Opera is my favorite browser for several reasons. First, it’s small and fast but full of clever features. The download for the latest Windows version 10 is 6.6 megabytes, which includes not merely the browser but mail, news, and Internet Relay Chat (IRC) clients. It also includes many essential enhancements for which other browsers need add-on programs. Among many useful features are a built-in search bar that goes directly to Google (or to seven other search sites), a password manager, a cookie manager, a handy print preview mode, and “fast forward” and “rewind” buttons that select the most likely next or previous pages.

One very welcome feature is a set of options for dealing with those annoying pop-up ads. You can accept them, refuse them, or (most useful of all) open only those pop-ups you specifically request, globally or for specific sites. Yes, other browsers now include some kind of pop-up blocking, but Opera’s is simple, unintrusive, and effective. Microsoft’s pop-up blocking (belatedly added to Internet Explorer with Windows XP Service Pack 2) is nearly as annoying as the pop-ups themselves; comparing Opera’s implementation to Microsoft’s should be enough to convince anyone to switch. There is also the ability to block individual sites by right-clicking a link or image and selecting “Block Content.”

Second, Opera is an elegant alternative to the bloat and corporate arrogance of Microsoft’s Internet Explorer. Its thoughtfully-conceived user interface provides the maximum range of options to let you interact with the Web in the way that’s best for you. For mouse users, Opera makes full use of buttons and wheels, and offers a repertoire of special “gestures” for navigational functions. For those who don’t like (or can’t use) rodents, Opera provides a complete set of keyboard commands and shortcuts.

Opera has unique features designed to help users with imperfect eyesight enjoy the Web. Pressing the “+” key on the numeric keypad enlarges everything in the browser window by 10%; pressing it repeatedly will zoom up to 1000% (the “-” key zooms out). Opera also includes a set of “user mode” options that override a Web page’s fonts, colors, and styles, making text easier to read. You can select an “accessibility layout” that sets text to black sans-serif on a light green background, or choose a “high contrast” setting with black on white or white on black text. Other “user mode” options show HTML structures and elements to help Web designers debug their pages.

A large choice of “skins” and toolbar selections lets you extensively customize how Opera looks. Opera’s Web site includes pre-packaged “setups” to give Opera the look of Microsoft Internet Explorer, Firefox, or Safari, to ease the transition for users of those browsers. There is also an extensive and continually growing collection of user-developed skins. For Web site developers, Opera strictly adheres to World Wide Web Consortium’s standards for HTML and CSS. So developing and testing a site using Opera means the site should work well with any browser.

Third, because Opera doesn’t rely on Windows and Microsoft Internet Explorer components, it has none of the security vulnerabilities that regularly threaten users of Microsoft Internet Explorer. When they do discover a security problem, Opera Software acknowledges it and quickly issues a new version of the browser (they don’t use patches because the entire Opera installation download is smaller than most Microsoft patches). Opera 9 provides advanced security features to protect users against “phishing” scams, including a pop-up notification if a secure site’s certificate doesn’t match the site’s name.

Opera started out as time-limited shareware, and then went to an “adware” model that was fully functional but displayed an advertising banner until you bought a registration code to remove the banner. In September 2005, Opera’s executives finally conceded that it’s too difficult to convince people to pay for their browser when every other browser is free. With the release of Opera 8.5, they removed the banners and registration system and made it free for all to download and use.

The inevitable question is how Opera Software can make money now that their flagship product is free. As a small Norwegian company, they aren’t a monopoly like Microsoft or a media conglomerate like Time-Warner, with lots of lucrative products to offset what they give away. The answer is that they expect to attract enough users to make up in the long run the revenue they’re forgoing in the short term. They have a partnership agreement with Google that provides a small payment every time you use the convenient search bar (and most Opera users use it a lot). They also have licensed Opera’s “Presto” rendering engine to Adobe— if you use any of their Web tools you’re using Opera. They sell and license versions of Opera to mobile phone manufacturers for wireless Web browsing, as well as to various hardware developers for “vertical” applications (Opera for Toaster Ovens, anyone?).

