1. IE 7 includes tabbed browsing, a capability that Firefox and Opera have offered for a while.

2. IE 7 has an integrated search box similar to that in Firefox and Opera.

3. Both IE 7 and Firefox 1.5 offer an easy method for deleting personal browsing data (browser cache, URL history, saved forms) via one menu option. The feature already exists in Opera 8.5.

4. Firefox 1.5 offer automatic updates similar to IE.

5. Opera comes with several advanced features that you can get in Firefox only with add-ons, and that IE lacks entirely.

Erik Larkin compares the three most popular browser on earth – Internet Explorer 7 Beta 1, Firefox 1.5 Release Candidate 1, and Opera 9 Preview 1. He also mentions Flock, a social browsing application. Flock rethinks the browser as a tool for creating Weblogs, organizing bookmarks and connecting with others online.

IE 7: The new Phishing Filter in IE aims to warn users if they visit a known or potential phishing site–a function previously available only via third-party toolbars.

Firefox 1.5: Firefox 1.5 includes drag-and-drop tab reordering. The new Auto-update process is streamlined and smooth.

Opera 9: Users can choose which Web sites can run JavaScript or display images. And the impressive built-in RSS feed handler now supports Atom 1.0. However, Opera has no plans to introduce Firefox-like extensions, so if you don’t like the way it does something, you’re stuck.

Browser Market Share: Though IE has been losing market share, it remains dominant yet. The number of users jumping to Firefox has slowed recently leading some experts to suggest only a finite number of people are willing or able to try an alternate browser.

Read full story at PCWorld – Browser Face-Off

Konqueror

1. Click on the Settings-menu in Konqueror
2. Click on the Configure Konqueror… item in the menu – a new window opens
3. Click on Cookies on the left side (you might have to scroll first)
4. Set Accept all cookies by default.
5. Save changes by clicking Ok.

Lynx

1. Press O for Options
2. Look for General Preferences and Cookies (!)
3. Set it to either ask user or accept all (more convenient)
4. Save changes by selecting Accept Change

Mac IE 5.x

1. Click Edit
2. Select Preferences
3. Under the Receiving Files option, select Cookies
4. Under “When receiving cookies:” select the desired level of cookie acceptance
5. Under “When receiving cookies:” select the desired level of cookie acceptance
6. Click OK to finish

Mac IE 4.x

1. Go to My AOL on the menu bar
2. Pick WWW
3. Go to the Advanced Settings option on the Category menu
4. Click “Cookies”
5. When receiving cookies: Click “Never Ask”
6. Click OK

Internet Explorer 7

1. Open your browser, and select Internet Options from the Tools menu at the top.
2. Now on the option screen select the Privacy tab.
3. Make sure the slider is not set to “Block All Cookies”.

Microsoft Internet Explorer 6.0

1. Select “Internet Options” from the Tools menu.
2. Click on the “Privacy” tab.
3. Click the “Default” button (or manually slide the bar down to “Medium”) under “Settings”.
4. Click “OK”.

Microsoft Internet Explorer 5.x

1. Select “Internet Options” from the Tools menu.
2. Click on the “Security” tab.
3. Click the “Custom Level” button.
4. Scroll down to the “Cookies” section.
5. To enable:
   1. Set “Allow cookies that are stored on your computer” to “Enable”.
   2. Set “Allow per-session cookies” to “Enable”.
6. Click “OK”.

Microsoft Internet Explorer 4.x

1. Select “Internet Options” from the View menu.
2. Click on the “Advanced” tab.
3. Scroll down to find “Cookies” within the “Security” section.
4. To enable:
   1. Select “Always accept cookies”.
5. Click “OK”.

Mozilla Firefox (1.0 final release and earlier)

1. Go to the “Tools” menu.
2. Select “Options”.
3. Select the “Privacy” icon in the left panel.
4. Check the box corresponding to “Allow sites to set cookies”.
5. Click “OK” to save changes.

Netscape 7.1/Mozilla 5.0

1. Select “Preferences” from the Edit menu.
2. Click on the arrow next to “Privacy & Security” in the scrolling window to expand.
3. Under “Privacy & Security”, select “Cookies.”
4. Select “Enable all cookies”.
5. Click “OK”.

Netscape Communicator 4.x

1. Select “Preferences” from the Edit menu.
2. Find the “Cookies” section in the “Advanced” category.
3. To enable:
   1. Select “Accept all cookies” (or “Enable all cookies”).
4. Click “OK”.

Opera 7.x

1. Click on the Tools-menu in Opera
2. Click on the Preferences… item in the menu – a new window opens
3. Click on the Privacy selection near the bottom left of the window.
4. The Enable cookies checkbox must be checked, and Accept all cookies should be selected in the “Normal cookies” drop-down.
5. Save changes by clicking Ok.

Opera 9.x

1. Click on the Tools-menu in Opera
2. Click on the Preferences… item in the menu – a new window opens
3. Click on the Advanced selection on the top of the window. Then select Cookies from the left hand side
4. The Enable cookies checkbox must be checked, and Accept cookies should be selected
5. Save changes by clicking Ok.

