Pc Startup Troubleshooting Tips

PC Startup Troubleshooting Tips By Bob O'Donnell As wonderful as computers can be, they can also be incredibly infuriating. Probably the most frustrating problem computer users run into are startup problems, where your computer won’t boot. Equally annoying are error messages you constantly run into during your computer’s startup process. In this article I’ll give you a few tips on how you can avoid some of the most common problems that happen right after your computer turns on. Learning the Boot Before getting into the troubleshooting details, you need to know a little about when goes on during the startup process. The reason is, there are actually quite a few steps that occur in between flipping the power switch and hearing the familiar Windows 95, 98 or Windows ME startup sounds and seeing the Windows desktop. In fact, there are a whole series of files that are automatically loaded one after the other when you turn your computer on. The trick with troubleshooting startup problems is trying to figure out which of those files (or what step in the process) causes your specific problem to occur. If you don’t know approximately where in the startup process your holdup happens, you could end up wasting an inordinate amount of time (and even causing more problems) on something that’s irrelevant to your situation. So, here goes. When your computer is first turned on, it automatically loads a program called the BIOS, or Basic Input/Output System, which is stored on a special chip on your computer’s motherboard. The BIOS is essentially a combination of software and hardware in that it consists of software, but the contents of that software is stored in a hardware chip. On most recent computers, the BIOS can be updated via a process called flash updating, which uses a piece of software that’s stored on a special startup floppy disk to overwrite the contents of the chip with a new version. On older computers, however, to upgrade the BIOS software you need to physically remove and replace the chip itself. (BIOS upgrades are often a necessary step in solving hardware-related problems. See the "PC Hardware Troubleshooting Tips" article for more.) One of the first things you should see on your computer’s monitor when you start your PC is some type of message that’s akin to "Hit Esc to enter Setup," although instead of Esc it may say F2 or F10 or any number of other keys and instead of Setup it may say CMOS Setup or BIOS Setup or just CMOS. Make note of the key required to enter the Setup program because you may need that later (some startup problems can only be solved by changing some BIOS/CMOS settings via the Setup program). As the BIOS runs, it performs a number of tests on your hardware called POST (Power On Self Test), such as checking the memory. Then it lists any devices that it finds attached to your computer’s internal IDE controller(s). Typically, this is any internal hard

drives, CD/DVD-ROM drives, tape drives, etc. One common problem you can check for is to make sure that all the devices that are supposed to be attached to the IDE controllers are listed. If not—for example, if you just upgraded your hard drive or added a DVDROM drive and you don’t see a reference to them right after your computer turns on— then more than likely there is a connection problem between the IDE controllers on your motherboard and the device itself. You’ll have to open up your PC, check the cable connections at both the drive and on the motherboard and, if necessary, replace your IDE cable(s). In a few rare instances you may also have to make some changes to your hard drive settings in the BIOS Setup program mentioned earlier. Another possible problem if a drive that used to appear in that listing no longer appears is that a physical problem has occurred with the drive. (Computer peripherals don’t last forever after all—although you should get at least five years out of a hard drive or CDROM drive.) If that’s the case, check the cable connections—sometimes they just come loose—and then run a disk utility such as ScanDisk or Norton Utilities’ Disk Doctor from a boot floppy disk to see if you can salvage any of your data. More than likely in this scenario it’s time to start thinking about a new hard drive. If you have a SCSI controller installed in your PC, you will also see a message about any BIOS (separate from the main system BIOS) that it has, as well as a listing of all the internal and external SCSI devices attached to it. Again, if you don’t see a device listed, you need to double check the cable connections, or check the integrity of the devices themselves. One other possible issue with SCSI devices has to due with SCSI termination issues, which I discuss more thoroughly in the "Mac Hardware Troubleshooting Tips." (SCSI is SCSI, regardless of platform, so the concepts explained there are just as relevant for PC people as they are for Mac users). Mastering the Master Boot Record After the BIOS finishes it work, it hands control of the startup process to some specific files stored in the first sector of your hard drive. This special area is called the master boot record and it contains critical information about how to start Windows (or any other operating system(s) you may have installed on your PC) as well as the specific files needed to continue the startup process. Because of critical role it plays, your hard drive’s master boot record is also a favorite target of virus writers. Why you wonder? Well, if a virus corrupts or rewrites your master boot record, your computer won’t work (and people who create viruses seem to get some sick pleasure out of causing computers to "break"). Viruses that target this area are called, logically enough, boot viruses and they are probably the most common type of virus there is (other than Word macro viruses, that is, but those aren’t typically very lethal). Most anti-virus programs can take care of boot viruses, but only if you have a boot floppy disk with the anti-virus program installed on it (otherwise you can’t boot your computer to get to the anti-virus program to run it!).

If you’re really stuck, one other trick you can try is to use the Fdisk partitioning program, which comes with any version of Windows, to rewrite the master boot record. You’ll need a boot disk with the Fdisk and Sys programs on it for this trick to work (see "Creating a 'Real' Windows 95 (or 98 or ME) Boot Disk" for more on how to do this). You can find both those programs (as well as other DOS utilities) in the Commands folder inside your Windows folder. As with any startup floppy disk, you need to turn your computer on with this floppy disk in the floppy drive and then the computer will use it to start the computer instead of trying to use your hard drive. If your PC boots successfully from the floppy, you should be taken to a DOS command prompt that says A:\. To make anything work under DOS you have to type in commands that tell the computer what to do, so use the following commands to continue with this process (hit the Enter key after each one). c: fdisk /mbr sys c: The first command switches over to your main hard drive, the second command tells the Fdisk program to rewrite the master boot record, and the third command rewrites the critical startup files back to your master boot record using the DOS Sys program. I’ll warn you now that this technique doesn’t always work (and in some cases you may need to reinstall Windows), but if you’re desperate, it’s certainly worth a try. Starting the Startup Files If you’re not having any problems up to this point, and your computer continues to boot, then that means your master boot record and a few of the critical startup files (such as Io.sys) are fine. Next up are some old DOS startup files that—despite Microsoft’s claims to the contrary —continue to play an important role for Windows on many (though not all) PCs. Specifically I’m referring to the Config.sys and Autoexec.bat files, which are typically used to load device drivers and startup programs, respectively. (Device drivers are pieces of software that are used to communicate between the operating system and all the different peripherals inside your PC, such as your CD-ROM drive, sound card, modem, etc.) In addition, these files can be used to set up a few relatively obscure system parameters that usually don’t have any impact under Windows 95, 98 or ME. If you have Config.sys and Autoexec.bat files on your PC (again, not all systems do), then any commands that they contain are automatically loaded or any programs that they refer to are automatically started when they themselves are loaded. One very common problem that occurs when you install new software (or hardware) on your PC is that the installation program may add a line to one of these startup files. This line (or lines) tells

your PC to automatically load some software that the new program (or piece of hardware) that you just installed requires. Unfortunately, sometimes those new startup programs conflict with other software you already have installed on your system and cause the boot process to stop or your system to crash. (See the "PC Software Troubleshooting Tips" article for more on how to deal with those types of problems.) Another related problem can occur when you remove software (or hardware) from your system. Depending on how you removed it or how the program’s uninstaller works, you may end up with a situation where there are references in one of these startup files to programs that no longer exist on your hard drive. In most cases like this you’ll get some type of error message that says something about being unable to find a particular file, but if you hit OK, your system often continues to boot and works just fine. In other words, it’s not causing any real problems for your PC, but it can be somewhat disconcerting or, at the very least, annoying. In either the first or second example (but particularly the latter), the solution to the problem usually involves removing the reference to the problem line in your startup files. In other words, you tell your computer not to try and load the files that are either causing a conflict or no longer exist. You can do that by either finding the specific line in your Config.sys or Autoexec.bat file and deleting it or by essentially telling your PC to ignore that particular line by "remarking it out." There are two ways to do this as well. You either type rem and then the space character at the very beginning of the offending line, or use the semi-colon (;) character at the beginning of the line. So, for example, a line that previously said: Device=Nec_bm.sys Would change to: Rem Device=nec_bm.sys or ;Device=nec_bm.sys. To edit your Config.sys or Autoexec.bat files (which should find in the root directory of your main drive—typically C:\) under Windows 95, you need to use some type of text editing program, such as Notepad, or the DOS Edit program. The easy way to do it, however, is to open the Run… command window off the Start menu and type in Sysedit, which launches the System Configuration Editor. This utility program automatically opens the five main startup files (including Config.sys or Autoexec.bat) in five overlapping windows for easy editing. Regardless of how you open files, you can make any necessary changes to the file(s) and save it (or them). Windows 98 makes the process much easier by providing a program called the System Configuration Utility that lets you individually turn on and off lines in any of your startup files, including Config.sys and Autoexec.bat (as well as lots more that I’ll be getting to in just a bit). You can find the powerful System Configuration Utility via the System Information Tool, which is available off the Start menu via Programs, Accessories, System Tools. Launch System Information Tool and then from its Tools Menu you’ll find

the System Configuration utility (as well as other very useful troubleshooting tools—this is one area where Windows 98 and Windows ME offer clear benefits/improvements over Windows 95). The easy way to launch the program it is to select Run… from the Start menu and type in msconfig.exe. And More Startup Files After the Config.sys and Autoexec.bat files (if present) load, the System.ini and Win.ini (again, if present) load. While these files are primarily leftovers from the days of Windows 3.1, they too can have a role in the startup process. And similarly, the exact same kinds of problems that can occur with Config.sys and Autoexec.bat can occur with these files—specifically, references to startup files that conflict with other programs or no longer exist on your system. In addition, though rare, it’s possible that certain settings in the System.ini or Win.ini can cause startup problems. Thankfully, the exact same methods that you use to solve problems with Config.sys and Autoexec.bat can be used with System.ini and Win.ini. The respective System Configuration utilities I mentioned above will open them and let you delete or remark out references in those startup files as well. In addition, you can manually edit or, in some cases, simply delete settings in these files that are causing problems. (Of course, you may want to make a backup copy of them before you start deleting anything.) Reading the Registry The first "real" Windows 95/98/ME file to load is the massive Windows Registry, which keeps track of loading all the 32-bit Windows 95/98/ME drivers—their filenames typically end in .vxd—for all the peripherals in your PC. (The previously mentioned Config.sys only loads older 16-bit drivers that you solely need under DOS.) The Registry is in charge of loading a ton of other important operating system files as well. If you ever want to see just how many, take a look at the Bootlog.txt that you’ll typically find in the root directory of your C: drive. (You’ll have to turn on hidden files in Windows Explorer by going to the View menu, selecting Folder Options and going to the View tab in order to see it.) When you see the text message that Windows 95, 98 or ME is starting to load (and then when you see the Windows splash screen), then you’ll know that your PC has begun loading and "processing" the registry. If you don’t quite make it to the Windows splash screen, you could have a problem with one of the low-level drivers that form part of the Windows Virtual Machine Manager (or VMM). The VMM is what lets Windows run multiple applications at once (which makes your PC appear to be multiple "virtual" machines in one—hence the name). Typically, you’ll see a message that refers to being unable to load vmm32.vxd. In reality, vmm32 refers to a folder located inside your Windows/System folder on your main hard drive that holds several driver (.vxd) files. In some situations this error message may indicate that some of the low-level drivers stored in this folder have been corrupted, which

