C egrep -v '(ifcfg-lo|:|rpmsave|rpmorig|rpmnew)' | \

...
# before "Bringing up interface %s:" IF-CLAUSE:
# continue to skip... eth0 can be ignored if eth1 is up...
if [ "$i" = "eth0" ] ; then
 if [ `/sbin/ifconfig eth1|grep -c inet` -gt 0 ] ; then
   continue
 fi
fi
```

Wireless

Wireless worked by default for unencrypted networks. It uses the settings in /etc/sysconfig/network-scripts/ifcfg-eth1 (Managed, any)
For encrypted networks, my script did NOT work (which did in redhat 7.3): "iwconfig eth1 essid CS-WL-1 key 57blablabla; ifup eth1"
I wrote 2 scripts that setup wireless networks that are unencrypted and encrypted: (be sure to save them in /usr/local/bin and "chmod 755 /usr/local/bin/*" after saving them). I basically ended up having to restart pcmcia in /etc/init.d. Doing that to /etc/init.d/network resulted in an unrecoverable situation of "busy devices" which i solved by rebooting.
wireless-unencrypted - this script will setup eth1 for any unencrypted wireless network. It is called by the wireless-encrypted script first, and if it fails, then it will resort to encryption.
wireless-NAME-of-LAN (encrypted) - Create one of these for each encrypted wLAN that you have. Modify the key and essid inside them.
It would be nice to write a script to scan for available networks and automatically select the available network by essid name. Scanning with "iwlist eth1 scanning" would do the trick, but it isnt supported by the orinoco driver. Someone patched the 0.11b orinoco driver, but the current kernel version is 0.13, and i don't want to necessarily downgrade the driver just for that feature.

To the end of /etc/sysconfig/network-scripts/ifup-post (but before "exit 0"), add:

ntpdate bigben.cac.washington.edu
/sbin/hwclock --systohc
hostname weller  (change weller to whatever your hostname is)

One more issue I have - I can't seem to connect to Apple's airport with encryption turned on. I can do it from Windows, but it doesn't seem to work from linux. I'm using the exact same key! If anyone gets it working, let me know! I noticed that in Windows, there were some things set in properties... I'm not sure if any of this matter: network authentication: shared mode. enable network access using IEEE 802.1X: checked; EAP: Protected EAP -> use my smart card.

Iptables (Firewall)

By setting up the firewall the way I want using "iptables", i did: "iptables-save > /etc/sysconfig/iptables"

Here's the contents of iptables:

root@weller /etc/sysconfig 5:17:12pm # more iptables
# Generated by iptables-save v1.2.7a on Mon Jul  7 17:15:53 2003
*filter
:INPUT ACCEPT [323:92927]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [445:29720]
-A INPUT -i lo -j ACCEPT
-A INPUT -s 172.16.228.0/24 -d 0/0 -j ACCEPT
-A INPUT -s 10.0.0.0/24 -d 0/0 -j ACCEPT
-A INPUT -s 137.207.76.3/32 -d 0/0 -j ACCEPT
-A INPUT -p tcp --dport 6881:6889 -j ACCEPT
#-A INPUT -p tcp --dport 22 -j ACCEPT
-A INPUT -p tcp -m tcp --tcp-flags SYN,RST,ACK SYN -j DROP
-A INPUT -p tcp -m tcp --tcp-flags SYN,RST,ACK SYN -j DROP
-A INPUT -p icmp -m icmp --icmp-type 8 -j DROP
-A INPUT -p udp -m udp --dport 53 -j ACCEPT
-A INPUT -p udp -m udp --sport 53 -j ACCEPT
-A INPUT -p udp -s 140.142.16.34 -d 0/0 --dport 123 -j ACCEPT
-A INPUT -p udp -j DROP
COMMIT
# Completed on Mon Jul  7 17:15:53 2003

Note that 172.16.228.0 is a logical network, and is needed for vmware's samba shares. 140.142.16.34 is my NTP server. Port 53 is for DNS. I open things up as I need them.

Iptables (Router)

This is if I connect a cross-over cable from my laptop to another laptop, and I want to give them internet access (via my eth1).
I set my eth0 (my eth1 network is 10.0.0.x): ifconfig eth0 inet 192.168.0.1 netmask 255.255.255.0
The other machine was set to 192.168.0.2, gw: 192.168.0.1, dns: 10.0.0.1 (my router)
I needed to recompile iptables for my kernel.
The reason why I had to recompile iptables was I kept getting this error with MASQUERADE.
bunzip2 -cd iptables-1.2.9.tar.bz2 | tar -xvf -; cd iptables-1.2.9; make ; mv /sbin/iptables /sbin/iptables.old ; make install
/etc/init.d/iptables stop
echo "1" > /proc/sys/net/ipv4/ip_forward
/usr/local/sbin/iptables -A FORWARD -m state --state RELATED,ESTABLISHED -j ACCEPT
/usr/local/sbin/iptables -t nat -A POSTROUTING -o eth1 -j MASQUERADE

Minimizing Drive Noise

See the kernel upgrade (below). that has laptop_mode which is very useful!! Actually, block_mode is very useful in tracking drive activity.
Type "vmstat 1", and you will see under the "bo" column that you'll get a non-0 # and a drive-click. I want to minimize that. Under ext3, it was clicking every 5 seconds (one reason why I wanted to switch to reiserfs). To troubleshoot this, "ls -ltr /var/log", and then "tail -f /var/log/.... &" for the last 4 logs (most busy) to find out cause.
Every 10-20 seconds, I get this in /var/log/XFree86.0.log: (II) PM Event received: Power Status Change... and this in /var/log/messages: CROND[3421]: (mail) CMD (/usr/bin/python -S /var/lib/mailman/cron/qrunner.
"mv /var/spool/cron/mail /var/spool/mail.cron; killall -HUP crond". you can reactivate mailman if you actually plan on using it!