Safari

1. Click on the Safari-menu in Safari
2. Click on the Preferences… item in the menu – a new window opens
3. Click on the Security icon near the top of the window.
4. Under Accept Cookies the Always or Only from sites you navigate to must be selected.
5. Save changes by clicking Ok.

• McAfee VirusScan
• Kaspersky Internet Security
• eTrust EZ Antivirus
• Trend Micro PC-cillin Internet Security 2007
• ZoneAlarm Pro

McAfee VirusScan Review

McAfee is an industry leader in computer protection and VirusScan is their #1 defense against viruses. This software comes with a ScriptStopper, to prevent virus to propagate from one computer to another via email, and WormStopper. Get McAfee VirusScan. Click here.

Ease of Use: 4/4

If you use the default settings, setup is a breeze. The default settings provide very good protection against viruses, Trojans, worms, malicious ActiveX controls and Java applets.

The SecurityCenter has several tabs besides the VirusScan tab. There is one for Personal Firewall+, Privacy Service and SpamKiller. These are other programs offered by McAfee.

Effectiveness: 4/4
VirusScan boasts one of the best virus detection records. The software received the VB100% award for 2005, the ICSA award for 2005 and passed both level 1& 2 Checkmarks from West Coast Labs.

Updates: 4/3

McAfee VirusScan 9.0 automatically downloads updated virus definitions from the Internet on a regular basis. Staying up to date will dramatically decrease any possibility that a virus will infect your computer. On Average, McAfee comes out with new virus definitions on a daily basis.

You will have to register and login to get updates-this makes the process longer than it needs be.

Feature Set: 4/3.5

McAfee VirusScan offers a rich list of features. You can set VirusScan to scan your computer at a preprogrammed scheduled time. The program also scans for spyware and adware.

You can’t add several tasks to the Scheduled Scan. This would be a nice quality so you could scan all files once a month and only program files weekly or any other tasks you would like to set up.

McAfee has an Antivirus Emergency Response Team that continually monitors the worldwide virus activities to provide you with the utmost safety. The always-on protection guards against viruses, spyware and other Internet threats that may enter your PC via e-mail, instant message attachments, Internet downloads, and web browsing.

Ease of Installation: 4/4

We didn’t encounter any problems during the installation process.

Help/Support: 4/2.5

McAfee has plenty of help options, including documentation, email and phone support. We were disappointed with their email response time-it took 5 days to respond and they didn’t answer all our questions, hence the low rating.

Summary:

McAfee is a name brand product that does an excellent job at detecting and removing viruses. Their update process can be somewhat tedious and their support department needs to answer their emails in a timelier manner.

Kaspersky Internet Security Review

Kaspersky also receives a perfect score on our review, passes all of the major virus tests, has hourly definition updates and offers an extensive feature set. This is one of the best anti-virus software. Download award-winning antivirus & anti-spyware software from Kaspersky Lab now

Ease of Use: 4/4
Kaspersky has a very clean and easy-to-use interface. The program is divided with three tabs- protection, settings and support. Within each tab, you can choose from several other functions.

This software is perfect for the average home user and advanced users. For home users, who don’t want to get too technical, there is recommended settings and an easy-to-find scan button that will instantly start to scan your computer based on these settings. More advanced users can go into the settings tab and completely customize what they would like the program to do.

Kaspersky comes with three types of scan levels: Maximum – thoroughly scans all files, self-extracting archives, startup objects, disk boot sectors, system memory and email format files. Normal – covers everything the maximum setting covers minus email format files. High Speed – scans everything minus the email format files, archives and startup objects.

Effectiveness: 4/4

Like the #1 product, Kaspersky passed the VB100% award, the ICSA 2005 and the West Coast Labs Checkmark level 1 and 2. These awards mean the software detected all of the “in the wild” viruses, generated no false positives, prevented virus replication and disinfected all of the viruses that could be disinfected.

Besides being very thorough and reliable, it only took eight minutes for us to scan an 80 GB hard drive. This anti-virus software is so fast you won’t have to set aside hours to scan your computer or set it up to run at night.

Updates: 4/4

By default, Kaspersky updates their definitions on an hourly basis, which blows away the competition. This leaves a very small window for a virus to be created and infect your machine. Alternatively, you can configure the program so you can receive automatic updates or you can manually update your program.

Furthermore, the updates will resume if your Internet connection is interrupted, and the cumulative updates, the ones that contain the entire virus database, are only half the size they used to be. The latest version employs a special technology so the program will use already existing anti-virus database for a cumulative update.

Feature Set: 4/4

You can choose which files you would like Kaspersky to scan by simply right clicking on the file. You can even select a number of files to scan or just scan your entire computer.

If the software detects a virus there are a number of things you can have the software do. It can just delete the file, notify you that it has found an infected file or log the information for later use.

During each scan, there is an estimated time to finish so you can plan accordingly.

Ease of Installation: 4/4

We didn’t encounter any problems while installing this program.

Help/Support: 4/4

Kaspersky has all of the support you need. There is an email address, an online forum, phone support and a FAQ. They answered our email within 12 hours.