Unfortunately, you’ll probably need to keep Microsoft Internet Explorer around. Many Web sites were developed using the popular Microsoft Front Page software. The resulting pages liberally employ proprietary extensions and characteristics specific to Microsoft Internet Explorer. This means some sites will have problems, or even be unusable, with anything other than Microsoft Internet Explorer. With one notable exception, most of those troublesome sites are the work of amateurs, who click all the fancy formatting and script-extension doo-dads on Front Page’s menus and tool bars without understanding what they’re doing. The “notable exception,” of course, is Microsoft. With typical monopolistic arrogance, some of their Web sites intentionally block or distort pages when a user has a non-Microsoft browser.

Another obstacle is that the Web sites of some banks and corporations check your browser when you enter. If you’re using something other than the current version of Microsoft Internet Explorer (or perhaps a recent version of Netscape or Mozilla) you might see a page informing you that you’re using an “unknown,” “incompatible,” or “obsolete” browser. They typically offer a convenient link to download Microsoft Internet Explorer.

Opera provides a handy way around that problem: Press F12 and you’ll get an option menu that lets you set Opera to identify itself as Microsoft Internet Explorer or as Mozilla (Netscape) for that site. This little white lie is usually enough to get past the gatekeeper. You have the choice of leaving “Opera” in the identification Opera sends to Web sites, or “masking” the identification to exactly duplicate what the other browser sends. “Masking” is useful for those sites that reject anything other than a “genuine” browser, but it’s better to let Webmasters know you’re using Opera unless they give you no other choice. Professional Web developers will only recognize Opera as a “known,” “compatible,” or “up-to-date” browser if it shows up in their logs.

Don’t let compatibility worries put you off. I’ve been using Opera since 1998, and it’s been my sole browser most of that time. I find that version 9 (identifying itself as Opera) works perfectly well about 98% of the time. Changing the identification with F12 usually gets around the checkpoints (and works fine) when it doesn’t. Microsoft-specific formatting or proprietary script extensions only very rarely force me to resort to Microsoft Internet Explorer. You’re probably better off avoiding such sites anyway, especially if they use ActiveX, a proprietary feature of Windows and Internet Explorer that lets Web sites run programs (and possibly wreak havoc) on your computer. If you must resort to lying, it’s probably best to identify as Mozilla rather than Microsoft Internet Explorer. That may avoid Microsoft-specific features on the Web site that could cause problems.

Once you try Opera, you’ll surely find that its cleverness, speed, and usability more than make up for the occasional compatibility problems (which again, you’d get with any browser other than Microsoft’s latest). And whatever inconvenience incompatibility might cause is certainly less than the inconvenience of continually downloading patch after bloated patch from Microsoft to plug dangerous security holes in Internet Explorer. I suspect that the growing exodus from Internet Explorer and its security vulnerabilities will ultimately encourage Web designers to move away from Microsoft-specific features, to everyone’s benefit.

The current free version of Opera is available for Windows, Macintosh OS X, Linux, Solaris, and FreeBSD. Older versions are available for OS/2 and QNX.

PocoMail

I use PocoMail for e-mail and Usenet news. It’s small and very clever shareware ($40; $30 for an upgrade to the current version 4.8), written by Slaven Radic of Victoria, British Columbia. Originally a project he wrote himself (with a support staff of four), Radic (as Poco Systems Inc) now aspires to challenge Microsoft with Barca, which adds a calendar, planner, notes, and diary to PocoMail’s e-mail capabilities for corporate users.

You might wonder why it’s necessary to buy an e-mail program when Microsoft so generously includes Outlook Express with Windows (and bundles the full Outlook with Microsoft Office). There are two reasons. The first is that many Microsoft products are what I call “arrogant software.” The underlying philosophy seems to be that Microsoft has determined the best— and only— way to do things, and users are thus obligated to adapt themselves to the software. Outlook isn’t quite as arrogant as Word (with which I have lost many a bloody battle), but I still find it unpleasant to use (in part because I find some of its “friendliness” quite annoying).