I found the source of the "Power Status Change". It's in the kernel at: /usr/src/linux-2.4.21-0.13mdk/arch/i386/kernel/apm.c. If you look at /proc/apm, you'll notice that the estimated battery life changes from 94-95-94, etc... it's sensitive and everytime it shifts in time remaining, it logs it to the file!!
Lowering the LOG verbosity solved the problem!:
OLD INSTRUCTIONS: "emacs /etc/X11/xdm/Xservers":
:0 local /bin/nice -n -10 /usr/X11R6/bin/X -deferglyphs 16 -logverbose 0
I noticed that if autologin is set to yes, this fix doesn't work. you need to also "emacs /usr/X11R6/bin/startx" and add "-logverbose 0" so that the line looks like:
xinit $clientargs -- $serverargs -deferglyphs 16 -logverbose 0 &
now the drive only clicks every 60 seconds :)
NEW INSTRUCTIONS: Instead of lowering the log verbosity, I added this line to the "ServerFlags" section of XF86Config-4:
Option "NoPM" "True"
At 4am, my drive would go nuts. You know, I really could care less about the daily stuff. I just went:
cd /etc; mv cron.daily cron.junk

MP3 Player USB drive

mkdir /mnt/removable
mknod /dev/sda1 b 8 1
modprobe usb-storage
mount /dev/sda1 /mnt/removable

Power Suspend/Resume (I don't use this anymore)

"chmod 4755 /usr/bin/apm" to SUID the power management so that normal users can run it.
right-click on blue power icon in system tray. select "configure klaptop".
Click "Battery" tab. Set "Not Powered" to "Suspend"... wait for "5 min" (Standby does not work)
Set "Powered" to "Suspend"... "20 minutes".
You can manually suspend at the unix prompt with "apm -s". Turning on will resume from where you left off. You might have to "CTRL-ALT-F7" to get back to the GUI. It's no big deal.
Actually, on second thought, I turned this off. the suspend feature is really flaky. it freezes 60% of the time and i end up doing a normal boot. If it does work, i usually lose the network, sound occasionally. sometimes eth0 and eth1 switch on me. very weird.

Time

Somehow my hw clock got set to jan 2000. two things will fix this. ntpdate will synchronize the OS clock, and then hwclock can set the hw clock based on the system time.
to end of: "/etc/sysconfig/network-scripts/ifup-post":
ntpdate bigben.cac.washington.edu ; /sbin/hwclock --systohc

Toshiba Utilities

First, I needed: xpm-devel
rpm -ivh xpm-devel-3.4k-2.i386.rpm
./configure
make depend
make
make install
Added "toshiba" to /etc/modules
modprobe toshiba
cd src; cp alarm dispswitch fan fnfind hotkey ownerstring svpw tbacklight tdocked thotswap tpasswd tuxtime-conf wmtuxtime /usr/local/bin/
I went through all of these, and the only one that seems "ok" is "fan"!

Toshiba Command-line utility
- Download toshset
- Install:
```
./configure
make
make install
```
- I got a core dump! Downloading the binary instead works fine.
- Here's the output of toshset -q"
- I've tried these, and they work: "toshset -lcd [bright,semi,super]", "toshset -d [turn off display in # min}
- "toshset -fan [off,low,high]", "toshset -video [int,ext,both,tv]" (i havent hooked up an extra monitor yet)
- "toshset -cpu [slow,fast]" - /proc/cpuinfo shows no change in MHz, so i'd have to run a benchmark to verify...
- I wrote a little script to automatically toggle between the 3 brightness levels, and created a shortcut on my desktop for /usr/local/bin/lcdtoggle
- chmod 755 lcdtoggle
- chown root toshset
- chmod 4755 toshset (in order to run as a normal user!)
- toshset -d 20 (set lcd auto-blank-screen to 20 minutes - note that this goes by IDLE time of the keyboard! i had it set to 3 minutes, but i noticed it would turn off when i'm just surfing the web using the mouse without touching the keyboard).
- toshset -cpu fast (use 1.7GHz. I tried slow, and it runs 2.5x slower (~ 700 MHz), but /proc/cpuinfo will still show 1695MHz).
Fan Setup
- Installed tosh-utilities (above)
- cd /etc/rc5.d
- ln -s /etc/init.d/fan S96fan
- cd /etc/rc3.d
- ln -s /etc/init.d/fan S96fan
- created /etc/init.d/fan
- Test with "/etc/init.d/fan start". The fan is on when AC power is there. If I switch to battery, the fan turns off, and resumes when I plug it back in again.
Modem

I haven't tested this yet, but my swiss friend managed to get his modem working:

> Grab slmdm-2.7.10.tar.gz (no newer version, they're buggy), compile and
> install it (you might need to create some dirs), and then let loose. I've
> made the modem work that way, I'm getting 52k connects.

Fonts

openoffice bullets would show up as 0's instead of arrows. It's the lack of certain windows fonts (I think wingdings).
Used sourceforge as a guide.

downloaded cabextract utility and msttcorefonts.spec.

emacs msttcorefonts-1.3-4.spec
 (changed path: < %define ttmkfdir /usr/sbin/ttmkfdir )
rpm -ivh cabextract-0.6-1.i386.rpm
rpmbuild -bb msttcorefonts-1.3-4.spec
rpm -ivh /usr/src/RPM/RPMS/noarch/msttcorefonts-1.3-4.noarch.rpm
rpm -ql msttcorefonts  (to check)
/etc/init.d/xfs restart
xlsfonts | grep ^-microsoft'  (to check)

followed these instructions to get openoffice to see the new fonts.
/usr/lib/openoffice/spadmin -> fonts -> add -> /usr/share/fonts/msttcorefonts
from openoffice: tools->options -> openoffice.org->fonts -> ?

Applications

Mplayer

followed instructions for "COMPILING MPLAYER WITH QUICKTIME LIBRARIES SUPPORT"
unpacked them into /usr/lib/win32, and put them in the root of that directory
Downloaded cvs mplayer
./configure ; make; make install
"mplayer -ao arts" should work, but i need to "artsdsp mplayer" instead.
sometimes, it doesnt work the first time, but the second execution worked. weird.
fyi, use "mplayer -dvd 1" for DVDs and "mplayer -vcd 1" for VCDs.