Summary:

Kaspersky is an easy-to-use program for the average person with plenty of customizable options for an advanced user. Just install and set up this software and you’ll never have to worry about a virus or worm infecting your computer.

eTrust EZ Antivirus Review

eTrust’s EZ Antivirus is an easy-to-use program that’s good for computer novices and does a good job protecting your files and your computer from viruses. Download Your FREE Optimize Scan from CA!

This software doesn’t scan for riskware, adware or spyware.

Ease of Use: 4/4

The interface uses an Explorer-like file tree that you use to select individual files and folders for scanning. This program is user-friendly with four tab categories. This product is ideal for the average user, but advanced users may find the software has limited advanced features.

Though you can exclude and customize specific files and folders to scan, the software lacks the advanced features that Bitdefender offers.

Effectiveness: 4/4

In Virus Bulletin’s tests with live viruses, previous versions of eTrust’s EZ Antivirus earned the VB100% rating in seven of its eight most recent Windows tests. The only test that didn’t receive 100% was while running on the Window ME operation system.

EZ passed ICSA Labs testing for August in both the OnDemand scanning tests and OnAccess tests.

The software’s compressed file detection is limited and it only scans email when you open an attachment.

Updates: 4/4

eTrust updates their virus definitions on a daily basis. You can choose between automatic and manual updates.

Feature Set: 3/4

EZ includes a new feature called Snooze that disables the program’s real-time protection for a user-specified period-from one to 99 minutes. This is great for new software installations or to troubleshoot a system performance glitch or problem. In previous versions of EZ Antivirus, you had to reboot after disabling virus monitoring. To deactivate Snooze, simply click the Wake Now icon in the System Tray. If you forget, the timer will automatically turn off Snooze after the allotted time and restart the EZ Antivirus protection.

EZ Antivirus real-time scanner operates like most other AV software and runs in the background without much of a performance hit. EZ is lacking a “quarantine” function for infected, unclean files. This feature is common on many other AV packages.

Ease of Installation: 4/4

We didn’t have any problems during installation.

Help/Support: 3/4

eTrust has great support with their support page, Knowledge Base Search and FAQs page. The support team responded to our email within 36 hours. However, the telephone support is very expensive at $49.95.

Trend Micro PC-cillin Internet Security 2007

The good: PC-cillin Internet Security 2007 offers solid protection against viruses, hacker intrusions, and other online threats. It also costs less than its competitors and provides free phone support. Download award-winning internet security and anti-spyware software from Trend Micro

The bad: PC-cillin Internet Security 2007 could provide more aggressive spyware protection and limit or disable its own security pop-up messages.

The bottom line: PC-cillin Internet Security 2007 has a few shortcomings, but overall it’s an affordable and feature-packed security suite that reliably defends against online threats.

Trend Micro PC-cillin Internet Security 2007 is a better buy than top-selling competitors Symantec Norton Internet Security 2007 and McAfee Internet Security Suite 2007. For starters, it costs $20 less than Norton or McAfee while bundling a similar toolkit, including virus, spyware, and phishing protection and a firewall. It produces less of a drag on system performance, and PC-cillin users get free phone support–a rarity in the industry. But PC-cillin 2007 also has its shortcomings. It should be more proactive in blocking spyware downloads; it leaves some security options turned off by default to reduce the hit on your system resources; and it peppers you with annoying pop-up security messages (you can disable this feature, but the deactivation process could be much easier). Despite these gripes, PC-cillin Internet Security 2007 is a worthy upgrade and a solid buy for new users.

Setting up PC-cillin Internet Security 2007 takes just a few minutes, although it takes a bit longer if you install the optional Trend Micro Anti-Spam toolbars for Microsoft Outlook and Outlook Express. We’d prefer a single setup routine for all components. The toolbars appear to have been an afterthought, and you have to rummage through folders on the setup CD to find the Anti-Spam installation files. Once setup is complete, PC-cillin downloads updated virus definitions and program files and recommends that you run a system scan. The initial system scan is optional, as it is with McAfee, Norton, and other security apps.

The PC-cillin 2007 interface is nicely organized. As in Norton and McAfee, the main Overview screen displays a snapshot of your security status. Green checkmark icons indicate that your computer is properly protected against the usual threats. Large buttons titled Update Now and Scan Now make it easy to download the latest program files or scan for security breaches. Dig deeper, and you’ll find a level of control in PC-cillin that its best-selling competitors don’t match. For instance, you can configure the real-time virus protection to scan compressed files (ZIP files, for example) up to six layers deep. The real-time scanner doesn’t scan ZIP files by default, but the manual and scheduled scans do.

Surprisingly, PC-cillin 2007’s default configuration leaves many key security tools turned off, including the Web Site Filter, which blocks specific sites or categories of sites–gambling, sex, job search, and so on–and Privacy Protection, which prevents personal information, such as your name and your telephone and credit card numbers, from being sent without your permission. It’s easy enough to activate these features via the AntiFraud Wizard, but they should be up and running right out of the box. According to Trend Micro officials, some security features are turned off to reduce PC-cillin’s drag on system performance. While it’s true that ultrahigh security makes for a slower PC, we found that PC-cillin 2007–even with every security tool activated–was less of a speed bump than McAfee or Norton.