The more important and practical reason is that the prevalence of Microsoft Outlook means virus writers and hackers too frequently exploit it and its security weaknesses to commit their crimes. It’s analogous to modern agriculture, where the “monoculture” of single standard varieties of crops makes the entire crop vulnerable to disease. Many if not most e-mail viruses and worms will stop in their tracks when the recipient uses something other than Microsoft Outlook.

PocoMail, when used with a little common sense, is essentially immune to worms, viruses, and common techniques spammers use to invade your privacy. Unlike Outlook and some third-party mail programs, it includes its own HTML display routines rather than the Internet Explorer “engine” built into Windows. That completely avoids the Microsoft’s security vulnerabilities. You can also tell it to ignore HTML entirely and just show you the text. PocoMail also protects your privacy by not loading images from Web links embedded in mail (unless you specifically enable image loading). This circumvents the “beacons” spammers often insert to confirm receipt, verify “live” addresses for sale to other spammers, and possibly track your Web activity. And it will not open attachments unless you specifically ask it to; even then it warns about the potential risk and asks for confirmation.

Early in 2003, a computer infected with the “Yaha.E” virus sent me over 1000 infected e-mails. They piled up ominously in my mailbox and made me rather anxious, but never affected my machine. PocoMail let me safely examine the messages without opening or running the noxious attachments. My virus scanner confirmed what they were and deleted all the attachments, so no harm was done. Then I set PocoMail to delete anything from the sender of the virus from the mail server without downloading it. Had I been using Microsoft Outlook, my computer would have been infected— and sent the worm to everyone I knew— before I realized what was happening.

Security is not the only reason for preferring PocoMail. Originally the product of an individual, it’s small, lean, and efficient, with a freshness of approach you’ll seldom see from a large team of corporate coders. But it offers a lot of control over mailboxes, message styles, signatures, fonts, colors, spam filtering, and just about every other aspect of its operation. It’s the opposite of arrogant. The message editor provides intelligent quoting and a built-in spelling checker. PocoMail includes a powerful scripting language to control distribution and delivery of messages, but I have not found a need to try it. You can also customize PocoMail’s appearance with a multitude of “skins.” The latest version 4 has a snazzier but slightly simpler user interface, better options for searching through large collections of mail, and faster performance.

PocoMail includes advanced junk mail filtering that combines key word scoring with Bayesian filtering. You can edit a list of key words in headers and body text, assigning them a score. If the scores from key words and various other characteristics is high enough, PocoMail sends the e-mail to a “Junk Mail” folder as it comes in. Augmenting that is a Bayesian filter that you “train” to recognize the contrasting statistical characteristics of valid e-mail and spam by clicking “Junk Mail” or “Good” buttons. I find that the combined filtering works quite well, especially since so much spam includes misspellings, random spaces and punctuation, and strings of random words specifically intended to confuse Bayesian filtering. The filtering doesn’t work with e-mails that contain only a link to a graphical message, but PocoMail won’t display the image unless you request it.

If you’re among the many people who use Microsoft Outlook because it was there when you got your computer, or because your Internet provider’s installation CD set it up for you, it’s well worth exploring alternative e-mail programs. Not only can you avoid many viruses and worms, but you might find a better way to do e-mail.

Take Command

Even though the only officially-sanctioned way to interact with modern computers is by pointing and clicking on hieroglyphic icons, there are many tasks more easily done at an “old fashioned” command prompt. JP Software’s Take Command for Windows (new users, $100, upgrades $50) is a powerful replacement for the CMD.EXE that runs when you open a “Command Prompt.” If you spend any time at all using the “Command Prompt,” Take Command is indispensable. It adds hundreds of powerful commands and enhancements that Microsoft never bothered to add to their command-line interpreters, and replaces a slew of DOS or console-mode utilities.