NOTE: when I watch movies, I must do this:
"/etc/init.d/laptop stop" (to turn off laptop_mode - it caches so far in advance, that the drive ends up sleeping. when it has to spin up again, the movie will pause for half a second every minute (very annoying!)
Also, "toshset -d 0" to turn off bios-screensaver.
"toshset -fan high" since the CD heats the laptop up quite a bit (it's better to be safe).
I noticed that mplayer SOUND doesnt work all of the time. When it does work, however, I STILL get the error: "audio_setup: Can't open audio device /dev/dsp: Device or resource busy", followed by "AO: [esd] 44100Hz 2ch Signed 16-bit (Little-Endian) (2 bps)".

My process list looks like:

weller   10419     1  0 00:36 ?        00:00:21 /usr/bin/artsd -F 10 -S 4096 -s 60 -m artsmessage -l 3 -f
weller   11595     1  0 00:49 ?        00:00:07 /usr/bin/esd -terminate -nobeeps -as 2 -spawnfd 31

One more thing i noticed - when I do FULL SCREEN 1600x1200, the AUDIO becomes out of sync with the video quite often. This doesnt seem to happen at 1024x768. When I play a movie on my TV, I need to be in 1024x768 mode anyways. What I end up doing is hooking up the RCA out to the TV, and then hit ctrl-alt-BACKSPACE to restart X in 1024x768 (it detects the dual out and drops in resolution).

I haven't quite figured this out yet, but sound goes through artsd which holds onto /dev/dsp... if that fails, it tries esd... then it that fails, it doesnt work at all. I've documented all 3 cases:

DOESNT WORK for arts OR esd: (repeat mplayer command several times and it works)
audio_setup: Can't open audio device /dev/dsp: Device or resource busy
/dev/dsp: Device or resource busy
AO: [esd] esd_open_sound failed: Connection refused
AO: [null] 44100Hz 2ch Signed 16-bit (Little-Endian) (2 bps)

WORKS via artsd:
AF_pre: 44100Hz 2ch Signed 16-bit (Little-Endian)
AO: [oss] 48000Hz 2ch Signed 16-bit (Little-Endian) (2 bps)
Building audio filter chain for 44100Hz/2ch/16bit -> 48000Hz/2ch/16bit...

WORKS  via esd, after artsd fails:
Can't open audio device /dev/dsp: Device or resource busy
AO: [esd] 44100Hz 2ch Signed 16-bit (Little-Endian) (2 bps)

I have no idea what spawned esd. it's orphaned and owned by the kernel.

LICQ

copied over my old ~/.licq directory
modified licq.conf, and changed: "Plugin1 = qt-gui"
Menu -> Options -> OnEvent -> Command -> "artsdsp play"
options -> preferences -> Output Plugin: aRts Driver 0.4 [libartsout.so]

Java

J2SDK 1.4.1.03-linux-i586-rpm.bin
h j2sdk-1_4_1_03-linux-i586-rpm.bin
rpm -ivh j2sdk-1_4_1_03-fcs-linux-i586.rpm
cd /etc/alternatives/
mkdir old; mv jar java javac javadoc javakey javap old
ln -s /usr/java/j2sdk1.4.1_03/bin/jar .
ln -s /usr/java/j2sdk1.4.1_03/bin/java .
ln -s /usr/java/j2sdk1.4.1_03/bin/javac .
ln -s /usr/java/j2sdk1.4.1_03/bin/javadoc .
ln -s /usr/java/j2sdk1.4.1_03/bin/javap .

OpenOffice

downloaded
su -
tar -xzvf OOo_1.1beta2_LinuxIntel_install.tar.gz
cd OOo_1.1beta2_LinuxIntel_install
sh ./install
cd /usr/local/bin
ln -s /usr/local/OpenOffice.org1.1Beta2/program/soffice .
as normal user: soffice
it complains. i need JRE1.3+. i installed it (see above)
GUI install: local install -> use ~/.soffice
-> use installed JRE -> browse to /usr/java/j2sdk1.4*/jre

Mozilla

Mozilla 1.3.1 had some annoying bug. If I go to www.di.fm and use SHIFT-scroll-wheel to change font size, it crashes.
downloaded mozilla 1.4 RC.
tar -xzvf mozilla-i686-pc-linux-gnu-1.4rc1-sea.tar.gz
cd mozilla-installer/
./mozilla-installer (installed to /usr/local/mozilla)
cd /usr/bin; mv mozilla mozilla.old
ln -s /usr/local/mozilla/mozilla .
cd /usr/local/mozilla/plugins
ln -s /usr/java/j2sdk1.4.1_03/jre/plugin/i386/ns610/libjavaplugin_oji.so .
downloaded Flash
tar -xzvf install_flash_player_6_linux.tar.gz
cd install_flash_player_6_linux ; ./flashplayer-installer
installed to /usr/local/mozilla
NOTE: sometimes mozilla or some apps FREEZE up. they just stop responding to mouse-clicks. I found out what it was! xfs crashes a lot! just restart with "/etc/init.d/xfs start". then problem goes away. I think i'll manually upgrade xfs some day.

Alvaro's MSN

I didn't like GAIM, so I won't even post the instructions.
Alvaro's MSN
untarred it to /usr/local/msn, and made a link:
ln -s /usr/local/msn/amsn /usr/local/bin
I thought I'd need to poke a hole in iptables to be able to receive files through MSN, but it's the sender's end who has to poke the hole. I was iptabled securely, and I connected to port 6891 on the other end.
I get a lot of these messages: "You are running a version of messenger that requires an immediate security update. Please visit..."
The latest AMSN fixed that problem. It installed to ~/.amsn/plugins/tls1.4. I moved that to /usr/local/msn/plugins for the sake of globality.