typically requires you to re-install Windows. However, I have found that simply turning the computer off, letting it rest a few minutes, and then turning it back on sometimes takes care of the problem all by itself. (In fact, this is a good piece of advice for startup problems in general.) In other situations, stopping right before the Windows splash screen could indicate a corrupted Windows Registry. That’s never a good problem to have, because it often requires re-installing Windows and all your applications (because application preferences are also stored in the Registry). You might be able to fix some Registry problems with Microsoft’s free RegClean if you can find it (it's no longer available on the Microsoft web site) or, if you have Windows 98 or ME, by running the Scanreg (DOS) or Scanregw (Windows) utilities bundled with that OS. If your registry is completely hosed, you can go back to a previous version (or any one of five or more previous versions under Windows 98 or ME), by renaming the Registry backup files that Windows automatically creates every time it successfully starts up. The two files that actually make up the Registry are called System.dat and User.dat and the main (or most recent) backups of those files are called System.da0 and User.da0. You’ll find all of these files at the main level of the Windows folder. To make use of the backups you’ll need to boot to a DOS prompt and then rename the System.da0 and User.da0 files to System.dat and User.dat respectively, overwriting the other files in the process. To do that in DOS, you use the following commands: Rename system.dao system.dat Rename user.da0 user.dat Once you’ve done that, you’ll need to restart the machine (holding the Control, Alt and Delete keys simultaneously always works), and then reboot from the hard drive to make the changes take effect. This is a good technique to use if you just installed a new application and start having Registry-related problems immediately afterwards because it will revert the Registry to the state it was in before you did the install. If you made the installation a few days (or, more importantly, a few "restarts") ago, then the backup versions will have the same problems as well, so this technique may not work in those situations. Registry Drivers If your computer makes it to the Windows splash screen before it stops, you probably have a driver-related problem. Many PC startup problems are the result of drivers that don’t load properly from the Registry, either because of a conflict with another driver, because a driver has been corrupted or accidentally deleted, or other possible problems. Unfortunately, there’s no easy to way to edit the Registry in the same way you can with other startup files, so Windows provides other options if your computer stops the boot process at this point.

The most common one is known as Safe Mode, which is essentially a limited, or minimal version of Windows that loads when your PC is having problems. Thankfully, in many situations Windows is smart enough to know when a problem has occurred and will automatically start in Safe Mode (or at least give you the option to start in Safe Mode) the next time you restart. In some situations you may also want to force your computer to start in Safe Mode. To do that, hold down the F5 key when you start up and keep holding it until you see that your PC has booted to Safe Mode. You can easily tell this because your screen resolution will be reduced to the VGA standard 640 x 480 resolution and you’ll see the words "Safe Mode" all over the screen. (Another way to do this is to hold down the F8 key during boot time until you’re presented with the Windows Startup Menu and then choose Safe Mode from the list of options presented there.) When you’re in Safe Mode Windows skips the Config.sys and Autoexec.bat files and loads a minimal set of drivers that lets your PC function at a basic level. However, usually you can’t print, use your modem or do lots of other things you would normally otherwise be able to do. You can, however, run most of your applications, so if you’re desperate to get some work done and your machine keeps crashing, you may want to consider simply working in Safe Mode for a while. The idea of Safe Mode, however, is to let you do things like install driver updates or make other software changes (such as changing references to start up files, as discussed earlier) while in a familiar Windows environment. Once you’ve made your changes, you can restart the machine and check to make sure that everything works. Another common startup troubleshooting technique is to take advantage of the Step-byStep Confirmation mode that Windows also offers you when you use F8 to boot your machine. In step-by-step mode, Windows asks you before it loads certain programs called for in your startup files. The idea is that by loading files one at a time, you can more quickly tell what’s causing the problem. Basically, you say Yes to each prompt until your machine freezes and you’ll learn the culprit. Note that this mode processes the entire Registry at once, however, so if the problem is a particular file called for in the Registry, this method won’t tell you which one it is. Still More Places for Programs to Run Believe it or not, we’re still not done yet, particularly if you’re trying to track down all the different programs that can be loaded automatically when your computer boots up. Some applications and utility programs, for example, take advantage of the fact that in addition to loading drivers, the Windows Registry can also automatically load applications (or small programs that continuously run in the background while your PC is on) in a kind of "hidden" way. The only way to find these is to open RegEdit, the built-in Windows Registry editor and go to the HKEY_LOCAL_MACHINE\SOFTWARE\ Microsoft\Windows\CurrentVersion section of the Registry and click on the Run, RunOnce, RunOnceEx, RunServices, and RunServicesOnce keys inside the Registry. Each item you see listed on the right hand side is started up each time Windows boots.

To prevent any items listed in any of these keys from starting up, you need to delete the specific settings from the key by right-clicking on it and selecting Delete from the context menu that appears. Select all but "Default" if you want to get rid of all of them and simply quit RegEdit when you're done. (Note that deleting these settings will not delete the software they're pointing to, just the command to launch them at boot time.) I don’t recommend you edit the Registry unless you know exactly what you’re doing, but this simple procedure can be helpful. Finally, the last step that occurs in the Windows startup process is to automatically launch any programs that have shortcuts stored in the Startup folder, which you can find inside the C:\Windows\Start Menu\Programs directory. As with Registry’s Run keys, some applications and utility programs take advantage of this feature to automatically launch program’s every time you start Windows. Using the Startup folder is a more "public" way of doing it, however. Unlike the Registry, fixing problems related to the Startup folder is easy—simply drag the shortcut of the offending application (or anything else you want to keep from starting automatically) out of the Startup folder and onto your Windows desktop. If you prefer, you can even delete it—as long as it’s just a shortcut it won’t affect the real application. The next time you restart your machine and Windows loads, whatever you’ve taken out of the Startup folder simply won’t start automatically. You can start it manually any time you want, however, by simply double-clicking on the shortcut icon you dragged to your Windows desktop. The End As you can see, there’s quite a bit to the Windows startup process. In fact, if you had to go through all these different techniques it could literally take you days to figure out what your problem was and what was required to fix it. If you can narrow down where in the process your particular problem occurs, however, you can save yourself a lot of time and effort. Ultimately, all computer troubleshooting problems—including those related to startup— are solvable, but that doesn’t mean they’re necessarily logical, nor easy. (If you feel like you've invested enough time, you can always just start over and reinstall everything. See "Starting Over: Repartitioning, Reformatting and Reinstalling" for more.) As with all difficult procedures, perseverance, common sense and a bit of good luck is what you’ll need to get through your startup problems and get back to enjoying your PC. Good luck.

PC Software Troubleshooting Tips By Bob O'Donnell Most problems that occur with computers are the result of software-related issues: programs that lock up, printers that suddenly won’t print, operating systems that won’t

boot and lots of other common quandaries are all somehow related to the operation of software on your PC. So, if these kinds of problems happen to you, take comfort in the fact that the good news (if you can call it that), is that you’re not alone. Unfortunately, the bad news is that troubleshooting software problems on PCs is unquestionably the hardest problem-solving task you’re ever likely to run into (on your computer, that is). The reason for this is because of the Byzantine-like structure of today’s PC applications and operating systems. In essence, the problem can be boiled down to this: All the various pieces of software on your computer—your main applications, the drivers for your specific peripherals, and the various Windows system files—work together in an enormously complex web of inter-relationships. Adding, removing or altering one piece of that puzzle can (and often will) lead to software that just doesn’t work. That would be fine and good if changes to this puzzle were few and far between, but for most people, these types of changes occur on a surprisingly frequent basis—and very often these changes are out of your control. Sometimes it’s a matter of simple user error (oops, shouldn’t have deleted that file), but often it’s due to conflicts between two or more different pieces of software, or even as a result of simply installing or removing applications. The end result is that software-based problems are an all-too-frequent occurrence on virtually all Windows computers. Keep Current One good general rule of thumb to keep in mind when doing software-related troubleshooting is to make sure you have the most recent versions of your applications, drivers and operating system. Usually that means taking an inventory of the software and hardware you have installed on or in your computer, and then trolling the web sites of the companies who make those products in search of updates, patches, service packs and other types of bug fixes. If you're using Windows 98, Windows 98 SE or Windows ME, you should take advantage of the built-in Windows Update feature to help you in this regard. Be aware, however, that as handy as Windows Update may be, it will only provide you with operating system and some driver updates. It will not get all the drivers you need nor will it get any application updates you might require. So, use it as a starting point, but not as a sole resource. Also remember to check for viruses. Most of the times you get a virus on your PC you'll know about it, but some of them work "stealthily" in the background, so you always have to be careful. Make sure that any computer you work on has good anti-virus software installed and properly configured so that it's regularly checking for viruses. Even more importantly, make sure that you keep the virus definition files for the program current. Unless you update those definitions on at least a monthly basis, your data will be at risk. When is an Application not an Application?

