Monitor Resolution

Amazing what you have to do to sometimes to coax windows into doing something it steadfastly refuses to do, isn't it? Just today I had an HP pavilion back in the shop (that I had just put a new WD harddrive into before xmas) that the customer had evidently royally f***ed up the file system trying to use Diskkeeper defragging software. Buggered up the files really good. Config.sys was full of garbage characters and many files and folders were missing, programs wouldn't run, (you get the picture). basically a prime candidate for a factory recovery. So after the recovery, I'm installing this update and that update, latest IE, antivirus, etc. Then I go to shut down, and the thing (running Win98 gold) shuts down so fast, you don't even see the "windows is shutting down" screen. Well, upon starting back up, scandisk runs. EVERY time. And evidently because it is shutting down too fast, and the harddrive cache isn't getting flushed, leaving the volume "dirty" so scandisk wants to run at every boot as a result. If you recall (dust off those old Win98 update memories) Win98SE and WinME had this very problem and MS released a patch to deal with the issue (which merely updated a system file and added a 2 second delay to the shutdown sequence). Problem is, this system was running Win98 gold, and the patch refused to install on it because it was written for Win98SE and WinME only. So I used WinRAR to extract the contents from the MS patch file, and inside there are a couple of .inf install files, new system file, and the .reg files that enable the 2 second delay setting in the registry. Well, low and behold, if you use the .inf file to install the patch (right-click-install) it would install the update without the OS version verification. Long story short, shutdown-scandisk problem is now SOLVED because I forced the update to install on a version of Win98 it wasn't released for. I know I could have just disabled scandisk from running at startup by using the option in MS Config, but that would have also prevented it from running in the event of a real problem, and I didn't want that happening. So I decided to try this instead. Tomorrow will see some serious stability testing to make sure this fix doesn't pose any other problems. But so far it looks great. :-)