Apache

modified /etc/httpd/conf/commonhttpd.conf -> changed "Options" under directory /home/*/public_html to "Options All" (i needed CGI)
/etc/init.d/httpd restart

Real Player

Downloaded Realplayer for Linux 2.x (libc6 i386) RPM
"mv rp8_linux20_libc6_i386_cs2_rpm rp8.linux20.libc6.i386.cs2.rpm"
"rpm -i rp8.linux20.libc6.i386.cs2.rpm"
"cd /usr/lib/mozilla-1.3.1/plugins/"
"ln -s /usr/local/netscape/plugins/rpnp.so ."
Despite this, I still can't get realplayer to work as a plugin (tested with mtv.com) - about:plugins lists audio/x-pn-realaudio-plugin
Since realplay is NOT artsd aware, you have to run "artsd realplay". I cannot figure out how to get mime-types/mailcap to call up "artsd" instead of realplay. It ignores my changes. I tried making a front-end, but that failed. If you just have "arstdsp realplay" running already, then mozilla will USE that app.

Wine + Kazaa + Nero

I got WINE working with kazaa and nero! (For kazaa, see my old redhat instructions
I am using the latest CVS for wine. Very quick instructions: "setenv CVSROOT :pserver:cvs@cvs.winehq.com:/home/wine ; cvs login; (cvs=pass); cvs -z 0 checkout wine; cd wine; ./configure; make depend; make; make install"
I used these instructions, which works for nero551042.exe
When I ran "wine nero551042.exe", it "hung" on installation. I had to WAIT for the drive activity to stop before hitting CTRL-C (important!). Ensure that nero entries were added to the wine registry (try "grep -i wine ~/.wine/*.reg"). Move the expanded directory to ~/.wine/c/Program Files/Ahead. If you don't wait or dont have nero entries in the registry, when you try to add audio files, you get "Could not find nero plugin manager. Please reinstall nero".
There were 2 other minor problems. NERO will not detect the cdburner, but it will if you run "cdrecord -scanbus" first. I added this to my .tcshrc:
alias nero 'cdrecord -scanbus; artsdsp wine ~/.wine/c/Program\ Files/Ahead/Nero/nero.exe'
Another problem is I had to add "artsdsp" in front of wine to get sound to work. artsd ties up the sound device, so you have to force rewrites. I noticed that with some games, artsdsp SLOWS down the whole program with sound. The proper way is to edit ~/.wine/config, and change the WinMM section:
```
[WinMM]
#"Drivers" = "wineoss.drv"
"Drivers" = "winearts.drv"
```
Unfortunately, I get an error: "err:module:BUILTIN_LoadModule loaded .so but dll winearts.drv still not found". Until they fix that bug, I need to stick with artsdsp in front of wine.
There's another problem that I havent resolved. I can't drag and drop from the file browser onto my CD layout. I have to right-click and go "Add file" to setup my CD layout.

VMWare

installed VMWare workstation 4
rpm -ivh VMware-workstation-4.0.1-5289.i386.rpm
this adds vmnet and vmmon kernel modules to /lib/modules.
/etc/init.d/vmware loads these, and starts vmware-smb, which basically lets you see the linux drive from windows by sharing via samba.
a virtual interface is added: ifconfig vmnet1: 172.16.228.1. vmware-smb is running on this interface.
had to add to /etc/sysconfig/iptables to allow samba to work:
-A INPUT -s 172.16.228.0/24 -d 0/0 -j ACCEPT
Basically, you can run "vmware" and kill artsd, or run "artsdsp vmware" to filter it. Either way, my sound is distorted. My MP3s play like chipmunks. Let me know if you know what's wrong.
I added 2 OS's: windows xp and windows 98se. you need the install CDs.
windows xp takes 2 minutes to load, and it created a 1.6gig file; windows 98 takes 20 seconds and takes only 470 meg.
UPDATE: I added 512M RAM in my machine, bringing it to 768M. I up'd windows XP physical RAM to 384M. Windows XP loads in about 50 seconds now, which is a LOT more tolerable!
when you create the hardware for these OS's, set the following: floppy: UNCHECK start on boot; sound: /dev/dsp; network: use NAT
once OS is installed, in VMWare, go: file -> install VMWare tools. This will load the OS drivers for the fake hardware. Unfortunately, for windows 98, you have to go to creative labs and download the ES 1370 drivers. maybe I got the wrong one. It says to get Ensoniq. I tried 1370 and 1371. Both staticy.
I had to run "smbpasswd -c /etc/vmware/vmnet1/smb/smb.conf weller" to set my samba password to use within Windows 98. In windows xp, you dont need this. you can simply right-click on vmware icon, open it, and go to shared folders (if you edit OS properties inside VMware and go options, you can set what Linux folders you want shared). The default shares setup with vmware-samba are hostfs and your_username. I mapped S: to / and Z: to ~, and saved these mappings.
As far as sound goes, I tried adding sound.maxLength and sound.smallBlockSize to 1000000 and 2048 respectively (set in ~/win98/win98.vmx). Still choppy. reducing these to values below 512 didn't help either.
I tried the generic MS sound driver. it says it cant find it.
I noticed that if you go into network properties and change it from "network login" to "windows login", the bootup time is reduced from 40 seconds to 20 seconds for windows 98. Also, you don't need a password to login.

Here is a clarification that I sent to a friend... I'll share it with you, in case you had the same problems that he did:

Hmmm... i have no idea how shorewall works.  iptables is very
low-level and built into the kernel.  maybe shorewall is a user-friendly
front-end for iptables or ipchains.  i feel like i have more control
by using the kernel-level firewall.