Many of the software problems that people run into stem from the confusing nature of today’s Windows applications. Instead of consisting of one large chunk of code (as they generally used to do, and still in many cases do on the Mac), today’s software programs are actually made up of lots and lots of little pieces held together by an umbrella application. This umbrella app is typically the main program file you double-click on to get an application started. To confound matters further, these different pieces are often strewn across several different locations on your hard disk, which creates even more possibilities for errors or other breakdowns. Many of these individual pieces are .DLLs, or Dynamic Link Libraries, which are chunks of program code that are called upon to perform various functions as you run an application. Another name for them is Application Extensions. Ideally, you never have to worry about these files—they just do their magic in the background while you do your work. And, when everything works right, that’s exactly what happens. If one of those files is deleted, changed or somehow corrupted, however, the application that uses them won’t work, and you’ll be in trouble. Let me explain the installation process to give you an idea of the issues involved. When you install an application, what usually happens is that the installer creates a folder for your new program, copies some of the application’s pieces into there, copies other pieces into the Windows folder, creates shortcuts to your program on the Start Menu (or in the Program Manager for Windows 3.1), makes additions to the Windows Registry, and may even make some additions or alterations to your startup files, such as autoexec.bat, win.ini, etc. (In many cases, these changes are the cause of problems that keep your PC from booting properly. See the "PC Startup Troubleshooting Tips" article for more.) The complexity of this process is the reason, by the way, that uninstaller applications were developed in the first place—it simply got to be too hard to keep track of where all the different pieces of an application were stored (if you could even get that info in the first place). The files copied into the Windows folder may be .DLLs or they may be other files that the program’s developer has decided should be kept there. Unfortunately, so many developers have decided to store bits and pieces of their applications in the Windows folder that it’s become a dumping ground for a whole lot of junk. (Even worse, there’s no easy way to separate the wheat from the chaff—the useful, necessary stuff from the unneeded "bloatware"—so Windows folders just keep growing and growing….) Some of the .DLLs will probably be unique to the program you installed and will only be used by that application. Others will be shared files, which means they might be used by more than one application. And that’s where lots of problems arise. Imagine, for example, that application A uses a shared .DLL called ABC.DLL. Now, let’s say that you install application B, which also happens to use the shared ABC.DLL. But, it turns out, application B uses a later version of ABC.DLL and so it overwrites the existing ABC.DLL file when it’s installed (without any warning to you). Now, depending on how application A is written, the next time it’s run and it looks for ABC.DLL, it may

seamlessly work with the later version of ABC.DLL installed by application B, or it might crash (a likely possibility). If the latter occurs, it will undoubtedly lead you to wondering how that could’ve happened, since it worked just fine the day before. Sound familiar? A similar type of situation is possible in reverse. Let’s say you try to remove application A from your system using Windows' Add/Remove Programs Control Panel (or even the application’s own uninstaller). At some point during that uninstall, the program may ask you if it should remove any shared .DLLs that are not being used by other applications. The default choice is to remove them and that's what most people do, but that can lead to more problems. Again, depending on how the uninstaller works in conjunction with the application being removed, it’s possible that application A would remove ABC.DLL because it didn’t know that a newer version had been installed by application B and therefore would think that it wasn’t being used by another application. The result would be that after uninstalling application A, application B wouldn’t work because it was missing the ABC.DLL file (and you might even get an error message to that effect). In both of these instances, the only real solution to these problems would be to reinstall the apps in question because there’s usually no simple way to reinstall only a shared .DLL (and even if there is, it may not be at all obvious where it needs to be located). In the first example, you’d probably have to first get an updated version of application A from the vendor’s web site and then reinstall that, hoping that it works with the new version of the ABC.DLL (and praying that it doesn’t add any new shared .DLLs that might break other applications!). In the second example, you’d just need to reinstall application B. Windows System File Problems A related, but even worse version of the problem can occur if an application overwrites some of the shared system files used by Windows 95—although that often means the operating system won’t load. (The same types of problems can also happen with Windows 98 or Windows ME, although Win98's System File Checker utility can help take care of these problems much more easily.) If that’s the case—it is relatively rare— you’ll probably need to reinstall Windows itself. (And if you do, you'll probably need to have a boot disk complete with a CD-ROM driver. Here's an article that tells you how to create one.) If you do think it's time to reinstall Windows, you might try the Windows Setup program's Verify option first—it should save you time and maybe even a few extra headaches. The Verify option is supposed to be available any time you run Windows 95, 98 or Windows ME Setup on a machine that already has the same version of the operating system installed (although I've found that it doesn't always show up for some reason).

What happens is, the Setup program detects that Windows is already installed and then asks if you'd like to Verify your installation. If you do choose to Verify, it goes through an installation log file called Setuplog.txt (you can find it in your main hard drive's root directory--e.g., C:\--if you want to look at it) to see what should be installed, confirms that all the required files exist and aren't damaged, and then silently reinstalls any components that it finds are missing or damaged. In some cases, this will solve Windows system problems. Diving In If your willing to get your digital fingernails a bit dirty and you know the specific name of a .DLL or other Windows system file that's causing a problem (perhaps because an error message keeps telling you the problem file's name), there's another option. You can try to individually reinstall particular files. The process is not trivial, however, because all the Windows system files are stored in compressed Cabinet (.cab) files either on your hard disk, the Windows 95/98/ME CD-ROM, or both. What you have to do is find the compressed file within a particular .cab file, decompress it, and then copy it to appropriate directory inside the Windows folder. Microsoft includes a DOS-based .cab file decompressor called Extract.exe on the Windows 95 CD and installs it in your Windows folder as part of the default installation. However, as you'll be able to quickly see from this article in Microsoft's Knowledge Base on how to use the Extract program, the program is not easy to use. Nevertheless, it can help you find the files you need. Another more straightforward option is to get the CabView utility from Microsoft (it's party of their handy PowerToys utility pack), and use it to specifically find and then decompress the component you need. Another improvement in Windows 98 and Windows ME is you can view the contents of CAB files and extract individual elements just by using Windows Explorer. Most shared .DLL problems can be avoided if the programmers have done their homework, but given all the possible combinations of programs and .DLLs that are out there, it’s almost impossible for them to avoid all problems. And unfortunately, as frustrating as it may be, the only real answer to these problems is to reinstall the software. In fact, I’ve even heard or read some people recommend that you reinstall Windows and your apps every 3-6 months or so to avoid problems. I think this is overkill, but the thought has passed through my mind on more than one occasion as I’ve struggled through trying to resolve software problems on my own or other people’s machines. Another possible solution is to not remove any shared .DLL files when you uninstall an application, but while that may work in a few instances, it eventually leads to a Windows folder full of unnecessary, potentially problematic stuff, so it’s not a great long-term solution. Frankly, there’s no great long-term solution other than to slug your way through, keeping your software as up-to-date as you can (and even that doesn’t always work— sometimes it’s better just to find a combination of software that works and stick with it). Getting into Conflict

Oftentimes, software problems are due to conflicts between two applications, which basically means one or both of the programs in question does something in the course of their normal operation that causes the other application to crash or to not function properly. This is related, though not identical, to the .DLL problems described above. Some conflicts are relatively easy to troubleshoot. If you install a new application and every time you try to run it while your word processor is open the new program crashes, but the new application works fine if the word processor isn’t running, then you’ve got a software conflict. The only way to fix it is by getting an update for one (or sometimes both) of the applications in question. Hopefully you won’t end up in a situation where each vendor claims it’s the other’s fault and nothing gets done, but be aware that that exact scenario sometimes happens. Other conflicts are much harder to determine. One thing you should check is to see if the conflict is related to applications that are running in the background. Unbeknownst to most computer users, quite a few small applications are usually running in the background on typical Windows 95/98/ME machine. This is true even if you don’t see any applications running on the Taskbar (which is why this can be so confusing). Under Windows 95, you can see a list of all applications running at a particular time by holding down Ctl-Alt-Del simultaneously (just do it once, though—if you do press all three again, it will restart your machine without prompting you to save any open files). The Close Program Window that pops up lists all the currently running tasks. While some are easy to understand, many of the programs listed only use obscure names like Systray or Rnaapp (which happen to be System Tray—which is the system application that puts all the icons in the lower-right corner of your Taskbar—and Remote Networking—which is used by Dial-Up Networking whenever you connect to the Internet via a modem). Figuring out what some of these various programs are and do can be extremely difficult because there’s usually very little information available about them. Again, Microsoft has improved this situation in Windows 98 and ME by including a handy utility called Microsoft System Information that gives a thorough, understandable breakdown of all the different software components that are open and running on your machine at the time you run the utility. One possible option for Windows 95 users is to pick up a copy of a shareware utility from SiSoft called Sandra. Sandra's Processes Information module can not only give you more information about each task (technically called a process) currently running on your machine, it can also tell you how much memory each one is using. Very handy. Another option for Windows 95 users is to get a copy of the WinTop utility, which is part of Microsoft's Kernel Toys for Windows 95. What WinTop does is give a quick overview of what applications and other software processes are open on your computer at a given time and shows you how much of the processor's time they're demanding.

For troubleshooting purposes under Windows 95, 98, or ME you can quit applications individually from the Close Program Window mentioned earlier by selecting one and clicking on End Task. As you go through and quit each one you can see if that resolves your problem, but it’s a tedious process that doesn’t always give you the results you want. Another way to prevent (or see) some, though not all, of the applications that run in the background is to check your Startup folder, which is buried inside the Windows folder (it’s in the Programs folder, which is inside the StartMenu folder). Many applications place small utility programs inside the Startup folder, so you might be surprised at all the stuff you find there. All the applications that are listed in the Startup folder are run in the background every time Windows 95 starts. Some of them put an icon in the System Tray to let you know that they’re running, but many of them don’t provide any visual indication whatsoever that they’re currently in use. If you want to prevent them from loading at startup to check for conflicts with certain programs, or if you just want to remove them completely, just remove the shortcuts listed in the Startup folder, either by deleting them or moving them to a different place. (Note that deleting a shortcut has no impact on the real file that the shortcut points to—it will still be on your system, but it won’t load at the system’s startup.) I’ll Take the Driver So far, I’ve only touched on applications, but there’s still more (hey, I told you this was hard…). Drivers, which are small pieces of software that allow your computer and operating system to "talk" to various peripherals that make up your system, such as your video card, printer, CD-ROM, etc., are often at the heart of software-related computer problems. Like regular applications, drivers are subject to bugs, or errors, in them that can cause them to not function properly. More often than not, however, driver problems are due to conflicts or incompatibilities either with specific applications, other drivers, or some piece of system software. As I described in the beginning of this piece, today’s software is like a very complex puzzle and changing one piece in a seemingly unrelated area of the PC’s total software system can create problems somewhere else. Driver problems can manifest themselves in a number of different ways. For example, if an application that used to be able to print can no longer print, it might be the application’s problem, but there’s also a very good chance that something has happened to the printer driver. Either a new piece of software that was recently installed has created a conflict that prevents the driver from working properly, or the driver file may have somehow been corrupted. (File corruption is a baffling, confounding problem that can happen for any number of reasons, including a quick electrical surge, an error in an application, a hardware error, and more.)