PC-cillin Internet Suite 2007 packs an impressive arsenal of security weapons. Like its competitors, it protects against viruses and spyware, cracker intrusions, and phishing attempts, the last of which are designed to purloin your credit card and bank-account numbers. This year’s upgrade brings improved phishing protection, including a new antiphishing toolbar for Internet Explorer. Another plus: PC-cillin detects new users on wireless networks. New users are increasingly common in home and small-business wireless networks, and they often operate without any security whatsoever.

That said, some of PC-cillin’s security tools need sharpening. Its spyware detection, for instance, should be more aggressive. Like Norton Internet Security 2007, PC-cillin allowed us to download and install WeatherCast, a free utility that contains adware. (After WeatherCast’s installation, both Norton and PC-cillin identified the program as a potential threat and offered to delete it.) By contrast, McAfee Internet Security Suite 2007 took a more proactive approach by identifying WeatherCast as a threat before we downloaded it.

PC-cillin also displays too many pop-up messages. When the Privacy Protection feature is activated, a pop-up box notifies you each time PC-cillin prevents a site from accessing your personal information. This gets annoying in a hurry, and you can end the messages via the Popup Manager in the main console. A suggestion: We’d rather click a button in a pop-up box to silence the messages (McAfee offers this). Trend Micro officials say future versions of PC-cillin will likely include this feature.

Here’s a big plus: Trend Micro provides free phone support for PC-cillin users–a rarity among security-software vendors. Phone times are Monday through Friday from 5 a.m. to 5 p.m. PT, and (get this) the calls are toll-free. For weekend and after-hours help, PC-cillin’s online knowledge base is helpful and worth a look.

Unfortunately, we had mixed experiences with PC-cillin support. The phone experience was excellent. We experienced no wait times, and a knowledgeable and polite tech rep helped us fix a glitch involving PC-cillin and our 2Wire home-network diagnostic utility. However, one of the several e-mail queries we sent via Trend Micro’s support site was never answered. Still, Trend Micro’s level of support is the highest we’ve seen in the security space.

ZoneAlarm Pro Review

This powerful firewall/Antivirus software program not only protects your computer from hackers, it also protects your personal data by blocking those nasty viruses and Trojans. Whether you’re a novice or a computer professional, ZoneAlarm Pro makes it easy to keep your computer data safe. Download ZoneAlarm Pro Here

The new 6.5 version of ZoneAlarm Pro has strengthened their Anti–Spyware feature and added new features: Identity Theft Protection, Spy Site Blocking and Game Mode.

Feature set: 4/4

Your financial data and password lists are stored inside your computer, but a novice hacker could easily get into your system to copy these without detection. How do you keep prying eyes locked out and your private data locked in?

ZoneAlarm Pro has an option called MyVault; here you can enter any personal information you’d like protected. You can also enter a list of trusted sites that you will allow access to this private data-such as your bank website. If another source attempts accessing or transmitting MyVault data, ZoneAlarm Pro will block access.

We tested this feature by creating a test bank account number and password in MyVault. Then we tried to send that data in an email. ZoneAlarm Pro detected the attempt to send the sensitive data and blocked the transmission.

Two bonus features in ZoneAlarm Pro include the ability to block viruses and offer email protection. Most firewall software programs require you to uninstall existing anti-virus software before installing if their company also offers an anti-virus product. (This way, they force you to buy their anti-virus software as well.)

But ZoneAlarm Pro will incorporate your current anti-virus software into their firewall, allowing you to use the anti-virus product of your choice. And, if you don’t already have anti-virus protection, ZoneAlarm Pro will install their version, saving you time and money.

Not many firewall programs include email protection, but ZoneAlarm Pro uses Mailsafe software to protect you from receiving or sending viruses by email. Most people worry only about incoming viruses, but nothing is more distressing that learning a virus has sent itself to all the friends in your email address book-outgoing email protection is just as important.

You can customize settings within each field (except the anti-virus field) to tailor ZoneAlarm Pro. You can create a safe list of senders’ addresses, allowing ZoneAlarm Pro to act automatically as an email filter. This is an excellent bonus when you consider that Trojans and viruses often enter your computer by email.

Ease of use: 4/4

We at TopTenReviews rate how easy a program is to use from a computer novice perspective. We ask, “If I’m a beginner, would I be able to understand and use the options in this program quickly, without questions?”

ZoneAlarm Pro does exactly that. The well-designed control panel explains and displays available choices simply. Each field (such as program control) offers a brief description of the main choice and then shows other choices available. New users are not overloaded with irrelevant information.

Ease of installation & setup: 4/4

We experienced no difficulties while installing ZoneAlarm Pro.

Reliability: 3.5/4

We tested ZoneAlarm Pro’s ability to protect a computer system’s vulnerable open ports. An open port is a small doorway in your computer system that allows data to enter and leave. When you use a chat program online, for instance, your computer opens four ports to allow data to flow back and forth quickly. ZoneAlarm Pro was effective at protecting ports. ZoneAlarm Pro passed the Trojan test with flying colors too; no destructive Trojan programs made it past the firewall.