just try this and see if you get entries:
root@weller /home/weller/public_html 5:16:58pm #  iptables -nL
Chain INPUT (policy ACCEPT)
target     prot opt source               destination
ACCEPT     all  --  0.0.0.0/0            0.0.0.0/0
ACCEPT     all  --  172.16.228.0/24      0.0.0.0/0
ACCEPT     all  --  10.0.0.0/24          0.0.0.0/0
DROP       tcp  --  0.0.0.0/0            0.0.0.0/0          tcp flags:0x16/0x0
2
DROP       tcp  --  0.0.0.0/0            0.0.0.0/0          tcp flags:0x16/0x0
2
DROP       icmp --  0.0.0.0/0            0.0.0.0/0          icmp type 8
ACCEPT     udp  --  0.0.0.0/0            0.0.0.0/0          udp dpt:53
ACCEPT     udp  --  0.0.0.0/0            0.0.0.0/0          udp spt:53
ACCEPT     udp  --  140.142.16.34        0.0.0.0/0          udp dpt:123
DROP       udp  --  0.0.0.0/0            0.0.0.0/0

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination
root@weller /home/weller/public_html 8:33:21pm # ipchains -nL
ipchains: Command not found.

so, depending on your system, it could be ipchains or iptables.

you can try temporarily turning off your firewall to see if
things work without it and then you know where your problem lies.
with iptables, it's /etc/init.d/iptables stop
then when you do a "iptables -nL", you don't see anything.

if you type:
ifconfig vmnet1
 do you get this:
vmnet1    Link encap:Ethernet  HWaddr 00:50:56:C0:00:01
          inet addr:172.16.228.1  Bcast:172.16.228.255  Mask:255.255.255.0
etc...?

from windows, try: start -> logoff user -> then log back in as
"dave" or whatever you username is in linux.

login with the same pass that you set here:
 smbpasswd -c /etc/vmware/vmnet1/smb/smb.conf -a dave

(replace dave with whatever you use)

then try:
start -> run -> \\172.16.226.1\
 then right-click on a folder and go: map network drive. map it to
something like Z:

VMWare - UPDATE + running the same XP off your partition

Ok, I'll first tell you the MINIMUM needed to get this working... then, I'll fill you in with my problems!!! Here is my story on getting VMWARE to BOOT the Windows XP that is already on your harddrive.
Upgraded VMWare workstation 4
rpm -Uvh VMware-workstation-4.0.5-6030.i386.rpm (all defaults were selected. I did have to point it to my kenerl build though - at /usr/src/linux-2.4.21mw10/include (which will be different for you))
Boot into Windows. control panel->system->hardware->hardware profiles->COPY (profile) -> call it VMWARE.
reboot. select VMWARE profile. go: device manager -> IDE controllers -> right-click IDE controller -> update driver -> from a list -> dont search -> select "Standard Dual Channel PCI IDE Controller". without this selected, if you boot into windows, you'll get a blue-screen of death, even going into safe mode.
Do not upgrade to service pack 1!!! if you do that, your CD burner will not work in VMWARE. I kept getting "insert a writable disc. this disc isn't writable". I really regretting having SP1. I didn't archive my files :(
go into vmware (as root - i need to access all of /dev/hda). Go: New Virtual Machine -> custom -> windows XP Prof -> 384M -> Use NAT -> Use physical disk -> use entire disk (NOTE: Yes, I know we only need /dev/hda1. If you select just hda1, you will get L 99 99 99 99 instead of LILO because the 2nd stage boot loader is not on hda1!!!)
floppy drive -> deselect "connect at power on" (vmware will fail because we dont have a floppy drive)
options - > shared folders -> add -> root -> / ... -> add -> weller -> /home/weller (in windows XP, you dont need samba for sharing. the vmware tools enable this)
BOOT into windows. select File -> Install VMWARE Tools. This installs VMWARE drivers for your display adapter and allows shared folders.
Edit /etc/lilo.conf. Comment OUT "timeout=" with a #. Also, set "default=windows". Then type "lilo". This is pretty much ALL you need to be up and running! I'll explain why you want that lilo stuff below.
Below is just a story of my woes. Stop here if you want.
When I added windows XP and booted it, I got a phone call. I left the computer, and guess what? My LILO default was set to Linux and it tried booting linux. Guess what happened? Never ever ever mount the same filesystem twice! You'll get major inconsistencies (kind of defeats the purpose of mutex locks on writing to files). Anyways, linux would NOT boot anymore, not even to single user mode. I had to boot off CD, type "linux rescue", then reiserfsck /dev/hda1. Phew.
That is why windows is my new default!! The reason why I turned off the timeout, was because most of the times, I want linux... Secondly, I started vmware once, and I walked away. I forgot that I WANT the VMWARE profile! The last thing you want is to go into the native profile and have it start uninstalling shit on you.
When I booted into windows via VMware, it wanted me to reactivate Windows. The stupid key on the bottom of my machine didn't work. I had to call microsoft. I read them a bunch of numbers and then they gave me a bunch of numbers. They asked me what changed on my computer. I told them I was running vmware in linux, so all my hardware is fake. The micropsoft rep was probably like "whatever". Then, when I booted into Windows natively, and it wanted me to reactivate again! I thought, "screw this". I'm going to get a crack so it won't make me activate everytime i go back and forth. I applied the crack, and guess what? I couldnt boot into safe mode or normal mode!
I booted off my windows XP boot CD (from vmware), and hit "r" for recovery console. Then it asked for Administrator's password. Arrgh. I didn't know it!
So, I figured I'd risk the risky NTFS-write mode from linux. All I wanted to do was "copy c:\windows\winlogon.bak c:\windows\winlogon.exe". Well, "mount -o rw /dev/hda1 /mnt/windows" didn't work. I got permission denied.
I got the latest kernel - 2.4.23, and patched it for the rewrite of NTFS 3 (which is in the 2.5 kernel). Still no go! Apparently, write-support only exists for copying files onto a file that is EXACTLY the same size!! You can't add new files or copy onto a file that is bigger or smaller. Not to mention, you have to run "ntfsfix /dev/hda1" after any changes because it creates an inconsistency (the kernel docs say this).
After all this work, it turns out that booting off the Windows 2000 CD and going into recovery console does not ask for Administrator's password! So, problem was finally solved and i was able to boot back into windows again.
My next problem is that I couldnt get my CD burner to work! But it used to work. One of the differences between the windows xp that worked and the one that didnt was that the native partition has SP1 installed. So, I tried re-installing Windows XP since I can't uninstall SP1.