As with application problems, the only way to resolve many driver issues is to get an updated version of the driver and reinstall it. (For more info on installing or updating drivers, see the "PC Hardware Troubleshooting Tips" article.) The End As difficult as it may be to accept, there’s no magic solution that can tell you exactly what’s causing the problem when a piece of software doesn’t work right. Even utilities that claim to solve system software problems for you are only partially successful and sometimes they introduce new incompatibilities or other problems of their own. That’s part of the reason why software troubleshooting is so tough—and often such a frustratingly inexact science. If you keep your calm and use your head, though, you can usually get through lots of seemingly tough situations on your own (or perhaps with the help of a vendor-supplied update). Remember this maxim: when in doubt, reinstall! (For more on reinstalling everything, see "Starting Over: Repartitioning, Reformatting and Reinstalling.") Good luck

PC Hardware Troubleshooting Tips By Bob O'Donnell Though they tend to cause some of the nastiest symptoms—computers that won't boot, scary and/or confusing error messages, etc.—computer hardware problems are usually the easiest type of problems to solve. The trick, as with all troubleshooting ventures, is to figure out where to start and then focus your efforts. First, of course, you need to check the stupid stuff. You'd probably be amazed how many "problems" are solved by connecting the cables, or turning on the power switch that you swear you just did. Beyond that, double-check the snugness of your connections— jiggling in a new add-in card or screwing in a cable connection can (and often does) make a difference. You may even want to check the integrity of your cables and connectors. I once solved a baffling SCSI problem by noticing that one of the pins in a miniature 50pin SCSI-II connector was bent. I had mistakenly presumed, prior to that, that a bent pin would have prevented me from making a connection at all, but it didn't. Unfortunately, as a result, I wasted several hours on something that could've taken two minutes. Finally, whenever you install something new, whether it's more memory, a new drive, a plug-in card or what-have-you, and something doesn't work, it's more than likely because you made a mistake somewhere in the installation process. Step back through the process again, double- and triple-check your connections, and then try one more time. In the case of RAM that doesn't work or isn't recognized, it could be an incompatibility with the specific manufacturer of the RAM and your motherboard, so see if you can try a different brand before you give up hope. Finally, in some instances installing something new causes a conflict with something else--which is what the rest of this article is all about.

Driver Dilemmas Once you're past the basics, it’s on to step two. As long as your computer boots, then there's a good chance the problem is related to missing, damaged, incompatible or improperly installed driver software, otherwise known simply as drivers. (If your PC doesn't start up, you may want to create a boot floppy disk—see "How to Create a 'Real' Windows 95 (or 98 or ME) Boot Disk" for more. In addition, you should check out the "PC Startup Troubleshooting Tips" article for more.) Virtually every piece of hardware located inside or connected externally to your PC requires a driver to communicate and function with the operating system, applications, and other hardware components in your machine. Drivers essentially translate messages back and forth between the hardware in question and the operating system, thereby allowing your computer system to work as a unified whole (at least, in theory). The truth is, though appearances may suggest otherwise, any computer system is actually made up of a bunch of specialized pieces that don't speak the "native language" of other components and, therefore, require a great deal of translation to communicate and work effectively with them. When any of these various levels of translation break down, well, that's when you get problems. The Device Manager is Your Friend If you're running Windows 95, 98, or ME, your first stop after the computer finishes the startup process should be the Device Manager, a piece of operating system software that helps you manage the various pieces of your PC. You can get to the Device Manager in several different ways: the two easiest are right-clicking on the My Computer icon on your desktop and selecting Properties from the context menu that pops up when you do this, or by going to the Start menu, going up to Settings, selecting Control Panel from the list of choices and then double-clicking on the System control panel. Either way, you’ll be presented with a tabbed dialog box; click on the tab that says Device Manager, and you’re there. The first thing to look for is a yellow exclamation point or red international no sign (you know, the circle with the slash through it) next to one of your devices. You may need to click on all the little plus signs next to each category of devices to see the full outline-like list of everything in your computer. If you see the yellow or red symbol, you know something is amiss. My first suggestion is to highlight the offending device and click on the Remove button. What this does is essentially erase the driver software associated with the device as well as references to it within the Windows 95/98/ME Registry. (The Registry tracks all the hardware and software you install, the preferences you set for each, as well as lots of other stuff). The fancy term for this is "logically" removing the device because, even though it may still be physically attached, the computer no longer has any record of its presence. Next, you should restart your computer, let Windows' Plug-and-Play feature "find" the device again, and redo the driver installation process by either using the driver

suggestions that Windows finds and makes or clicking on the Have Disk button and using the installer disks that came with the device. (If both options are possible, I’d go with the disks that came with the hardware—unless you’re sure they are an older version.) If the computer doesn’t find the device, then I would suspect the connections or the device itself. Buy another cable, unplug and re-plug the connectors, remove and then physically reinstall any new plug-in cards or do whatever you have to to ensure that the connection is solid. Once you’ve done that, you’ll find that many times, simply reinstalling the drivers solves the problem. Other times the problem is due to an older version of a driver conflicting with something else on your system. It’s always a good idea to check for and get the latest versions of drivers that your PC needs. Check the manufacturer’s web site first and if you can’t find the driver, or drivers, there (or if the company is no longer in business), try searching for it at one of the web’s driver repositories, such as The Driver Zone, WinDrivers.Com, WinFiles, or Frank Condron's World o' Windows. If you're running Windows 98 or Windows ME, you may also be able to get a new driver for your hardware via the Windows Update feature off the Start menu. If you find and download a new driver version, you can often update the driver by clicking on the device listing in the Device Manager, clicking on Properties, selecting the Driver Tab and then clicking on the Update Driver button. From there, depending on what version of Windows you’re using (the original Win95A, the revised Win95B, Win98, or Windows ME), you’ll either need to manually find where the file is located on your hard disk, and then select it, or the computer will attempt to find it for you. Either way, once the correct file is found, you initiate the update process by simply clicking a button. Some driver updates force you to create new floppy disks, which you then use to update the driver. Frankly, though it’s more work initially, this is a good solution in the long run, I think, because if you ever have to completely reinstall Windows (such as, when you buy yourself a big new hard drive), you’ll generally want to have copies of your drivers on floppies anyway. Mind your Ps and IRQs Finally, if that still doesn’t work, you may need to futz with the dreaded IRQs, or Interrupt Requests. Most hardware devices on your computer need attention from the processor on a fairly regular basis (to check their status) and the mechanism for doing that is called Interrupt Requests (because the device politely asks the processor to interrupt what it’s doing at the time and give it some attention—well, sort of). Because of the need to maintain compatibility with older hardware, today’s PCs are still limited to 16 IRQs (numbered 0-15), which is turning out to be a fairly big problem on many newer, well-equipped computers (read one of my InfoWorld Electric "Plugged In" columns to learn more on the subject).

The general principle with IRQ troubleshooting is that two devices cannot typically share an IRQ (an important exception is with some PCI-based add-in cards), and if they try to, one or sometimes neither of the devices will work properly. If you find that you have an IRQ conflict, where two ISA cards or other non-PCI devices are trying to use the same IRQ, you’ll need to change the settings on one of the devices to an open IRQ. The problem is, not every device is able to use every IRQ, so even though you have other IRQs available, the problem device may be incapable of using one of the open IRQs. If that’s the case, you may need to move another device using one of the IRQs that the problem hardware does work with first, and then free up an IRQ for the problem hardware. So, for example, if IRQ 8 is open but you have a SCSI card that only works with IRQs 9 or 11, you may first need to move whatever’s on 11 to 8, and then set the SCSI card to IRQ 11. Unfortunately, this can sometimes lead to an infuriating puzzle game where you try to match devices with your IRQs. To find out what IRQs are in use by your computer, double click the computer icon at the top of the Device Manager. You’ll see a new window pop up that shows which devices are using which IRQs. Unfortunately it doesn’t leave a blank for any IRQs that are not in use, so if you need to find an available IRQ you’ll have to look close, count the numbers and see if any are missing. If a number between 0 and 15 is missing, that means that particular IRQ is available. Some older plug-in cards require you to change the IRQs by setting tiny DIP switches on the card itself, or via a dedicated configuration utility. Most newer Plug-and-Play cards can be changed via Windows 95 or Windows 98. To do so, go to the Device Manager, highlight the product in question, click on the Properties button and then go to Resources Tab. Generally, you’ll have to deselect the Use Automatic Settings button to make any changes. Most devices offer several Basic Configuration choices, which is what you should try first. These are different combinations of IRQs, memory ranges, and I/O ranges. All you typically need to worry about is the IRQ. Some devices also let you adjust these parameters individually by clicking the Change Settings button. Thankfully, you rarely have to worry about IRQ problems if you’re using Windows 95, 98 or ME because they all do a pretty good job of automatically fixing them before they arise. In fact, this is one of their most important, yet little discussed improvements over Windows 3.1. Unfortunately, Windows 3.0 and 3.1 users still have to worry about this kind of stuff on a semi-regular basis. The problem is not unheard of under Windows 95, 98 or ME, however (I ran into it myself), which is why I’ve included it here. The End These tips won’t solve all the hardware problems you may run into, but they should solve a good number of them. The important thing to remember when doing any troubleshooting is that computers really are logical devices and there’s always a logical reason for why something isn’t working. Discovering what that reason is and then applying the right solution isn’t always easy (or intuitive), but if you think about the problem logically and work through it step-by-step, there’s a good chance you’ll be able

to solve it on your own. And, if worse comes to worse, you can always just nuke everything and start over (see "Starting Over: Repartitioning, Reformatting and Reinstalling" for more on exactly how to do that). Good luck!

Starting Over: Repartitioning, Reformatting and Reinstalling By Bob O'Donnell There’s nothing quite like the feeling you get when you first turn on a new computer and begin to use your unspoiled machine. Except in very rare occasions, it’s one of the few times you can be virtually guaranteed that everything will work, that the software already installed on the hard disk won’t crash and that you can get something useful done. Of course, this technological honeymoon never lasts terribly long, because you invariably install some new software, add new hardware, make some configuration changes or do something that—though it should work fine—eventually leads you down the slippery slope of seemingly inevitable PC problems. The desire to recreate that "fresh" feeling often leads people to start over with their computers by reformatting their hard drive(s) and reinstalling their applications from scratch. In fact, I’ve heard of several people who do this on a semi-yearly or even more frequent basis as a regular form of system maintenance. In addition, many computers now come with special boot floppy disks and installation CDs that are specifically designed to bring your system back to its pristine, shipped-from-the-factory state. Another reason for pursuing this strategy is that no matter how hard you may try, there are times when your system reaches a point where it’s simply not worth expending any additional effort trying to figure out why programs keep crashing or other strange problems keep occurring. I know that dedicated PC troubleshooters never want to give up, but one of the hardest lessons you can learn is that sometimes it really is better to start over. Now, I wouldn’t throw in the towel too quickly because starting from scratch is a fairly time- and effort-intensive project. But if you’ve tried the techniques I describe in the "PC Hardware Troubleshooting Tips," "PC Software Troubleshooting Tips," and "PC Startup Troubleshooting Tips," articles and have still been suffering through several difficult days, weeks or, God forbid, months of problems that just don’t seem to get any better, then you’re a good candidate for a fresh start. The problem is, nobody every really tells you how to make that fresh start. Oh sure, you hear the basics: "Just reformat and reinstall," but you don’t really ever hear exactly how the whole process is done. Well, worry no more, because this article will take you stepby-step through the process of starting over with your PC.