The only test ZoneAlarm Pro failed was the refer test. The referrer test checks to see if the firewall allows general computer information (such as your operating system, browser, location, IP address) to be seen by others online. Almost every firewall program we tested failed the refer test.

Help/Support: 4/4

ZoneAlarm Pro’s documentation is well organized, making it easy to find answers fast. Learning to use the different fields and options that ZoneAlarm Pro offers is simple too, since this software is designed for a novice to understand.

ZoneAlarm Pro also has instant chat support available online. This allows you to enter then fine-tune a question. Then the support staff guides you to the right answer. ZoneAlarm Pro also has premium phone support available for a fee.

Summary:

There are hundreds of software firewalls on the market today. Each claims to protect you from intruders. But few also protect your computer and your email from hackers, Trojans, viruses, porn ad images, and unwanted cookies. Like having a guard dog at the door, ZoneAlarm Pro delivers peace of mind by taking care of all these undesirables for you.

Maybe so. But if that’s true, it’s all the more reason to make some noise about it and get open source developers more interested in delivering those tools. Because the alternative — turning to proprietary software for the management piece — could have fairly unpalatable side effects.

By way of example, let’s take a look at another closed-source tool that can manage open source software. Vintela Authentication Services, part of Quest Software’s suite of identity management products, makes it easy for network administrators to centrally manage user access accounts on Linux systems. And, I must admit, I’ve seldom felt as conflicted about an IT product before.

The reasoning goes that most IT organizations have two types of administrative staffers. The Windows admins are accustomed to the day-to-day maintenance of Windows networks, including end-user support for desktop workstations. Linux admins, on the other hand, tend to be a more rarified breed. They have different core knowledge and experience — not to mention being more expensive.

When employees join or leave an organization, however, provisioning their physical workstations is usually only one of the IT tasks involved. They also inevitably need access to a whole variety of servers, including mail servers, in-house Java applications, even databases. Many of those servers are maintained by the Linux side of the IT organization, yet adding and deleting user accounts is seldom the favorite part of those admins’ jobs. More importantly, it’s also probably not the best use of their time (dollars, again).

This is where Quest comes in. Vintela Authentication Services, in conjunction with Quest’s other identity management products, makes it possible to manage those Linux user accounts through Active Directory and Microsoft Identity Integration Server. It allows Windows admins to automate and maintain central control of these functions, freeing the Linux admins to concentrate on more critical systems maintenance tasks. What’s more, centralizing in this way makes it easier for IT to deliver audits of account creation and deletion, for compliance purposes.

Personally, I think this is brilliant. It eliminates bottlenecks and reduces IT drudgery, and does so in a sane way using proven tools. I’m sure that many admins on mixed Windows/Linux networks will be thrilled to discover it. But there’s a more troubling side to it, too.

“We take the essence of Samba and apply that to identity management,” Jackson Shaw, senior director of product management for Quest’s Active Directory products, told me in a recent meeting. “We don’t use Samba, but that’s the idea.”

And there’s the rub. Quest doesn’t use Samba because it doesn’t have to. Quest has a close partnership with Microsoft, so it doesn’t have to reverse engineer Microsoft protocols the way Samba does. In fact, Microsoft encouraged Quest to acquire the former Vintela in 2005, after investing US$10 million of its own in the company.

Why would Microsoft be so interested in a company that provides management tools for Linux? Consider the ramifications.

By encouraging a policy in which all identity management and account provisioning happens via Active Directory, Quest’s products effectively make Linux servers subservient to the Windows infrastructure. They also effectively elevate the Windows admins to become the gatekeepers of those Linux resources. In other words, they keep Linux right were Microsoft wants it: at the back of the server room, out of sight.

Are you willing to make that compromise? Some will be. But to others, this kind of solution represents the worst kind of vendor lock-in.

Which brings us back again to the topic of open source management tools. Should we really manage open source software with tools and protocols that are closed and proprietary? One could argue that this is, in fact, the last place you’d want to use a proprietary solution. So if open, standards-based alternatives truly don’t exist, I don’t see that as a failure of the open source community. Rather, it’s a tremendous opportunity.

It was such a nice change, I began to wonder how well Ubuntu would run on some of my older PCs. Alas, the resulting installations were a little slow, and in one case installation failed completely. But a little research turned up tricks to get Ubuntu sprinting even on antiquated systems. Here are a few basic things I did to adapt Dapper Drake to old hardware.

Speed

Get the alternate CD: To install Ubuntu, you must first download the live-CD version from one of the mirror sites listed on Ubuntu’s page (www.ubuntu.com/download) and burn it; you then boot your PC with the disc, and install. However, PCs with less than 192MB of RAM may not be able to boot with the standard CD, so you’ll need to download the alternate install CD (lower down on the same mirror page) instead.