Guess what? The Windows XP corporate CD will NOT BOOT on the laptop! I assume there's some proprietary toshiba driver needed... so, I installed windows XP from vmware (since it's fake hardware, not toshiba hardware). Voila. worked like a charm... kind of :(
Now my laptop wont even boot natively into windows :( Not even into safe mode :( I ran windows in LOGGED mode, and checked it out in linux in /mnt/windows/WINDOWS/ntbtlog.txt. The last thing it got stuck on was "Loaded driver agp440.sys Did not load driver Advanced Configuration and Power Interface (ACPI) PC"
So, I got the toshiba rescue disks out. Arrgh. it wanted me to wipe out the whole disk!!! I checked out the disk and saw that there was a .GHO file on there. Aha. I figured I'd install norton ghost and just ghost explorer it.
What more can go wrong? It was password protected!! Here is how to get that. It's 2559 for the 6100.
After peaking into the GHOST image (it made me stick in CDs 4 and 5 since it spanned 3 CDs), I extracted the following files: (I got the hint of what files to get by going into driver details on "Computer -> ACPI Uniprocessor:" and "IDE Controllers" and "CDROM drive":
windows\system32: hal.dll ntkrnlpa.exe ntoskrnl.exe
windows\system32\drivers: atapi.sys acpi.sys acpien.sys cdrom.sys agp440.sys
I backed up the original files (just in case), and overwrote them from windows in VMWARE. I then rebooted and it worked! No more black screen hang! It actually boots.

Sniffer
- Downloaded tcpdump.tar.gz and libpcap.tar.gz
- Downloaded snort.tar.gz
- ran this on all 3: ./configure; make; make install (did the libpcap first)
- to sniff a particular connection, I simple found the port with: "lsof|grep -i wish|grep EST" (to find that msn was on 1863), then went: "tcpdump -i eth1 -X port 1863"
BitTorrent and Python

Now I remember why I hate RPMs! I got circular dependencies... I need libPython to install python. i need python to install libpython. gay.
I gave up with the "--force --nodeps". it just keeps giving me a longer and longer list. screw that.
downloaded python.tar.gz, and wxPython
ran "./configure; make; mv /usr/bin/python /usr/bin/python.old; make install; rehash; python -V" for python.
for wxPython, the standard "./configure; make; make install" didn't quite work for me. I had to actually read the docs this time, since I'd get ImportError: No module named wxPython.wx.

Steps:

  1134  13:08   setenv WXPREF /usr/local/lib/wxPython
  1135  13:08   mkdir build
  1136  13:08   cd build
  1137  13:08   ../configure --with-gtk --prefix=$WXPREF --enable-rpath=$WXPREF/lib --with-opengl --enable-geometry --enable-optimise --enable-debug_flag
  1146  13:19   cd ..
  1147  13:20   cd wxPython
  1148  13:20   python setup.py WX_CONFIG=$WXPREF/bin/wx-config build install

the last line installed wx and wxPython into /usr/local/lib/python2.3/site-packages. these have a __init__.py file, which make it a package. So, when a .py file calls "from wxPython.wx import *", it looks in site-packages for wxPython/wx.py.

downloaded bit torrent (.tar.gz)... also got bit torrent ++.
oops. apparently, I need wxwindows.
bunzip2 -dc wxGTK-2.4.2.tar.bz2 | tar -xvf - ; cd wxGTK-2.4.2; ./configure ; make; make install
I give up with BT++. The toolbar is missing.

For BitTorrent:

mv BitTorrent /usr/local/lib/python2.3/site-packages/
cp bt*.py /usr/local/bin

You can download the text way with: btdownloadcurses.py file.torrent
In Mozilla: edit->pref->helper apps -> add new -> application/x-bittorrent, extension: torrent, execute: /usr/local/bin/btdownloadgui.py (do not add anything else)
Try browsing supernova for bittorrents. Mozilla will automatically call up the external app when you dload a torrent.
You need to open up port 6881-6889. /etc/sysconfig/iptables:
```
-A INPUT -p tcp --dport 6881:6889 -j ACCEPT
```
then /etc/init.d/iptables restart
I guess there are some dumb trackers that use your internal IP instead of external IP, so according to this article, you should launch with btdownloadcurses.py --ip EXTERNALIP#.

Mozilla Firebird
- cd /usr/local; tar -xzvf MozillaFirebird-0.7-i686-pc-linux-gnu.tar.gz
- cd bin; ln -s ../MozillaFirebird/MozillaFirebird .
- downloaded mimeTypes.rdf and copied it:
- cd /usr/local/MozillaFirebird/defaults/profile; mv mimeTypes.rdf mimeTypes.rdf.orig; cd US; rm mimeTypes.rdf; ln -s ../mimeTypes.rdf .
- rm ~/.phoenix (it copies the mimeTypes over)
- well, i cant get the mplayer-in.so to work, so I'm going back to mozilla 1.6a

Mozilla Upgrade

Downloaded Mplayer plugin

  1503  16:15   mkdir mplayerin
  1504  16:15   cd mplayerin/
  1505  16:15   cvs -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/mplayerplug-in login
  1506  16:15   cvs -z3 -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/mplayerplug-in co mplayerplug-in
  1507  16:16   ls
  1508  16:16   cd mplayerplug-in/
  1509  16:16   ls
  1510  16:16   ./configure
  1511  16:16   make
  1512  16:16   cp mplayerplug-in.so /usr/local/MozillaFirebird/plugins/

  1587  16:55   cd /usr/local
  1588  16:55   mv mozilla mozilla1.4
  1581  16:54   tar -xzvf mozilla-i686-pc-linux-gnu-1.6a-sea.tar.gz
  1583  16:54   cd mozilla-installer
  1585  16:54   ./mozilla-installer
  1587  16:55   cd /usr/local/mozilla/plugins
  1595  17:09   mv ../../mozilla1.4/plugins/libjavaplugin_oji.so .
  1596  17:09   mv ../../mozilla1.4/plugins/flashplayer.xpt .
  1597  17:09   mv ../../mozilla1.4/plugins/libflashplayer.so .
