Joseph Maddison's Computer Page
Here are some solutions that I've figured out to weird problems that nobody seems to have run into before. It's possible that noone else will ever run into these situations, but I'm posting here on the chance that somebody will. If you get here by finding a keyword through a search engine, maybe this will help. Good luck. Feel free to email me if you need additional info (see below).
1/1/2003 - I updated my Windows 2000 installation to Service Pack 3. Everything seemed to go okay, but when I rebooted, I got a strange error and the Blue Screen of Death (BSOD). It was something like:
STOP: c0000221 Unknown Hard Error
\SystemRoot\System32\ntdll.dll
I also sometimes got a similar message about tcpip.sys. I was thoroughly stuck, so I tried some things. Replacing the ntdll.dll from another Windows 2000 SP2 system got things working again. I'm not certain why this worked, but then I downloaded the rest of the Windows updates and I got a similar crash. I renamed the new one back and it worked for one reboot, then I got the old BSOD crash again. I was thinking that perhaps the updates needed to be installed in a different order or something.
An article from http://www.scotsnewsletter.com suggested that the Maxtor ATA100 card's drivers (Promise Technology mfr) were the cause of problems with SP3. I downloaded a new driver from http://www.promise.com. That may have been it finally. My current configuration now uses the netdll.dll and tcpip.sys from SP3 and I'm operational. Cross your fingers the next time I reboot! If I had to do it again, I'd have upgraded to the new driver before upgrading to SP3!
2/1/2003 - Sigh, I finally could not get it to boot, again. The only recourse was to uninstall SP3 and go back to SP2. It is possible that my problem was due to the additional hard drive controller card on my system, but I did not care to expriment further! A correspondent who found this page reported a similiar problem that went away with a different motherboard. Another correspondent said that he was having the problem under Windows XP.
6/3/2004 - Fast Forward a year or so... my computers haven't been real stable, but eventually they
got impossible to work without crashes. I clean installed Windows XP Professional and things actually
got worse after a while. I starting getting errors like "UNMOUNTABLE_BOOT_VOLUME" and data
corruption.
My problems increased with the newer and faster hard drives I bought. It turns out that my
motherboard has an ATA100 controller, and the drive was an ATA100 or even ATA133. I use
only the good 80 pin cables, so it appeared that everything was correct. Almost. It turns out that
the mobile racks (removable drive bays) meet only an ATA66 standard. Even worse, the data
speed doesn't slow down if one of the components is lower-rated (like noise on a telephone
line slowing a modem connection). Instead, this was causing wide-spread data corruption.
Computer repair folks say they have little confidence in a $15 rack. (This is only after
I bought it from them) I have now direct wired my hard drives and stability has returned.
Crashes are now very infrequent.
I want to eventually return to using mobile racks. It seems like Lian-Li RH-42 racks have
a good reputation. It may be more expensive than the "no-name" racks, but it may cost
more in the long run to not have a stable system.
Visor Pro
4/14/2003 - I am having general stability problems with my Visor, probably due to the Hacks and other odd drivers sitting on it. The worst is, a crash that ends in a rebooting loop. The only way out of it is to restore from backup... fortunately, I have not one, but two backup modules. I was in the process of removing questionable software and running for a while without it, to see if perhaps one particular piece is responsible for my misfortune. I was moving files to my 8 MB Flash Module, when I inserted the Flash Module, and none of the programs showed in my main menu. I reset and got the dreaded:
"DataMgr.c, Line 3362, Null dbID passed"
After some trial and error, I figured out that having the StowAway keyboard driver and configuration app moved to the Flash Module, but not on the Visor (with the keyboard disabled, of course) would cause this problem.
I restored the Keyboard files to the Visor, then I was able to access the Flash Module again.
Deleting them from both the Visor and from the Flash Module then did the trick. Weird. It was fortunate that I had backups so that I could experiment.
6/3/2004 - Stability problems have been a thing of the past-- I changed to a
different Visor Pro and crashes and memory wipes are no longer the status quo. My
wife has the "bad" one-- it was an experiment. I fully expected it to fail for her and
then it would prove that the hardware was bad. Strangely enough, it has not been a
problem -- at all. It turns out that she usings maybe 2 MB of the 16 MB available.
I use almost all of the memory. Presumably, there's a problem with a memory
bank above where she uses memory. Surprisingly, there is no such application as
a memory checker for a Palm OS hand-held.
Palm LifeDrive
9/30/2005 - I switched to a Palm LifeDrive in May 2005 and discovered it was slow and unstable. It's still slow, but I discovered it is now much more stable, now that I've removed the Wireless Bluetooth Keyboard drivers entirely from the program space. Think Outside may eventually fix the problems, if I get a chance to report them.
10/2/2005 - I got an evil startup message from HotSync.exe after installing some extraneous desktop software conduits-- possibly ThoughtMgr or HanDBase, though I have not attempted to isolate it. The message Hotsync delivers, before it crashes is
"Ordinal 157 could not be located in UserData.dll."
It turns out that the good version of c:\Program Files\PalmOne\UserData.dll (file date 6/9/2004 / size 139,264) was replaced by an older version. Pull the correct file from a system backup, or else reinstall the Hotsync software to fix this.
Java 1.5
10/12/2005 - I recently got the following warning when compiling a source file that did not have any warnings under JDK 1.4:
warning: [unchecked] unchecked call to put(K,V) as a member of the raw type java.util.TreeMap
It turns out that the solution is to use the new generics capability in Java 1.5, for example:
TreeMap<String, Integer> keywords = new java.util.TreeMap<String, Integer>();
This ensures that the Objects inserted into the Map are checked for the proper type. It will still work without generics-- that's why this is just a warning.
Thanks for visiting,
Joseph Maddison
jdmaddison@spamcop.net