A complete list of added commands would be too lengthy to include here. But a (very) few of them are a text file viewer; a file/text finder; editable command and directory histories; file date/time “touch;” FTP capability to manipulate remote files as if they were local (I use it to manage this Web site); GUI dialog management; T and Y “pipe fittings;” and file descriptions (originally a way around the “8.3” file name limitation of DOS, but still very useful). You can also play audio files, or send e-mail or instant messages from the command line. Many standard commands are enhanced. For example, you can get a directory of files with today’s date (with file names colored according to their extension); specify a list of files to exclude from copying moving, or deletion; schedule copying, deleting, renaming, or moving “access denied” files for the next reboot; or set the computer’s clock from a time server. You can create aliases for commands or groups of commands, and map them to keys.

Take Command provides enhanced command line editing. You can recall the previous command with the up-arrow key, select command from a list, or use control-0 through control-9 to select an individual parameter of the previous command. Since Take Command is a fully graphical application, it provides a scroll-back buffer from which you can cut and paste text.

Take Command greatly enhances the Microsoft’s batch file capability into a comprehensive and powerful programming language. Control structures include block IF, DO loops, and subroutines. And an extensive set of system information variables and functions manipulate numbers, strings, file names, and even Registry keys. It integrates with REXX, Perl, and Ruby if that’s what you prefer. An interactive debugger is built in. And if that’s not enough, Take Command provides a plug-in interface to add your own commands, or download plug-in libraries. Despite all the enhancements, the developers have gone to great lengths to ensure compatibility with all the Microsoft quirks (documented and otherwise) often exploited in batch files.

As a small but useful example of Take command’s extensions, I have a simple alias to show the total size of all files in the current directory and all its sub-directories. When I’m working on a travel photo essay, I create a directory for all the pictures, with separate sub-directories for the 16-bit master files, full-sized JPEG versions, small JPEGs for the Web, camera raw files with their Adobe Camera Raw “sidecars,” and miscellaneous working files. When I’m finished with all the pictures, I archive everything to a set of DVDs. I need to know how large all that data is to determine how it fits on the disks. The alias is DIRSIZE and the code is ECHO %@comma[%@filesize[/s *.*]]. It uses two “variable functions:” @filesize does the actual scanning and totaling, and @comma formats the resulting number with commas. The old-fashioned ECHO command writes the output. Adding some parameters to @filesize (@filesize[/s *.*,bc]]) would format the number without the need for @comma, and I could get the result in kilobytes or megabytes rather than bytes by changing @filesize[/s *.*,bc]] to @filesize[/s *.*,kc]] or @filesize[/s *.*,mc]].

JP Software previously sold a separate console-mode command interpreter called 4NT. It had the same functionality as Take Command, but could run some text applications faster because it was purely text based without the GUI windowing overhead. I used it to run the combination of batch files and Pascal filters that creates the index of pictures for this Web site. They also offered TCI, a graphical container that ran console applications in tabbed windows. Beginning with version 9, Take Command combines these functions into a single program that also includes a basic graphical file explorer with directory tree and file view panes. One or more instances of the command interpreter (now renamed TCC) run in tabbed windows within Take Command. Any other console-mode applications can also run in tabbed windows. You can still launch TCC separately as a console application, but there seems little reason to do that. JP Software claims that a console application running in a tabbed window displays text faster than if it ran directly on the desktop. If you aren’t using the file explorer, you can quickly toggle one or both panes out of the way until you need it. You can also set up a toolbar with buttons to launch your favorite applications.

The integrated “graphical console” approach lets you easily switch between a graphical and a command line interface as you deem appropriate, all within in one application. The integration brings other advantages. You can drag and drop a file from the explorer (or from any other application or the desktop) to the command line window. The name of the file then gets added to the command line, enclosed in quotes if the name has embedded spaces. That’s certainly easier than typing long file names! The file explorer isn’t as powerful as some file managers— I use and recommend VCOM’s $40 PowerDesk® Pro.