  1600  17:12   mv ../../mozilla1.4/plugins/raclass.zip .
  1601  17:12   mv ../../mozilla1.4/plugins/rpnp.so .
  1593  16:56   cp /root/inst/mplayerin/mplayerplug-in/mplayerplug-in.so .

Over-Heating Problem!

After an electrical storm, my laptop crashed a couple times during a kernel compilation. It freaked me out. It turns out that it IS consistent... it crashes in windows AND in single-user mode! I believe it's the CPU fan that's blown. The outer fan runs just fine.
Here was a temporary fix. I turned the fan (outer) on high, and the cpu speed on slow. I disabled the fan daemon so that it's on permanently: /etc/init.d/fan now shows:
```
#       daemon /usr/local/bin/fan -d
/usr/local/bin/toshset -cpu slow
/usr/local/bin/toshset -fan high
```
I'm not sure what part I need, but the only component listed related to fans is Cooling Module for $72. Does anyone know how the cooling system works? is there even a cpu fan?
sigh. after getting it to work, I noticed that my X stopped working :( a temporary fix was to change the KDM to GDM, by changing both lines in /etc/sysconfig/desktop from KDE to GNOME
Afterward, I've discovered that there are missing files on my system!!! (maybe it was lost through fsck, improper buffer flushes??)

I wrote a script to compare the checksum on the RPM files:

#!/bin/csh -f
set list=(`rpm -qa|egrep -i "gcc|kd|XFree"|sort -u`)
 @ count = 0
while ($count < $#list)
 @ count++
 echo ""
 echo $list[$count]
 rpm -V $list[$count]
end

I scanned kd and xfree packages, and sure enough, there were a lot of missing files. I fetched the new rpm's and forced replacement of package with: "rpm --replacefiles --replacepkgs -ivh ~weller/kdebase-3.1-83.3mdk.i586.rpm"
Well, that didnt solve my problem. It appears that /etc/X11/prefdm loads up mdkkdm as my login manager. It is unable to execute /etc/X11/Xsession which calls up "startkde". I just get an xterm window and I have to manually type "startkde". Does anyone know why this stopped working?
Well, I sent my laptop back to Toshiba Support (in Memphis). They'll take it back if your warranty expired less than a month ago. They replaced my main system board and power supply.

Swap problem

there seems to be a bug with the 2.4.21 kernel. The system uses the RAM as CACHE, as seen with "free". When I use "display" as a slideshow, it caches all the pictures in RAM. Instead of "freeing" cache for RAM, it starts to use SWAP space:

weller@weller /home/weller 4:02:47pm > free
             total       used       free     shared    buffers     cached
Mem:        774816     751564      23252          0       4908     643768
-/+ buffers/cache:     102888     671928
Swap:       361420     361420          0

Then there's 0 SWAP left! i ran display on tons of files in a directory and monitored it with "free". the "cached" column fluctuated from 637M to 640M. it would do something like: 637->638->639->640->637 and so forth.
But then... I ran vmware, and the system crawled. the cache went up to 699! Then I had tons of instability... random programs started closing and then the cache dropped to 600M.
so basically, it "freed" the cache, but there must be a bug, since it freed memory that was in use. i lost a lot of running programs, including the kde toolbar! i was able to ctrl-alt-f1 and get a text prompt to see what was going on. that's how i discovered the cache was only 600M.

To test reproduce my problem, I used my aliases to slideshow a directory of JPG's:

alias disp4 '\ls -1 *.jpg|xargs display -delay 100'
alias disp2 find . -maxdepth 1 -type f -name \'\*.jpg\'  -exec display -delay 10 {} \\\;

DVD-R read problem

I had problems with reading some DVD-R's. I posted this to a forum and followed up on it:
Hi, I bought a spindle of 100 DVD-Rs for $60, and I can read them fine on most DVD drives, except the DVD/CDRW combo drive that came with my Satellite Pro 6100 :(
I can read DVD+R's just fine. Do you think it's the specific brand of DVD-R that it doesn't like? The reason why I think this is because the info below shows that it supports DVD-R Reading.
I don't see any firmware upgrades for UDJA720, except for ones that come with IBM notebooks.

Here is the info that cdvdinfo reports:

DRIVE VENDOR : MATSHITA
DRIVE PROID : UJDA720 DVD/CDRW
DRIVE REVISION : 1.00
VENDOR SPECIFIC : 11009tKME12040292311
INTERFACE : ATAPI
RPC2 SCHEME : NO
CDR READ SUPPORT : YES
CDRW READ SUPPORT : YES
CDRM2 READ SUPPORT : YES
CDR WRITE SUPPORT : YES
CDRW WRITE SUPPORT : YES
DVD-ROM READ SUPPORT : YES
DVD-R READ SUPPORT : YES
DVD-RAM READ SUPPORT : YES
DVD-R WRITE SUPPORT : NO
DVD-RAM WRITE SUPPORT : NO
MAX CDR READ SPEED : 24
MAX CDR WRITE SPEED : 8

I figured it out:
I got it working. I don't know what fixed it because I did 2 things:
1) I upgraded the firmware on my IO Magic DVD-writer from 2.15 to 2.30.
2) I used Nero to burn a dual ISO/UDF disc and burned at 1x (just to be safe).
I was then able to read the DVD-R just fine in Windows and Linux (I had to compile the UDF module for my linux kernel) from my Matshita UDJA720 drive.
Also, if I boot with the DVD-R already inserted, I have to eject and then reinsert it. Otherwise, the directory comes up empty.