Cut down on the eye candy: Ubuntu’s default graphical interface is the Gnome desktop environment. Though it’s very slick, it can be slow on older graphics adapters. One way to speed it up is to avoid complex themes and backgrounds. Choose System-Preferences-Theme, select a lightweight theme like “Simple”, and click on Close. You can also save memory and avoid swapping to disk by skipping desktop background images: right-click the desktop, choose Change Desktop Background, select No Wallpaper, and click Finish.

Not all roads lead to Gnome, of course – you may prefer to use the KDE environment with Ubuntu instead. The Kubuntu distribution (www.kubuntu.org) replaces Gnome with KDE but otherwise installs and works much like Ubuntu. You can also just add KDE to an existing Ubuntu installation using Synaptic Package Manager, and choose one or the other environment from the log-in screen’s Options menu. Like Gnome, KDE can be slow on older hardware. To speed it up, choose Settings-Control Center in the KDE menu, click on Theme Manager under “Appearance & Themes”, click on Style under “Customize your theme”, select the Effects tab, and uncheck “Enable GUI effects”. Then click OK and Apply to finish.

Try a slimmed-down window manager: Another way to accelerate Ubuntu is to choose a window manager that is fundamentally less hardware hungry, such as the Xfce 4 Desktop Environment. You can add Xfce to your existing Ubuntu installation through Synaptic, or download the Xubuntu distribution (www.xubuntu.org), which by default installs Xfce in place of Gnome. Xubuntu is also a good choice for machines with very little memory, as Xubuntu’s alternate install CD works on computers with less than 128MB of RAM.

Dig deeper: Performance is more than skin deep. Ubuntu’s online forums contain additional tips and how-tos on regaining even more precious processor cycles by disabling unneeded services, speeding up Firefox, deleting unnecessary files, optimising OpenOffice, and more. Start with the “Improve performance in Ubuntu” thread at www.ubuntuforums.org/showthread.php?t=189192/ and don’t stop until Ubuntu has reached maximum speed.

Safety

Linux has a reputation for being relatively impervious to attack via the Net. But is it really immune to the threats that stalk Windows? The answer has long been, and still is, a qualified yes. Sure, occasionally viruses, worms and rootkits are written specifically to compromise Linux systems. But if you avoid a couple of major blunders, those relatively few threats are unlikely to do you any real harm. To sleep utterly soundly, however, you may still need to take some steps to malware-proof your Linux system.

The first reason your Linux system is probably safe from attack is that recent versions of Linux (kernels 2.4 and 2.6) include a built-in firewall called “iptables” that simply drops all uninvited incoming connections by default. If a worm or a person tries to break into your Linux box from afar, or not so afar, iptables simply turns away and ignores the incoming connection – the attacker won’t know whether a system even exists at the attacked address.

If you don’t run mail, Web, FTP, or other servers on your Linux system, you’ll probably never need to modify iptables’ default settings. However, if you use the Samba server suite to enable file and printer sharing with other local systems, you will have to enable incoming connections to the Samba server in iptables.

In SuSE 10.1, the Yast2 configuration tool lets you make this change in a few clicks: select Firewall in the “Security and Users” section, click Allowed Services in the left panel, pick Samba Server from the “Service to Allow” menu, and then click Add-Next-Accept to make the change. If your distribution lacks a similar interface, try Luigi Genoni’s graphical iptables configuration interface, Knetfilter (http://expansa.sns.it:8080/knetfilter/). And, of course, you can configure iptables from the command line; Troy Johnson’s Samba and IPTables page (http://troy.jdmz.net/samba/fw/) shows you how.

Antivirus

By default, most Linux distributions wisely set you up as a lower-privilege user, with a type of account that generally can’t allow malicious code to take over the system. Were you to somehow permit a Linux virus or worm to run on your computer, the fact that you are not logged in on the all-powerful root account prevents the malware from attacking the system’s (and other users’) files. That’s why you typically have to log in as root (with the root password) to change Linux system configurations. In contrast, the default Windows XP user account is the full-privilege administrator, which gives viruses and other threats greater leeway to infect and damage the PC.

However, should someone you know lose their head and log in as root, and then run a program that happens to be infected with one of the few Linux viruses known to exist in the wild, that virus could definitely destroy or steal user data. And even if you think you’re smart enough to avoid infection, the files moving through your mailbox, Samba file shares (shared files mounted as drives), and other storage mechanisms could still contain viruses destined to infect other systems. These are valid reasons why you might want to use antivirus software on your Linux computer.

Although commercial Linux antivirus products exist, save your money and start with one of several excellent free utilities instead. Even better, pick one that is not only free but open source: the delightfully named Clam AntiVirus (www.clamav.net/binary.html#pagestart). In addition to binary packages (no compiling necessary) for most major Linux distributions, ClamAV is also available for Windows and Mac OS X.

If you’d prefer antivirus software with a commercial track record, try the Linux version of Grisoft’s renowned AVG Anti-Virus Free Edition (http://free.grisoft.com/). Because of Linux’s inherent safety, the system you save may not be your own, but you’ll be doing your part to curb malware altogether.