Take Command version 11 is the latest in a line of advanced command interpreters that began in 1989 with 4DOS, a replacement for the primitive COMMAND.COM that was the text-based face of MS-DOS in the days before Windows made computers user-friendly. I started using 4DOS in 1990, on a 12 MHz 286 machine running MS-DOS 3.3. I’ve relied on it, and its various descendants, ever since. Over the years I’ve seen the command interpreter evolve an impressive array of features. But I suspect it may now have reached its full maturity in terms of truly useful features. So my guess is that future versions of Take Command will focus on enhancing the integrated file explorer to make it competitive with products like PowerDesk.

If you don’t need the full functionality of Take Command, Take Command LE is a $40 “entry level” version. LE includes most of Take Command’s command-line enhancements and its complete windowed GUI. It omits all the Internet access and batch-file debugging capabilities, interfaces to REXX, Perl, and Ruby, and a few advanced functions and variables in the batch-file programming language. There is also a freeware console command interpreter, Take Command Console LE. This is the same command interpreter as Take Command LE, but without the windowed GUI (it’s like the old 4DOS or 4NT). Without the mouse and easy cut-and-paste capabilities, it’s less convenient to use than the windowed GUI; but for the price there’s no point in complaining.

UltraEdit

IDM’s UltraEdit is an excellent shareware text editor for programmers and Web developers ($50 for new users; $25 annually for upgrades). Unlike many Windows products, it has a lot of features without excessive bloat. It can work with multiple files, including searching and replacing text on a selected set of disk files. It includes check-as-you-go spell checking, word-wrap, and text formatting. For Web development it includes a browser view (using the Internet Explorer libraries integrated into Windows) and an integrated version of HTML Tidy, an HTML validator. UltraEdit can handle various flavors and formats of ASCII (for Unix and DOS) and includes a full macro and scripting capability.

While there are several other editors that offer these features, UltraEdit includes a full hex-file editor, integrated ftp (useful for publishing Web files and debugging CGI scripts), and configurable syntax highlighting for up to ten different programming languages (including HTML, C/C++, and Perl). The latest version (14) adds “environments” configured for different types of users and tasks, including Web development, technical writing, programming, and system administration. You can configure any of the “environments” to your own preferences.

A shareware competitor to UltraEdit well worth considering is NoteTab. NoteTab comes in three versions, “Light,” “Standard,” and “Pro.” The “Light” version is freeware, but it is anything but light on features. It’s a great replacement for Microsoft’s crippled Notepad editor, and handles large and multiple files in any available font. The “Standard” version costs $20 and includes a spell checker and thesaurus, better text formatting, and the ability to search and replace text in files on disk. The “Pro” version costs $30, and adds HTML syntax highlighting, faster text replace, bookmarks, outlining, and numerous other features.

It can be difficult to decide between NoteTab Pro and UltraEdit. NoteTab is cheaper and, perhaps, slightly friendlier to use. But I think UltraEdit has enough additional useful features to give it an edge. Try both and decide for yourself. Even if you don’t need a full-featured editor like UltraEdit or NoteTab Pro, it’s well worth downloading the free NoteTab Light.

Terabyte Image for Windows

A “drive image” backup is an exact copy of an entire hard disk or partition on CD, DVD, tape, or some other external media (preferably with compression to save space and writing time). It can be a real life-saver— or at least a system-saver— not if but when a hard drive suffers a catastrophic failure. That has happened to me twice (so far). Just plug in a new hard disk, restore the image from an external hard drive or set of CDs or DVDs, boot the restored disk, and you’re up and running. It’s an essential complement to a file-level backup that allows easy access to individual files.