Step 1: Backup Before you even think about doing anything to your hard drive, you need to back up all your critical files. This means not only all your data files (you did organize them all in a single location, didn’t you?), but also those application files and other software pieces that took some time and/or effort to acquire. Included on this list should be updated driver software, applications patches, service packs, bug fixes and any other enhancements that you’ve downloaded off the web (and don’t have available on CD or in some other handy form). I also recommend you save your browser bookmarks which, if you're using Internet Explorer, can be found in the Windows/Favorites folder. One file that's commonly overlooked (because it isn't stored in an obvious place) is your Outlook or Outlook Express e-mail file. The easiest way to find it and back it up is to search for *.pst off the Start menu. All Outlook files use the .pst extension and you can be sure to find yours this way, even if it doesn't have the default name of Outlook.pst. Generally speaking, your Outlook file should be in the C:\Windows\Application Data\Microsoft\Outlook folder. This is important to know because when you reinstall, you need to copy your .pst file back to this same directory. In addition, don't forget to write down all your network settings from any network log-in you have, as well as those found in the network control panel. If you have dial-up networking connections, remember to right down the settings for each of those as well. When you go to re-establish your network settings, you'll be awfully glad you did. You’ll probably run into a problem with applications that automatically update themselves over the web, because they don’t necessarily have an easy way to find the update files they’ve downloaded. If that’s the case with some of your applications, you’ll probably have to simply let the application "re-update" itself after you re-install it. Windows 98 or Windows ME updates that occur via the Windows Update feature may also present this problem, although you might be able to find them in your Windows directory in a hidden folder called msdownld.tmp (at least, that’s where they were on my machine). To view hidden files, open Windows Explorer, select Folder Options from the View menu, select the View tab, and click on the Show all files radio button. You don’t need to back up all your applications because you can more easily install them off their original CDs. In fact, part of the point of this exercise is to re-install your applications so that all the right files get put in all the right places. For this reason, I also don’t recommend that you make a complete disk copy, or disk image before you do a reinstall of all your software. If you do, and then you restore that copy, you could end up with the same types of problems that led you to take on this procedure in the first place. Just back up what you need. Step 2: Create a Boot Disk

The next step is to create a bootable floppy disk that includes all the programs you’ll need to get the next few steps. I cover how to do this in my "How to Create a ‘Real’ Windows 95 (or 98) Boot Disk" article. One additional point I’ll add here is that you need to make sure both the Fdisk.exe and Format.com DOS utilities are on your newly created Windows 95 or Windows 98 boot floppy. If they aren’t (the standard Windows 98 floppy still needs Format.com), you may need to copy them over from your hard drive onto the boot floppy—you’ll find them both inside the Command directory inside your main Windows directory. (In fact, Windows/Command is where you’ll find all the important DOS-based utilities.) One other option for Windows 98, Windows 98 2nd Edition and Windows ME users is that the Windows 98/ME CD is bootable, meaning it has all the necessary files to start your computer stored in the right places, much like a boot floppy disk. Your computer has to support booting from the CD-ROM and you have to enable this feature (which you do in your computer’s BIOS or CMOS Setup program) in order for this technique to work, but it can be a handy option. If you’re unsure whether or not your computer supports this, look for a reference to the El Torito BIOS standard—which this feature is sometimes called—or look around in the Boot Options section of your computer’s BIOS Setup program. Also remember that after you’re done with this procedure you’ll want to change this BIOS setting back to booting from your floppy drive and hard drive (usually in that order). Whether you go with the floppy or the CD, be sure you try it out at least once before you begin the partitioning process. The next step in this process will erase all of your computer’s data, so you want to be sure your computer boots from the disk/disc before you continue. Step 3: Partition and Reformat The crux of the process occurs here in Step 3. The first part of this step is called partitioning your hard drive and it’s usually done with the DOS-based Fdisk program bundled with all versions of Windows. (Some people prefer third-party partitioning programs such as PowerQuest’s powerful PartitionMagic or QuarterDeck’s Partition It or Partition It Extra Strength for this process.) Partitioning involves organizing a single hard drive into logical chunks called partitions, as well as setting an overall file structure to be used on each partition, such as FAT16 or FAT32. The second half of this step is called reformatting and it basically wipes any existing data from each partition and prepares the partition to accept new files. (By the way, this is not the same thing as a true low-level hard drive format—these days that can typically only be done—and should only be done —at the factory.) Formatting is done with the DOS–based Format program, or simply within Windows itself, just as you do with a floppy disk. Before getting into specific steps, you need to know a bit more about partitions, such as the fact that there are two main types: primary and extended. The most important difference between them is that primary partitions can be used to boot your computer and extended partitions cannot. In addition, unlike primary partitions—which actually hold

data—extended partitions are themselves just containers for yet another kind of structure called logical DOS drives. So, for example, you might find that your hard drive is divided into one active partition and one extended partition and the extended partition contains two logical drives "inside" of it. Each active partition and logical drive uses its own drive letter (i.e., C:\, D:\, E:\, etc.) and operates independently, so with multiple partitions, a single hard disk may "look" like multiple drives. In reality, however, it’s just one physical disk that’s organized into different containers. Of course, if you have multiple hard disks inside a computer, each of them uses a drive letter as well, so when you have multiple partitions on multiple disks, things can get kind of confusing. If you want to run multiple operating systems on your PC—such as Windows ME and NT, or Windows 2000 and Linux—you often need to have multiple primary partitions. In some instances, such as with Windows 95 and NT, it's possible to have just one primary partition with two operating systems, but both operating systems need to be able to understand the partition scheme--such as FAT or FAT32 (see below for more)--for this to work. The maximum number of primary and/or extended partitions you can have is four, but be aware that only one primary partition can be active (and therefore "visible" to the rest of your system) at once. On the other hand, other than the 26-letter drive limit—which does exist—there are no restrictions on the number of logical drives that you can have within an extended partition. More importantly, multiple logical drives within an extended partition can be used and visible on your system at once. So, for example, if your system has an extended partition with two logical drives and one primary partition (you always have to have one of those), you would be able to see all three drive letters at once. On the other hand, if you have two primary partitions and one extended partition with two logical drives, you might only see three drive letters because the other primary partition and any data or programs stored on it would be invisible if the two primary partitions were completely different types (such as Ext2 for Linux and NTFS for Windows 2000). Again, if both operating systems "understand" the same partition type, then you might be able to see all four drive letters. In many cases you’ll want to keep your entire disk as a single primary partition—and, therefore, single drive letter—although there are some cases where you can’t. Specifically, if you have a hard drive larger than 2 GB and you’re using the original version of Windows 95 or Win95A, you’ll have to use multiple partitions because of limitations in Win95 itself. (To find out what version of Windows you’re using, open the System Control Panel and look in the upper right portion of the General Tab. You should see a reference to Windows 95, 95A, 95B, 95C, 98, ME and 2000 underneath where it says System.) Other limitations you may run into on disk size limits may be as a result of your computer’s BIOS. Some older BIOS’s had a hardware limitation of around 2.1 GB (some

newer ones are limited to 8.4 GB), that prevents them from working with larger drives, but that can usually be fixed with a BIOS update. Check your computer manufacturer’s or motherboard manufacturer’s web site, or you can also try the Micro Firmware or Mr. BIOS sites. If you have both an updated BIOS and Windows 95B (sometimes called OSR2) or later— including Windows 98 or Windows ME—then you can take advantage of the FAT32 (File Allocation Table 32-bit) file system and have a partition (or even multiple partitions) larger than 2 GB. Without going into too much detail, the basic reason for this is that FAT32 is able to keep track of a much larger number of individual file elements than the older FAT16 file system (which is more commonly referred to just as FAT). This translates into the ability to work with larger partitions. Before you actually begin the partitioning process, you need to decide how you want to partition your drive—if you want to keep it all as one big drive, or if you want several different partitions/logical drives with one for data, one for programs, etc. In addition, if you plan to try out or regularly work with multiple operating systems (OS's), you’ll have to plan for that at this stage. You’ll also need something called a boot loader if you install multiple OS's—one comes bundled with PartitionMagic and another comes with Windows NT 4.0 and Windows 2000. A boot loader is a program that lets you decide which primary partition to make active at boot-up. The OS that is loaded from the active partition is the one that gets "control" over the machine for that particular session. Once you have a basic strategy figured out, you can move onto the specific steps. The following describes how the process works with Fdisk. (If you’re using PartitionMagic, or some other utility, you’ll have to follow different steps, but the concepts will be similar.) First you need to boot your computer with your boot floppy and then launch the Fdisk program as soon as you get to the A:\ prompt. To do that, just type in Fdisk and then hit Enter or Return. If you’re running the version of Fdisk that comes with Windows 95 OSR2 or later (including 98, 98 SE, or ME), you’ll first see a kind of obscure text message and question about having support for large disks to which you answer yes or no. Though there’s no specific mention of it, this question is asking whether or not you want to use FAT32. If you answer yes, you’ll get FAT32 and if you answer no you’ll get a FAT16-formatted drive. (Of course, if you have Windows 98, you can convert from FAT16 to FAT32 with the bundled FAT32 Driver Converter after the fact. If you have Windows 95, however, you’ll either have to start all over again to switch to FAT32, or purchase a third-party tool such as PartitionMagic.) Once you’ve answered the question, you’ll be presented with four numeric choices from which you can create a new partition, delete an existing partition, make one of the partitions active or get more information on the current partitions you have. In general, I’d recommend selecting option 4 first to get more information about your current partitions.

If you’re going to switch from multiple partitions to a single partition or if you want to adjust the size of your current partitions, you’ll first need to delete all but the primary active partition. Before you can delete an extended partition, however, you first need to delete any logical drives that are inside the extended partition. To make any of these deletions, select option 3 off the main Fdisk menu and follow the directions. You’re always given a warning before you do anything destructive, so if you take your time, you shouldn’t run into any serious problems. To create new partitions or logical drives or to resize the remaining primary partition, select option 1. If you want to use logical drives, you first need to create an extended partition to hold them and then you can create the logical drives. In all cases, you’ll need to know how large you want the partitions and/or logical drives to be in megabytes, so do your math ahead of time. If you’re resizing a single partition, simply make the partition the same size as the available disk size. Also remember that all hard drives use a certain amount of space for disk overhead so don’t get upset when your new 8 GB hard drive (or whatever size you have) doesn’t have eight full gigabytes (or whatever its advertised capacity is) for creating partitions. Once you’ve finished your partitioning, you can exit from Fdisk by simply hitting the Esc button at the main Fdisk screen. As the ensuing screen says, you’ll have to restart before the changes take effect and before you can re-format the newly created or resized partitions. By the way, if you opt for something like PartitionMagic, you’ll find the partitioning process more intuitive and more flexible than what Fdisk provides. For example, you can resize partitions graphically without having to first delete them, and you can easily switch a particular partition back and forth between FAT16 and FAT32, among other capabilities. Regardless of how you partition the drive, however, the re-formatting process is very simple. Once again, you’ll need to restart the computer with the boot floppy installed and when you get to the A:\ prompt, type in: Format C: /s What this command does is reformats the active primary partition on your main drive—in other words, it reformats your hard drive. The /s switch at the end of the command tells the computer to also copy over the basic DOS system files to the hard disk so that you can then restart from the hard disk and boot to a C:\ DOS prompt if you want. To continue onward with Step 4, however, you’ll probably want to boot from your boot floppy. Step 4: Reinstall Now that the hard part is over, it’s on to the drudgery of re-installing everything. Of course the first thing you’ll need to do is re-install the operating system from scratch.