In this article I’ll focus on PHP, the technology Oracle has chosen to incorporate into its products, and ASP.NET. I’ll overview the various strengths and weaknesses of each, discussing in particular those areas that will help you make your decision on which to go with for your development project. There are a lot of factors to consider, and different projects may appeal to a different technology. In conclusion you’ll find a point-by-point comparison in terms of price, speed and efficiency, security, cross-platform support, and the advantages of an open source solution.

What is ASP.NET?

The latest incarnation of ASP, ASP.NET, is not completely backward-compatible with previous versions of ASP, as it is a complete rewrite of the software. Previous ASP technology actually has a lot more in common with PHP than with ASP.NET, which is a complete framework for building Web applications. One of the principal features of this model is the flexibility to choose your programming language. ASP.NET works with scripted languages such as VBScript, JScript, Perlscript, and Python, as well as compiled languages such as VB, C#, C, Cobol, Smalltalk, and Lisp. The new framework uses the common language runtime (CLR); your language source is compiled into Microsoft Intermediate Language code, which the CLR then executes.

The framework also provides for true object-oriented programming (OOP), and true inheritance, polymorphism, and encapsulation are supported. The .NET class library is organized into inheritable classes based around particular tasks, such as working with XML or image manipulation.

Besides the programming language and the methodology, database access is a significant concern. When you program in ASP.NET, integration with databases can be accomplished through ODBC, which provides a consistent set of calling functions to access your target database.

Strengths and Weaknesses

ASP.NET’s strength lies clearly in its clean design and implementation. It is an object-oriented programmer’s dream, with language flexibility, and with sophisticated object-oriented features supported. In that sense, it is truly interoperable with your programmers’ existing skills.

Another strength of ASP.NET is the development environment. For instance, developers can use WebMatrix, a community-supported tool, Visual Studio .NET, or various Borland tools such as Delphi and C++ Builder. Visual Studio, for instance, allows setting of breakpoints, tracing sections of code, and reviewing the call stack. All in all, it’s a sophisticated debugging environment. Plenty of other third-party IDE solutions for ASP.NET are certain to surface as well.

But what you gain in robustness, you pay for in efficiency. ASP.NET is expensive with respect to memory usage and execution time, which is due in large part to a longer code path. For Web-based applications, these limitations can be a serious problem, because on the Web, your application is likely to scale to thousands and thousands of users per second. Memory usage can also become an issue on your Web server.

What is PHP?

PHP is a scripting language based on the model of preprocessing HTML pages. When the PHP preprocessor in your Web server notices a PHP language tag like the following, the PHP engine is invoked to execute that code:

< ?php
some code here
?>

PHP will be familiar to any programmers who have worked with imperative programming languages; you’ll notice syntactical similarities with Perl, C, and Java. Strictly speaking, Java is an imperative programming language, but it also makes use of object-oriented constructs and concepts. PHP borrows from this structure when it is convenient, but it is not a pure OOP language.

In the discussion of ASP.NET above, I mentioned the ODBC driver, and how applications can be built with database abstraction in mind. In PHP, you can also use ODBC to talk to databases, so you already have a whole list of supported databases to choose from. There are also native drivers for MySQL, Oracle, and Postgres. Furthermore, if you are connecting to Oracle, a special OCI8 library provides more feature-rich access to Oracle, allowing you to use such features as LOB, BLOB, CLOB, and BFILE.

You might ask, at this point, “Why are database-dependent libraries being touted as features of PHP?” Database abstraction, or independence, is a feature if you seek to build an application that works with multiple databases in one application or that can easily be ported to another database—when moving from development to production, for instance. And these are indeed valid concerns and considerations.

But, as Tom Kyte points out in his latest book, Effective Oracle by Design (Oracle Press), database dependence should be your real goal because you maximize your investment in that technology. If you make generic access to Oracle, whether through ODBC or Perl’s DBI library, you’ll miss out on features other databases don’t have. What’s more, optimizing queries is different in each database.

Zend Technologies, a commercial software company that contributes significantly to PHP, has created a commercial-development environment called Zend Studio that includes a sophisticated debugger, a profiler, and other features. It has also built the free Zend Optimizer, which, in combination with the Zend Encoder, compiles PHP code to speed performance. Additional commercial products also exist, such as the Zend Performance Suite, which can cache precompiled PHP pages, further speeding overall performance tremendously.

Strengths and Weaknesses

As of beta version 4, PHP 5 still has a few shortcomings, including its lack of exceptions, event-based error-handling instances that interrupt the normal flow of a program, jumping your code to a special error-handling section. Java also provides exceptions for error handling, while C++ provides exception handling via the try, catch, and throw syntax. You can, of course, still manage errors in PHP, but the structure is not standardized, so programmers are left to their own devices on how to implement error handling, leading to less consistency and a tendency to reinvent the wheel.

Another weakness is that PHP’s function names are case insensitive. Some programmers might find this feature annoying, though this isn’t a serious drawback.

I do have misgivings about PHP’s object model, however. PHP wasn’t designed to be an object-oriented language. Some of those features were added later, although care was made to keep backward compatibility with PHP 3, so you’re left with a bit of both models. In fact, many of these weaknesses are addressed in PHP 5. Keep your ears to the ground.