Searching for a Backup Solution

I first discovered Terabyte’s Image products in 2003, when I bought a Memorex “True 8x” DVD writer. The drive came bundled with Nero Express OEM, which included an adequate file-level backup feature. But it lacked a drive image capability. Version 2 of NTI’s Backup NOW!, which I had been using for years to make file-level and drive-image backups on a CD writer, didn’t support any DVD writers. So my first thought was to look at upgrading to the then-current version 3. Its Web site claimed “extensive backup device support” including the “widest range” of DVD writers. In the fine print was a note to check their Web site for a list of supported devices. My DVD burner wasn’t on the list, and their prompt reply to my e-mail query was an apology that they had no plans to support it any time soon. So off to Google I went.

Although the Web site for Symantec’s Drive Image claimed it would work with “virtually any internal or external drive,” my drive wasn’t one of them. And even if it did support my writer, it required Microsoft .NET Framework. Putting another reeking pile of Microsoft bloat and security holes on my machine wasn’t particularly appealing. Acronis True Image, PC Magazine’s “Editors’ Choice” for drive backup, could only write DVDs using external UDF packet-writing software— and the one that came with my drive can’t format DVD+R disks. Paragon’s Drive Backup couldn’t burn an image directly to DVD at all under Windows 98. (I have since bought a new computer with Windows XP, and use a USB external hard drive rather than DVDs for backup.)

Having exhausted the realm of commercial shrink-wrap software, I turned to shareware. I found exactly what I was looking for in Terabyte Unlimited Image for Windows. This is a straightforward, no-nonsense tool that creates a bootable CD or DVD with a compressed image of a hard disk partition. At $39, it’s about half the price of its cheapest commercial competitor. Its user interface isn’t fancy or pretty, but all the features are easily accessible. And mirabile dictu, it could write directly to my DVD writer! How could a little shareware company manage what NTI or Symantec (heir to the famous Norton Utilities) can’t? I don’t know. But it works. And it most likely works with any DVD writer (a real advantage of shareware is that you can find out if it really works on your system before you buy it).

The Image for Windows purchase price also includes licenses for Image for DOS and Image for Linux. The Windows package installs PHYLock™, a driver that can lock system files and files in use by applications, allowing the backup of a running boot (C:) drive from Windows. That would seem to make the other versions unnecessary for Windows users. But a stand-alone bootable floppy disk (with Image for DOS) or CD (with Image for Linux) is an absolute necessity. It provides the only way to restore a failed hard drive, or a corrupt Windows installation that won’t start. If you’re still running an ancient DOS-based version of Windows (95/98/Me), you’ll need the DOS or Linux version to back up your C drive because PHYLock won’t work.

All three Image versions have the same capabilities, options, and user interface, although the DOS and Linux versions use 1980s-style text rather than Windows graphics. Since they work directly with hardware, they can access just about any IDE or SATA internal hard drive or CD/DVD burner, or a USB or FireWire external drive. They can also read and write NTFS partitions. The Windows package includes a utility that creates a bootable floppy disk containing Image for DOS and TBOS, Terabyte’s DOS clone. The Image for Linux package includes a Windows installer that creates a “turnkey” bootable CD, DVD, or USB thumb drive with a complete Linux installation. You don’t need to be a Unix wizard, since it starts up with a user-friendly menu rather than an intimidating console shell (though a shell is available as one of the menu options). It also identifies hard drive partitions by volume name rather than as opaque Unix “devices.” You’ll need to use Image for Linux rather than the DOS version if your backup media is an external USB drive and you have a USB keyboard. DOS can’t handle both devices simultaneously.

Image offers optional encryption of backup files. It can also optionally omit the Windows paging and hibernation files to save time and space on an image of a system partition (those large files contain no useful data, and Windows will just re-create them when you start it from a restored partition). You can also make a differential backup of only the files that have changed since the last full backup. That should make frequent backups quicker and much more convenient. With either type of backup, Image can optionally verify the integrity of a compressed backup file, or do a byte-by-byte comparison of the backup with the source. Either option increases the time it takes to complete a backup; the byte-by-byte comparison exactly doubles the time. But it can provide an extra measure of assurance for critical backups.