To do that, after you start your computer with your boot floppy inserted into the floppy drive, you'll need to make sure you have the Windows 95, 98 or ME CD in your CD/DVD-ROM drive. Once it's there, just type the following at the A:\ DOS prompt you should see when the boot process finishes. Hit the Enter key at the end of it. (Note that you may have to type a letter other than "D" if your CD/DVD-ROM is assigned to a different drive letter.) D:\Setup.exe On a freshly formatted drive this process should go smoothly, but be prepared with any drivers or driver upgrades you have available on floppies or CDs. As Windows goes through the Plug-and-Play process of detecting your computer’s hardware and then attempting to install drivers for it, the OS should give you very clear signs whenever it needs input (or disks/discs) from you. If you want to, you can create a directory on your hard drive called Win95CAB or Win98CAB and then copy all the compressed .CAB (or cabinet) files you’ll find on the Windows 95/98 CDs (in the Windows 95 or Windows 98 folders respectively) into those directories. It takes a fair chunk of hard disk space—around 100 Mbytes or so—but it saves you from having to look for your Win 95 or Win98 CD down the road if you ever install anything and the installation process asks for the CD. Instead, you can just direct it to the CAB files on your disk and you’ll be all set. Thankfully, Windows ME does this for you automatically. If your PC comes with a program that automatically returns it to its factory fresh state, you’ll use that to install your OS instead. Doing so should automatically take care of installing the OS and applications that came with your computer. If you have any updated drivers or applications as part of your backup, however, you’ll have to re-install those manually, as explained a bit further down. If you have trouble during the installation, it could be that one of you drivers is out-ofdate. If so, you’ll want to check the manufacturer’s web site for an update (see the "PC Hardware Troubleshooting Tips" article for more). Once the OS installed, you should run any other driver installation programs you have. Occasionally these types of programs will tell that you need to reboot for the changes to take place. When you’re going through this re-install process I highly recommend you take their advice for each program that requires it. Even though constantly rebooting adds even more time to the process, it can be worth it in the long run. The reason is if you install multiple pieces at once that make changes to your system, those changes could conflict or counteract each other. Because the purpose of this exercise is to get everything working properly, you’re better off taking the conservative route here and letting each piece "take hold" one at a time. Once all your drivers are done, it’s time to reinstall the apps. Again, if at the end of the install the program says it needs to restart Windows for the changes to take effect, I would restart. The order that you install the applications in typically doesn’t matter, although I would probably install any that had been causing you problems first. Once the

main apps are in place, you need to reinstall all those lovely Service Packs, bug fixes and other updates that you painstakingly backed up in Step One. Remember also that some updates and Service Packs can only be done after a previous update to the same program has been made so make sure you do them in the proper order. Before copying over your own data, I suggest you try running a few of your favorite applications to make sure everything is working properly. In addition, make sure you double-check any previously problematic programs once everything has been installed. If a problem crops up now, it’s probably due to a software conflict with another application on your system. If that’s the case, you’ll need to check web sites for updates and see if that helps (see the "PC Software Troubleshooting Tips" article for more). Finally, after all the applications have been installed, it’s time to copy back over all your own data. If you haven’t already, I suggest you take advantage of this re-installation process and use the opportunity to organize all your data files in a single location, such as the My Documents directory. You don’t want to put everything at the main level of the My Documents directory, however, or you’ll be overwhelmed. Instead, to make that directory useful, you should first create sub-directories inside it and then use those directories to store your various types of files. Step 5: Enjoy When everything has been restored, it’s time to enjoy your new machine. Well, almost. Though it shouldn’t make any difference, it’s probably worthwhile to double or triplecheck any problem applications (you know—the ones that led you to take on this procedure in the first place) after you restore your own files. Once you’re confident that things are working well, you can take yourself and your PC back on a second (or third or fourth) technological honeymoon and get to know each other all over again.

Creating a "Real" Windows 95 (or 98 or ME) Boot Disk By Bob O'Donnell Though floppy drives and floppy disks are increasingly falling out of favor, there are many occasions when a single floppy disk can make the difference between being able to use your computer or not. I’m referring specifically to a boot disk, which is a floppy disk that allows you to boot, or startup, your computer without having to access the hard drive. A boot disk contains all the necessary operating system files your computer needs to get started. Boot disks can

be tremendously important for many different applications and, frankly, no computer user should be without one. For example, if your hard drive fails, or if you have the misfortune of contracting a computer virus that attacks your hard disk’s master boot record (as many of them do), or if some of your important operating system files get accidentally deleted or otherwise corrupted, a boot floppy will be your savior. (For more advice on addressing on common boot problems, see the "PC Startup Troubleshooting Tips" article.) In addition, if you ever want to reformat your hard disk or reinstall Windows 95 (or Win98, for that matter) from the CD, you’ll also need to have a boot disk. (For more on this issue, see "Starting Over: Repartitioning, Reformatting and Reinstalling.") Some older DOS games also work better if you boot your computer from a startup disk. In recognition of this importance, Microsoft made the process of creating a basic Windows 95 boot disk very easy. All you have to do is open the Add/Remove Programs Control Panel. To do that, click on the Start button, go up to Settings, select Control Panel from the list of available choices and when the Control Panel window opens, just double click on Add/Remove Programs. Once that's open, click on the Startup Disk tab and click on Create Disk… You’ll be prompted to insert the Windows 95 CD into your CD-ROM drive (in most cases) and to put a floppy disk into the floppy drive. After a minute or two, you’ll have a disk that includes all the critical files necessary to start your computer and "see" your hard disk. The basic boot disk also includes important disk utilities such as Scandisk.exe, Fdisk.exe and Format.com. Unfortunately, that disk will not be able to "see" your CD-ROM, which means you won’t be able re-install Windows 95 from a CD, run a CD-based DOS game or do any of a number of other possible applications. If you boot with this disk and try to switch to the CD you’ll get the heart-warming "Invalid Drive Specification" error message in DOS. The "simple" answer to this problem is to just install your CD-ROM driver onto the disk —that's the answer most magazines and books will give you. The problem is, they don't tell you how to do that. (By the way, Microsoft clearly saw the errors of its ways here because the regular Windows 98, Windows 98 Second Edition (SE) and Windows Millennium Edition (ME) boot floppies—which you create in the exact same way as you do under Windows 95— do include a CD-ROM driver. In fact, they include both a generic IDE/ATAPI driver as well as a generic SCSI CD-ROM driver. Even better, a boot floppy created under Windows 98 or Windows ME will boot a Windows 95 machine without any problems or conflicts at all. Now, why Microsoft couldn't have done this in the first place is beyond me....) Before I get to the specifics of how it's done with Windows 95, let me give just a bit more background. In order for a CD-ROM drive or any other device (such as a sound card) to be "visible" in DOS, you have to tell the operating system that it's there. You do that

through specific pieces of software called device drivers, or just drivers for short, that "communicate" between your operating system and these devices (such as the CD-ROM drive). There are different types of driver files, but for a boot floppy you need a "real mode" or 16-bit driver for your CD-ROM drive. ("Real mode" refers to a type of memory allocation system used by DOS and early Intel processors.) In order for DOS drivers to work on a boot floppy you need to both have copies of the appropriate drivers on the disk and put references to those driver files in some of the operating system startup files also found on the startup floppy. Without those references, the operating system doesn't know the drivers are there and won't be able to load them. And if it doesn't load them, well, we're back where we started—an "invisible" CD drive. In the case of a CD-ROM, you need two driver files, one that's specific to your machine and one that's included with the standard Windows 95 installation. The file that's specific to your machine is the "real mode" or DOS driver I referred to earlier. The startup files you need to adjust are Config.sys, which is included on the standard boot floppy, and Autoexec.bat, which is not. If you also wanted to add a sound card driver to this floppy in order to play games, you would copy over an additional file and then make additional changes to the existing Config.sys startup file. For the purposes of this article, I’m only going to describe how to install the CD-ROM driver for an IDE/ATAPI CD-ROM drive, which is what most computers have. The same basic principles apply to SCSI CD-ROMs and sound cards, although you may need to make some additional alterations to the startup files to get them to work properly. Check the documentation that came with your sound card or with your PC for more details. OK. Now it's time to describe how to create your CD-ROM driver-equipped boot disk. As I explained, the first thing to do is to find a copy of the real-mode driver for your CDROM. Note that this is different from the 32-bit Windows 95 driver that your CD-ROM uses while running under Windows 95. Most 32-bit drivers are called virtual device drivers and have the file extension .vxd, whereas many real-mode CD-ROM drivers end with a .sys at the end of their file name. For example, the real-mode CD-ROM driver for the Dell Computer I’m currently using, which has an NEC CD-ROM, is called nec_bm.sys. On some computer systems you can find the real-mode driver on a separate floppy that came with the computer (or with the CD-ROM drive itself, if you added it separately), but in many cases, particularly with newer computers, you’ll find that you don’t have it. In that case, you should first check the computer manufacturer’s web site, then call them if you can’t find it there. If you know who made the CD-ROM drive mechanism used in your computer, then you can use the same techniques with that company.