Kernel Upgrade

I tried 2.4.21 and applied 2.4.22-pre10 patch to it, but decided not to, otherwise the laptop patch for 2.4.21 (=rc8) would have too many mismatched lines.
fetched kernel source from kernel.org
downloaded laptop patch for 2.4.21
here's my kernel config. I'd rather not use mandrakes since there's too many unnecessary things to compile. I had to recompile the kernel 8 times before I got it right. I kept forgetting to include things! :( Your best bet is to type "lsmod > somefile", and check to make sure that those all exist. Some important things are usb-core (mouse-wheel wasnt working with usbmouse); orinoco_cs for wireless (even though i had orinoco) and cdrom (for cdrom - it maps to /dev/cdrom0 instead of /dev/cdrom); apm to allow OS to turn off power after shutdown... the only additional drivers needed looking after were vmware's vmnet and vmmon and nvidia's nvidia.o. reiserfs has to be compiled into the kernel because the laptop patch broke the module. I got unresolved symbols in /lib/reiserfs.o

basic steps:

cd /usr/src
bunzip2 -cd linux-2.4.21.tar.bz2|tar -xvf -
ln -s linux-2.4.21 linux
ln -s linux-2.4.21 linux-2.4.21-rc8
ln -s linux-2.4.21 linux-2.4.21-rc8-laptop
cat laptopkernel-2.4.21-rc8-laptop1.diff|patch -p0

I had to fix up a few things in the source to get it to work. here's the diff:

36      ./drivers/char/drm/i810_dma.c
< #define DO_MUNMAP(m, a, l)    do_munmap(m, a, l)
---
> #define DO_MUNMAP(m, a, l)    do_munmap(m, a, l, 1)

148     ./drivers/char/agp/agpgart_be.c
>       { PCI_DEVICE_ID_VIA_8375,
>               PCI_VENDOR_ID_VIA,
>               VIA_APOLLO_KM266,
>               "Via",
>               "Apollo Pro KM266 / KL266",
>               via_generic_setup },
5119a5126,5131
>       { PCI_DEVICE_ID_VIA_P4M266,
>               PCI_VENDOR_ID_VIA,
>               VIA_APOLLO_P4M266,
>               "Via",
>               "Apollo P4M266",
>               via_generic_setup },

Additional steps:

make mrproper clean xconfig
 load config file -> use kernel config
emacs Makefile
  (change EXTRAVERSION to mw8)
make dep bzImage modules
make modules_install
make install
cd /root/inst
sh ./NVIDIA-Linux-x86-1.0-4363.run
 (this COMPILES nvidia.o to present kernel! if you boot into the new kernel,
  you can do this in text mode)
vmware-config.pl
  (this COMPILE vmnet.o and vmmon.o for vmware)

I don't know why, but I had to link /dev/cdrom to /dev/scd0. it was pointing at hdc, which no longer worked.
I want to try this script which can put the system into laptop mode by modifying /proc/sys/vm/bdflush and /proc/sys/vm/laptop_mode.
I saved the script as /etc/init.d/laptop
cd /etc/rc5.d; ln -s ../init.d/laptop S99laptop
I ran "/etc/init.d/laptop start" and did "vmstat 1". Instead of using "toshset -hdd 1" for 1 minute spindown, use "hdparm -S6 /dev/hda" (for 6x5=30 seconds).
This is cool. You can monitor DRIVE activity to find out the block and paritions being written to whenever there's disk activity. Also, you can determine what file is READING from the drive.

Do this:

/etc/init.d/syslog stop; chown root /dev/console; ( xconsole & ) ; klogd -f /dev/console; echo 1 > /proc/sys/vm/block_dump

To bring things back to normal:

 (echo "0" > /proc/sys/vm/block_dump ) ; /etc/init.d/syslog stop ; killall xconsole ;/etc/init.d/syslog start

Actually, I wrote a script and stored it in /etc/init.d/diskspy

I must admit, this is amazing to be able to sit there and hear drive activity and know what caused it. For instance, here's some output that told me licq was writing to /dev/hda8 (03:08): (this is just an assortment of stuff to give you an idea)...

<4>licq: WRITE block 65680/8 on 03:08
<4>kdeinit: READ block 133800/8 on 03:05
<4>kdeinit: READ block 133808/8 on 03:05
<4>kdeinit: READ block 133816/8 on 03:05
<4>kreiserfsd: dirtied buffer
<4>kreiserfsd: WRITE block 17144/8 on 03:08
<4>kreiserfsd: dirtied buffer
<4>kreiserfsd: dirtied buffer
<4>kreiserfsd: WRITE block 17232/8 on 03:08
<4>kreiserfsd: dirtied buffer
<4>kreiserfsd: WRITE block 17240/8 on 03:08
<4>mozilla-bin: READ block 8651872/8 on 03:07
<4>mozilla-bin: READ block 8651880/8 on 03:07
<4>kswapd: WRITE block 87992/8 on 03:05
<4>kswapd: WRITE block 88000/8 on 03:05
<4>kswapd: WRITE block 88008/8 on 03:05
<4>kswapd: WRITE block 88016/8 on 03:05
<4>kswapd: dirtied page
<4>kswapd: dirtied page
<4>kswapd: dirtied page

As an afterthought, I realized I forgot to compiled smbfs.o as a module. I couldnt mount a windows machine. Here's a quick way to add a module without rebuilding the whole darn kernel:

cd /usr/src/kernel/linux
make xconfig  (put "m" next to filesystems->network->smbfs)
make dep
make modules SUBDIRS=fs
mkdir  /lib/modules/2.4.21mw10/kernel/fs/smbfs
cp fs/smbfs/smbfs.o /lib/modules/2.4.21mw10/kernel/fs/smbfs
depmod -a

Now I can use it!

Graphics stuff

This is just stuff I needed for my graphics course
I needed glut libs for openGL. I used: "rpm --replacefiles --nodeps -ivh Mesa-devel-3.4.2-10.i386.rpm". (It conflicted with XFree86-devel)
To compile, I had to use: "gcc -o newpaint newpaint.c -lglut -lGL -lGLU -L/usr/X11R6/lib -lm -lXmu". Use SAME for sierpinskiGasket. (must rename .cpp to .c, or else gcc will use g++)
In Solaris, I had to use: "gcc -o sierpinskiGasket sierpinskiGasket.c -I/apps2/ogl/include -L/apps2/ogl/lib/glut -lglut -lGL -lGLU -lm -lXmu -lX"