Image for Windows includes TBIView™, a Windows Explorer extension that lets you navigate and extract individual files from a backup image file. Double-click on an image file (with the .TBI extension) and it opens a navigation window, with the image’s directory tree in the left pane and the contents of a selected directory in the right pane. From there you can drag files or directories to your desktop or to another application. TBIView has saved me from stupid mistakes more times than I’m willing to admit. It effectively makes Image for Windows a reasonably convenient file-level backup program as well as a drive image backup tool.

Image in Action

I had been making regular backups with Image for over five years before I had any reason to use its restore capability. My old 160GB Seagate 7200.7 drive didn’t fail, but it was getting uncomfortably close to the end of its warranty. When I found an irresistible price on a 640GB Western Digital Caviar Black drive, I decided that Image would be the ideal tool for copying the old drive to the new one. It can resize a partition after restoring it to a larger drive, and I’d also have a complete backup when the upgrade was finished.

My hard drive has a complex setup, with a bootable partition and two data partitions. Terabyte’s documentation is extensive, but it isn’t entirely clear about how to back up that type of setup and restore it to a new larger drive. I posted a question on Terabyte’s newsgroup (available through an interface on their Web site) and promptly got exact instructions from one of their people. I’ll describe the process, as an example how Image works and some of the options it provides.

Using Image for Linux, I first backed up the entire old drive onto an external hard drive. Image provides the option of backing up either individual partitions or the entire drive at once; the latter option backs up all partitions. Then I powered down the computer, swapped the drives, and rebooted from the Image for Linux CD. As with the backup process, Image can restore the entire drive at once, with the option of resizing the restored partitions to fit a larger drive in proportion to their original sizes; or it can restore (and resize) individual partitions separately.

I restored the C partition first, selecting the options to restore the first track, adjust the Windows BOOT.INI, and make the partition active. Then I entered the new size of the restored partition (39 GB on the old drive, 48 GB on the new drive) in the “resize after restore” text box. Restoring the first track creates a master boot record (MBR) that includes all the old partition letter assignments. It also includes the hidden data some applications (such as Adobe Photoshop) write to the MBR as part of their anti-piracy schemes. Some disk imaging software creates a new MBR when restoring rather than saving and restoring the first track (Image can also do this if you select that option); so any applications that hide data in the MBR will have unexpected problems. The “make active partition” option makes the partition bootable. If you’re like most Windows users and have only one partition on your system drive, the restore process is done after Image restores this partition.

I un-selected the options I listed above when restoring the other two partitions. For each one, I entered the new size for the restored partition in “resize after restore.” That text box displays the minimum partition size for the restore and the maximum available free space on the drive. You can specify any new size within that range. Because I restored the MBR with the C partition, Image correctly created the necessary extended and logical partitions on the disk. It then adjusted their sizes after restoring the data. Note that Image reports free space and specifies partition sizes in Mebibytes, units of 1,048,576 bytes. Manufacturers of disk drives specify capacities in megabytes or gigabytes, units of 1,000,000 or 1,000,000,000 bytes respectively. So you’ll have to do some arithmetic when determining the partition size to select in Image.

When Image for Linux finished restoring the third partition, I powered down the computer, disconnected the external drive, and rebooted from the new hard drive. Windows started up flawlessly, although it informed me that it had “found new hardware” (i.e., the new hard disk), which it “installed” and asked me to reboot. I did that, and it again started up flawlessly. Because an image backup restores the sectors of a partition exactly as they appeared on the original drive, the system partition’s volume serial number is the same as it was on the old drive. That helps to avoid triggering Windows’ anti-piracy scheme, which requires “re-activation” if too many hardware items change. I found everything in order, including Photoshop and several other applications that require “activation.”

Commercial products may now have caught up with the DVD writing capabilities of Image. But I can definitely recommend Image for Windows as an excellent and inexpensive drive image backup tool that’s doubles as an excellent and inexpensive file backup and recovery program. And it’s impressive proof that software can be powerful and useful without unnecessary bloat or fancy bells and whistles. Image for DOS, which fits handily on a floppy disk, proves that the art of writing elegant and compact code is not lost. Microsoft and Adobe could learn much from Terabyte!