If you don’t know who made the drive you can often find out by typing in the FCC ID, which has to be on a sticker somewhere on the drive, into a special FCC database (linked here) that tracks all that information. Every company that sells a computer product has to have a unique ID, so by typing that ID in, you can find out the name of the company that sold the drive, as well as an address and a phone number. Once you’ve found out who made the drive, another possible source for drivers is one of the many driver sites here on the Web, such as The Driver Zone, WinDrivers.Com, WinFiles, or Frank Condron's World o' Windows. After you copy the CD-ROM driver over to the startup floppy, you’ll also need to copy over a Microsoft-supplied system-level CD-ROM driver called MSCDEX.EXE. You should find a copy of it in your Windows/Command folder on your C: drive. If it's not there, you'll definitely find it on the Windows 95 CD that came with your computer. The next step is to open the Config.sys file and add one line to it. To do that, you'll use the DOS Edit application that’s installed on the boot floppy as part of the standard Create System Disk routine. To get to the application, open an MS-DOS window by going to the Start menu and selecting MS-DOS Prompt from the Programs menu item. After the window opens, switch to your A: drive by typing A: and then Enter. Once you get an A:\> prompt, type in "edit" (without the quotation marks) and hit Enter and you’ll be presented with a basic text editing program. Open the Config.sys file on your A: drive. By the way, make sure you’re not editing the Config.sys file on your C: drive or you could have big problems. You should be able to easily tell because the basic Config.sys on the standard boot floppy has only one line in it and it refers to Himem.sys. Put your cursor underneath that line and type in: Device=A:\ /D:MSCD001 Don’t actually type —that needs to be replaced with the exact spelling of the driver file you previously copied onto the floppy disk. So again, in my case it was: Device=A:\nec_bm.sys /D:MSCD001 Save the Config.sys file and close it. The next step is to create a simple Autoexec.bat file on the floppy. While you’re still in the Edit application, go to the File Menu and select New... Then type in the following line:

MSCDEX.EXE /D:MSCD001 When you're done with that, go up to the File menu, select the Save As… command and save the file as Autoexec.bat (capitalization doesn’t matter) on the A: drive. Exit the Edit program and then close the MS-DOS window by clicking on the X in the upper right hand corner. That’s it. Now you need to try it out. To do that, insert the boot floppy (you did put a new label on it, didn’t you?) into the floppy drive and restart the machine. After the normal BIOS messages you should see some startup messages about MSCDEX and a few other things. Once you get the A:\> prompt, try switching over to the CD-ROM drive to make sure the driver worked by typing in D: (or whatever letter your CD drive is assigned to) and hitting return (make sure you have a CD-ROM disc of some sort in the drive—it doesn’t matter what). If you get the D:\> prompt, try typing a DIR (or directory) command to make sure you can see the contents of the drive. You should see a list of all the files and directories on the root level of the CD you have in the drive. If so, you’ve successfully created a bootable floppy disk with a CD-ROM driver for your system. Congratulations!

Mac Software Troubleshooting Tips By Bob O'Donnell These days, many people try to claim that Windows 95, 98 and/or Windows ME are "just  as easy" as the Mac, but as soon as you scratch below the surface of each operating  system, you’ll quickly discover that the Mac OS is still the better designed choice. Nowhere is this more apparent than when it comes to troubleshooting common problems.  Windows continues to show its ugly DOS­based history or some other arcane  architectural flaw as soon as you start trying to fix something that goes wrong on your PC  (see the articles "PC Software Troubleshooting Tips" and "PC Hardware Troubleshooting  Tips" for more). The Macintosh, on the other hand, is generally much less prone to  problems and those problems that do arise tend to have more straightforward answers. But that doesn’t mean problems don’t happen on the Mac because they still do. In fact, for  some people, they are maddeningly common. Sometimes the issues are hardware­related  (see the accompanying "Mac Hardware Troubleshooting Tips" for more), but most of the  time they have to do with software problems.

Staying Up to Date One area where the Mac really isn’t that different from PCs is in the need to keep your  software up­to­date. New application updates, operating system patches, and other types  of upgrades are often the key to solving frustrating problems or just some annoying  glitches. So, before you get too far into the troubleshooting process, take an inventory of  your software and hardware and check to see if any updates are available online from the  web sites of the companies whose products you use. One site you’ll certainly want to check is Apple’s Software Updates page. In addition,  don’t forget to check on any extensions you have installed, particularly ones that function  as drivers for any hardware you’ve added to your Mac.  Extending the Mac’s Reach Speaking of extensions, the vast majority of problems that pop up on Macintoshes are due  to what are called extension conflicts. Before I get into what that means, let me tell you a  little bit about how the Mac OS works. When Apple designed the Mac’s operating system they knew they wouldn’t be able to  provide all the features that people wanted, so they also built in a mechanism that lets  Mac users add to or extend the capabilities of the OS very easily. This mechanism  consisted of files called extensions that could, well, extend or alter the features of the  operating system. By placing these files into a special Extensions folder inside the System  Folder, they are automatically loaded and run whenever the Mac starts up. The only time most Mac users even become aware of extensions is during the startup,  when the little icons representing different extension files dance across the bottom of your  Mac’s screen. (Not all extensions have startup icons, by the way, so don’t presume that  because you counted 8 or 10 different icons during the startup, or boot, process that’s all  you have. Most Macs have at least 40 or 50 extensions and many have 100 or more.) The way that many—though not all—Mac extensions work is that they are loaded into  memory along with Mac OS at startup time and stay there until you shut down. So, for  example, if you have an extension file like Adobe’s Type Manager, it adds its type­ rendering features to the operating system as your Mac starts up, and stays there as an  "extension" to the operating system for as long as your Mac is on. The benefit of this arrangement is that any and all programs that run on your Mac and  need access to the features offered by the extension can get to them by interacting with 

the operating system, as they normally do. The downside, however, is that some programs  don’t operate correctly when used in conjunction with certain extensions or certain  versions of specific extensions. In addition, some extensions step on each other’s toes by  providing some of the same types of changes to the Mac OS. Others simply make changes  in a way that conflicts with each other. Finally, the last drawback of this arrangement is that if you’re using applications that can’t  or don’t take advantage of the features offered by certain extensions, you’re wasting  precious Mac memory. The reason is, once an extension has been loaded into memory it  stays there and continues taking up space until you shut down your Mac. As we move  towards larger and larger amounts of RAM this concern is becoming increasingly less  important, but it’s still there. By the way, if you ever want to see just how much memory your extensions use, try this  simple procedure. First, do a normal Mac startup and when you reach the desktop, go up  to the Apple menu and select About This Computer (or About This Macintosh or  whatever the message on your machine happens to be) to see a display of how much  memory each application on your Mac is currently using. Make note of how much the  Mac OS, or System, is using. Now, restart your Mac holding down the Shift key until you  see a message about No Extensions (I’ll explain what this means in a bit) and when you  get to the desktop, go to the Apple menu again and select the same menu item. You  should see a dramatically lower amount of memory being used by the Mac OS, or  System, in this case. The difference between the two is the amount of memory that your  extensions are using. Extensions Manager to the Rescue So, what’s the trick to solving Mac extension conflicts? Using the Extensions Manager  that’s been built into the Mac OS since System 7.5, or an alternative commercial utility  such as Casady & Greene’s popular Conflict Catcher. Either program lets you selectively  turn off individual extensions or groups of extensions at startup time so that you can try  and figure out what’s causing the problem. They also let you change the order in which  the extensions are loaded, which can also make a difference. Some extensions need to be  loaded last or near the end in order not to conflict with others. To get to Extensions Manager (which is actually a Control Panel), you can hold down the  space bar while you’re starting your Mac, or choose Extensions Manager off the Control  Panel submenu under the Apple menu while your Mac is already running. Conflict  Catcher lets you assign a hot key to launch it during startup but it also defaults to the 

space key. Essentially what you do with either utility is select a group of extensions that  you want to keep "on" and a group that you want to turn "off" and then after your  computer starts with the set of extensions you’ve chosen, you try to recreate the problem  you were having.  In some situations, extension conflicts can keep your Mac from booting at all. If you only  make it part of the way through the startup process and then your computer freezes up,  then that’s the classic sign of an extension conflict. (A partial boot can also be a symptom  of a hardware­based SCSI problem, though, so check the "Mac Hardware  Troubleshooting Tips" article for more.) If it does turn out to be an extension­related  problem, you’ll know you’ve found the problem when the Mac makes it through the  startup phase and brings you to the desktop. Some extension conflicts don’t occur during startup, however. If you’re having trouble  running certain applications or groups of programs, you’ll want to check those after  you’ve selected a different set of startup extensions. If the same problem occurs (or if  your Mac still can’t boot), then you haven’t found the culprit and you’ll need to select a  new set and go through the process all over again. As you might expect, finding extension conflicts can be very slow and tedious (aren’t  computers grand!), but Conflict Catcher makes it a bit easier by automating the testing  and restarting process to a certain degree. By the way, in addition to turning off certain extensions to avoid conflicts, you can get  more memory for your applications by turning off extensions that you aren't regularly  using. As the above example shows, extensions can take up a lot of memory and turning  off a bunch of them (or even sometimes just a few of them) may give you several  megabytes of RAM space back. Giving Yourself Enough Room to Work Speaking of RAM, another problem that can occur with Macs is running out of memory.  Like PCs, Macs take advantage of something known as virtual memory, which basically  uses some space on your Mac's hard drive to "fake" the system into thinking it has more  real physical RAM than it actually does. So, for example, a system with 32 MB that uses  virtual memory might appear to the MacOS as having 64 MB of working room. You can  turn virtual memory on and off via the Memory control panel. In most instances you'll  want it on, but there are some older applications that won't work properly with it turned  on, so you should check your Mac's settings and adjust them as necessary.