Note: I mention copy protection mainly to illustrate the frequent contention that, like so many “security” measures, it tends to burden legitimate users far more than it does “pirates.” But I also need to point out the potential legal pitfalls of tools like Terabyte Image. In the United States, Section 117 of the Copyright Act specifically allows backup copies of software that you legitimately own. But any contrary provisions a software publisher buries in its license “agreement” will take precedence. And Section 117 does not apply to other types of copyrighted files, such as music or movies. So making image backups may violate copyright law as well as the license “agreements” for one or more of the software programs on your hard disk.

That said, I believe that the very real benefit of making regular backups specifically to protect yourself from an inevitable disk failure outweighs the theoretical risk of becoming the sacrificial defendant in a test case (as far as I know, no appellate court has yet ruled on the issue of making disk backups). Don’t keep backups on networked drives that the RIAA’s or MPAA’s zealous snoops— or actual pirates— might find. Do keep backups on one or more detachable external hard drives in safe locations. This is, of course, completely different from using imaging software to “clone” hard drives for multiple computers without specific authorization or license, or to create images for unauthorized distribution. Anyone who does either deserves whatever misery copyright owners and their lawyers choose to inflict on them.

About Upgrading Serial ATA Hard Drives

I have an Asus A8V Deluxe motherboard that uses the VIA K8T800 Pro chipset. This chipset— specifically its VT8237 “southbridge”— is one of several older controllers that have a known and documented incompatibility with current SATA 2 hard drives. The motherboard also has a Promise PDC20378 controller (also called “FastTrak 378”), but when I was researching my hard drive upgrade I could find no specific information about its compatibility with SATA2. I’m including this note here (for the lack of any better place to put it) to answer that question.

The Promise controller does support SATA 2 correctly. My Caviar Black SATA 2 drive and an older Seagate 7200.9 both work perfectly with the Promise controller in “IDE” mode with no jumpers needed, though almost certainly at 150 megabits per second. I suspect the Promise controller also works with SATA2 drives in RAID mode, but I haven’t tried that. Since my old drive had been running from the Promise controller all along, it had the correct Windows driver installed. But if you are upgrading an older SATA drive that was connected to the VIA controller, you’ll need to first move it to the Promise controller, install the driver, and verify that it works before you make the backup. (By the way, the Promise controller’s SATA capability is only for hard drives. As the Asus motherboard manual clearly notes, it doesn’t support SATA CD or DVD drives.)

If you’re wondering what this all means, current Serial ATA hard drives use an interface that is usually (but incorrectly) called “SATA 2” or “SATA II.” It’s an upgrade to the original Serial ATA standard that offers a maximum data throughput of 300 megabits per second, twice the rate of the original standard. In practice this enhancement is yet another one of those features more useful for marketeers than for users. The old 150-megabit standard is faster than any mechanical disk drive can output data. The higher throughput rate could theoretically be beneficial when transferring a small “burst” of data from a hard drive’s cache.

The developers of the SATA 2 standard included a provision for compatibility with controllers that don’t support the faster rate. When the controller first detects an SATA 2 drive during initial start-up, it “negotiates” a data transfer rate with the drive. If the controller can’t support the faster rate, the drive will configure itself to the older standard. But some first-generation SATA controllers, such as the VT8237, were designed before the SATA2 standard and don’t support the negotiation protocol. They won’t even recognize an SATA 2 drive. To work around that problem, SATA 2 drives have included a jumper option that manually sets the drive to the old standard. But some drives— notably Seagate 7200.11 models larger than 320GB— don’t provide that jumper (all sizes of the newer 7200.12 drives have the jumper). And some others, like my Caviar Black, don’t fully document it. Western Digital’s knowledge base article describes the “OPT1” jumper, but the drive label and the “quick start guide” both state only that the jumper block is “for factory use.”