A practical benefit of virtual memory is that it gives your Mac enough room to open up  multiple applications. However, it doesn't necessarily impact how much memory each  application has to use, nor does it automatically increase the amount of memory allocated  to each application as needed. Instead, the amount of memory allocated to each  application is determined by an obscure setting in the application program's Get Info box.  To find this, click on the application you want to check in a Finder window and select  Command (the clover key)­I or select Get Info from the Edit menu. (You have to do this  on the original application file, not an alias.) Inside the ensuing dialog box you'll see a reference to the minimum amount of memory  required by the application to run, as well as a preferred amount. If you increase both  those numbers (they're often in Kilobytes, so divide by 1,024 to figure out the amount in  Megabytes), you'll increase the amount of memory the application has to do its own work.  This can be very handy if you're working with large or complex documents because if the  application doesn't have enough room to do its work, you may get an error message  saying there isn't enough memory even if you have lots of RAM installed and it's the only  program running. You don't want to increase the amount too much, however, because these numbers affect  how much memory the MacOS sets aside for the application. If the application has way  more than it needs, you'll just be wasting memory. So, the trick is to find an amount that  lets you work without getting out­of­memory messages, but doesn't take more than  necessary. By the way, if you have only one or two problematic documents, you can  always just adjust the application's memory allocations­­which is what these numbers are  called­­right before you work with them, and then change them as soon as you're done.  You can't change them while the application is open, but otherwise it can be done at any  time. Problems with Preferences Another fairly common problem on Macintoshes has to do with preference files, which  are files used by most applications to store certain settings they need to operate properly.  If you’re having problems with a specific application, it may be due to a bad or corrupted  preferences file. There’s no easy way to tell by simply looking at a preferences file whether it has a  problem, but there is a simple procedure you can try if you suspect a problem. Just open  the Preferences folder inside your Mac’s System Folder, look for the preferences or  "prefs" file for the particular application, and drag it into the Trash. Most applications 

will automatically create a new preferences files the next time you start them if they can’t  find an existing one in the Preferences folder. In some cases you may have to reset any adjustments you had previously made to the  program’s preferences or settings dialog box, but that’s a small price to pay to get your  application working again.  Viruses and Other Issues Mac users were very fortunate for many years because the Mac OS was fairly free of  computer viruses. Recently, however, that situation has changed and now Mac users also  need to be on the lookout for viruses (such as AutoStart Worms), Trojan Horse programs  (which hide bad things in a seemingly innocent package—hence their name), and other  nasty stuff that can wreak all kinds of havoc on your Mac. If you’re encountering strange problems, such as a Desktop Print Spooler that keeps  launching and won’t go away, you may have contracted a virus. To fix that problem and/or  avoid getting any viruses in the future, make sure you purchase and install an anti­virus  program such as Norton AntiVirus or Virex and, most importantly, keep the virus  definitions up to date. New computer viruses are being developed and unleashed at an  alarming rate and even if you have the most recent version of an anti­virus program, you  may still be at risk if the virus definitions—which is what looks for and erases specific  viruses—aren’t current. If you come across generic Mac error messages that have specific numbers in them, you  might be able to find more information about them in this article from Apple’s Technical  Information Library. The article is a bit out of date, but you still may find it helpful if you  want to find out about certain types of specific error codes. If you’re still coming across  Type 11 errors on your Power Mac, you might want to read this article (although updating  to a current version of the MacOS would probably help as well). Finally, if all else fails, there are several excellent Mac­specific troubleshooting resources  available here on the Web. Apple’s Tech Exchange and previously mentioned Technical  Information Library, for example, are the company’s answer to Microsoft’s extensive  Knowledge Base, and includes all kinds of articles describing specific problems and  offers solutions to those problems. The Basic Troubleshooting Article from the Tech  Exchange, in particular, offers some great general troubleshooting advice. You may also  want to check MacFixIt, Macintosh Software & Extension Conflict Troubleshooting, or  other Mac sites you see listed on my Troubleshooting Resources page.

The End There’s no question that the Macintosh and the Mac OS are great products, even if they  aren’t quite perfect—yet. If you do run across problems, bear in mind some of the  previous tips, use some of the available resources and you should be able to keep your  Mac in top working form.

Mac Hardware Troubleshooting Tips By Bob O'Donnell One of the Mac’s greatest strengths is its ease of use. A related, but not as widely touted benefit is that Macs are pretty easy to troubleshoot, especially when compared to Windows-based PCs. First of all, because Apple designed most of the hardware that goes into Macs and Mac clones as part of an integrated system, Mac hardware problems are much less frequent than Windows PC-based problems. Second, Mac hardware problems are often less complicated—loyal Mac enthusiasts will never have to worry about confusing subjects such as IRQs and the like. That’s not to say that Macs don’t have any problems, though, because they do. (For more info on software-related Mac problems, see the "Mac Software Troubleshooting Tips" article.) And when you run into hardware problems on a Mac, they can be just as frustrating as any problem you’ll run into on a PC. Take the First Step As with PCs, or any electronic device for that matter, the first thing to do any time you encounter what appears to be a hardware problem is to check the obvious stuff. Look for loose or unconnected cables—you’d be surprised how often a quick jiggle to the cable fixes a problem. Look at the cable connectors as well, particularly for things like bent pins. I recently solved a modem problem by noticing that the adapter for my PowerBook’s PC Card modem had a pin that was bent so far out of the way that I could still connect the adapter to the modem—but of course it didn’t work. Once I bent the stray pin back into place and reconnected the adapter, everything was fine. If the problem is with a new PCI plug-in card, try pulling it out and then reseating it back into the slot. The card needs to fit very tightly or it won’t work properly. Accelerator cards or processor upgrades can also be problematic on Macs, or other MacOS compatibles. Unfortunately, the problems here are usually due to incompatibilities between these cards and either the MacOS or individual applications—which means there really isn’t anything much you can do about them other than finding out if you can somehow upgrade your accelerator or processor upgrade card.

If you've just added new RAM and either the system won't boot or won't recognize the additional memory, either you need to pull out and re-seat the DIMMs (or SIMMs) or you may have to try RAM made by a different manufacturer. While all RAM should work the same, the reality is that tiny differences in design can cause a particular manufacturer's RAM to not work in a particular Mac model (or PC, for that matter). The Skinny on SCSI Probably the most common Mac-related hardware problem has to do with SCSI (pronounced "scuzzy") conflicts. Because the Mac has had a SCSI port since it’s earliest days, SCSI peripherals have always played a big role in MacOS systems. Hard drives, CD-ROM drives, scanners, removable drives and more are all available as SCSI devices for Macs, both in internal and external versions (because virtually all Macs have both internal and external SCSI connectors—even newer Macs that use IDE hard drives). SCSI problems often manifest themselves on startup--often times your Mac won't be able to boot because of them--but they can also show up while using a particular application or trying to access a SCSI device (such as copying a file to an external drive, scanning in a photo from a SCSI-based scanner, etc.) To completely understand the issues involved, you need to know a little more about what SCSI is and how it works. SCSI is a high-speed, bi-directional interface that allows you to connect up to 7 different devices to a Mac. Devices of connected components are referred to as SCSI chains and all SCSI chains begin at the SCSI controller. On most Macs, the SCSI controller is a chip on the motherboard. Some Macs include two of them: one for internal components and another one (which often runs at a different— typically slower—speed) for external devices. The controller chips support one of several varieties of SCSI: normal SCSI, Fast SCSI (sometimes called SCSI-2), Wide SCSI, Fast and Wide SCSI, Ultra SCSI (or SCSI-3) and Ultra Wide SCSI to name a few. These various flavors of SCSI differ primarily in how fast they’re capable of sending data down the line, although Ultra SCSI also ups the maximum number of supported devices on a single SCSI chain from 7 to 15. Regardless of the type of SCSI capability that the controller supports, however, you can attach any SCSI device to any SCSI chain because all SCSI components can work together without any problem. However, if you attach a fast device capable of supporting the 40 MB/second transfer rates of Ultra Wide SCSI to a Fast SCSI controller (the kind typically found on today’s Macs, which offers a maximum transfer rate of 10 MB/sec), the attached device will not be able to operate at its peak performance. This is because the maximum speed of any device on the chain is determined by the SCSI controller, not by any particular device’s capabilities. (This is why some people who need maximum performance opt to purchase plug-in SCSI controller cards for their Macs, even though they already have a SCSI port.)

For a chain of SCSI devices to work properly it needs to be terminated at each end. If it’s not, then all kinds of strange problems can result. And in fact, many SCSI problems (on both Macs and PCs) are due to improper termination. Since all SCSI chains begin with the controller, it has its own terminator, which is always on. Individual SCSI devices also come with terminators—on some devices it comes in the form of a short, adapter-like plug that attaches to one of a SCSI device’s SCSI ports. (Most devices include two SCSI connectors so that they can be included in the middle of a chain, although some devices have only one SCSI connector, which means they always have to be at the end of a SCSI chain.) On other devices, including all internal SCSI drives, there’s a small jumper switch on the back panel that you can turn on or off as needed. The latest development is automatic termination, where a device can tell whether or not it’s the last one in the chain and then turn the termination on or off as necessary. (Even these type of devices offer a manual switch in case you need to override their automatic settings, however.) All devices in the middle of a SCSI chain must have termination turned off and all devices at the end of a SCSI chain must have termination turned on. If you’ve got multiple SCSI chains, such as one internal and one external, you need to terminate the device at the end of each chain. If you don’t, the Mac may be unable to boot, or may crash randomly. Sometimes an improperly terminated chain will work for a while and then start to blow up, so if you’re Mac suddenly starts working strangely, check your SCSI chains. Even if all your termination is set properly you may still run into problems that are related to the SCSI devices connected to your Mac. The reason for this is that the physical connections between various SCSI devices can be very temperamental—cheap SCSI cables are a notorious source of strange problems, for example. If you continue to have strange problems that you believe are due to your SCSI chain, invest in some high-quality (i.e., expensive) cables and see if that solves them. You should also try switching your cables around—as strange as it may seem, sometimes reversing a cable so that the portion that was originally connected to the Mac is now connected to the attached device and vice versa can make a difference. If that still doesn’t work, I would consider investing in an active terminator, which is essentially a fancier terminator that is more effective at blocking signals that can bounce around the SCSI chain and cause problems (which is what terminators basically do). To help out with your Mac SCSI troubleshooting, I also suggest you download SCSI Probe, which is popular utility for finding SCSI problems on the Mac. One final trick to consider if it's your internal SCSI hard drive causing the problem is to boot from a MacOS CD. Most Macs can boot from a CD-ROM with a System Folder by holding down the C key when you start up the machine. Once you've done so you can use Disk First Aid (which should be somewhere on the MacOS CD) to check for disk errors and/or update your hard disk drivers.

Zap, Zap, Zap If your Mac serial ports seem to be causing problems—for example, if you can’t communicate with an external modem, printer or other device attached to your Printer or Modem ports, you’re a prime candidate for resetting your Mac’s Parameter RAM, commonly called PRAM. A Mac’s PRAM is roughly equivalent to a PC’s CMOS—it’s a small amount of battery-backed memory that stores various settings for your computer, such as the time, etc. On the Mac, it also has settings that can effect communications with devices attached to your serial ports. The process of resetting the PRAM is called "zapping" the PRAM. To zap the PRAM on your Mac you need to restart your Mac and then quickly hold down four different keys simultaneously: Command-Option-P-R. If you’ve done it successfully you’ll hear a quick chime (and if you keep holding them down a bit you’ll hear several chimes), and then your Mac will reboot. Most Mac people recommend that you let it ring a few times to completely flush out the settings stored in the PRAM. Often times zapping the PRAM can clear up a variety of different hardware-based Mac problems. If that still doesn't work, you may have a software-related problem (many times they're hard to distinguish). For more info on Mac software problems, check out the "Mac Software Troubleshooting Tips" article. Finale Most people buy Macs for their ease of use and so get frustrated if they run into any kinds of problems. Thankfully, paying the "Mac premium" generally does give you a computer that causes a lot less problems than Windows equivalents, but not always. If you follow some of these steps when you do come across a Mac hardware problem, you should soon be able to productively smile back at your computer’s happy Mac face.