What PHP lacks in a few areas, it makes up for by leaps and bounds in areas in which it excels. The price is right, so you don’t have to worry about licensing issues. It’s open source, too, so an entire community will keep a close eye on development, identifying bugs and making sure they get fixed. And if there’s a feature you don’t like, you can dabble with the code. What’s more, PHP works native with Apache: It can be compiled as a module or directly into the Apache binary.

But running on Apache means that, with PHP, you can take advantage of whatever server investments you’ve already made, because Apache runs on Windows, Linux, Solaris, and various other Unix platforms. Also, going with a web server with Apache’s track record means security remains a top priority. And, finally, PHP has a smaller code path, meaning there’s less server-side code executed to parse and execute your PHP page, which results in more efficient memory and usage and faster execution.

What’s New in PHP 5?

The fourth beta release of PHP 5 came out at the end of December 2003, and the change log makes it obvious that many bugs are being identified and ironed out. Although it’s still in beta, it’s definitely worth taking a look at for all the new features and advances.

PHP 5’s major new achievements come in the area of its exception handling and a new object that introduces features that bring true OOP to PHP. Exception handling was certainly one of the most noticeable missing features in PHP 4, and its addition to PHP 5 is certainly a sign of maturity. Exception handling means you have language defined and standardized ways of handling errors in your software. Just use the try, catch, and throw methods, and your PHP code becomes more robust and clean.

< ?php

class blue {

function openFile ($inFile) {
if (file_exists ($inFile)) {
# code to open the file here
} else {
throw new Exception
("Cannot open file: $inFile");
}
}
}

$blueObj = new blue ();

try {
$blueObj->openFile (’/home/shull/file.txt’);

} catch (Exception $myException) {
echo $myException->getMessage ();

# rest of exception handling code here
}

# rest of blue methods here

?>

The new object model has a number of positive impacts on programs written in PHP. In PHP 4, when an object was passed to a function or method, it was passed by value, unless you explicitly told PHP otherwise. This procedure meant that a copy of that object, all the data structures in memory, would have to be copied. This step used memory and made access slow and clunky. In PHP 5, however, objects are always passed by reference.

The new object-oriented features in PHP 5, including constructors and destructors, are noteworthy. As with C++ and Java, they provide a standard way to create the object, allocate memory, and do any necessary setup via a constructor method and perform cleanup with a destructor method.

PHP 5 also introduces more subtle control of methods and variables in your classes. In PHP 4, everything was public: You could access variables from your classes outside the class or in derived classes. In PHP 5, you can still make variables or methods public, but you can also make them private, so they’re used only within the class itself. A third option is to make them protected, which means that methods and variables can be viewed within the class or when subclassed.

Furthermore, PHP 5 introduces type hinting, or better type checking. When you pass an object into a routine, PHP can check that it is the right type and give a type-mismatch error if the check fails.

Additional features such as static methods and variables and abstract classes exist, so be sure to check the documentation for details.

Security Comparison

ASP.NET officially requires that you use IIS. Unfortunately, IIS has a long history of vulnerabilities, which makes many administrators reluctant to deploy it to handle their web site. Whether these weaknesses are because of Microsoft’s ineptness or because IIS is a real red flag to hackers is irrelevant: Those systems have a history of being hacked and compromised. PHP runs on Apache, too, which is fast and open source and has a good security track record. Also, as I mentioned, Apache runs on many platforms.

If you are considering ASP.NET but you want to use Apache for your front-door to the Internet, you are fortunate to have a few options. First, you can use Apache to relay requests to IIS running internally on another machine. Apache then handles static content and passes aspx content on to the IIS server, which isn’t exposed to the internet.

However, if you want to host ASP.NET with Apache, a couple of options are available that may or may not be supported by Microsoft. As a last alternative, there is Ximian’s Project Mono, which is working to build an open-source module. Check www.go-mono.com for more information.

Database Coding Examples

Connecting to a database is one of the first things you’ll consider doing in PHP or ASP.NET. With ASP.NET, however, it’s a little more complicated, because you have the option of any of a number of languages to choose from. Of course, these code samples would have to be embedded into an HTML page, the classes instantiated, and so on. The following information, however, will give you an idea of the coding styles for each.

PHP 5 Connecting to Oracle

Here’s a PHP 5 class that provides an Oracle connect-and-disconnect routine to show one way of connecting to Oracle with PHP 5 (other drivers, such as the ODBC driver, and generic database interfaces can be used as well):

class oracle_object {
protected $theDB;
protected $user;
protected $pass;
protected $db;

function __construct($u, $p, $d) {
$this->user = $u;
$this->pass = $p;
$this->db = $d;
}

function db_open () {
$theDB = @OCILogon($this->user, $this->pass, $this->db);
db_check_errors($php_errormsg);
}

function db_close() {
@OCILogoff($theDB);
db_check_errors($php_errormsg);
}

function __destruct () {
print (”so long…”);
}

}

ASP.NET Connecting to Oracle

If you’re looking to connect to Oracle with VB.NET (Visual Basic is Microsoft’s default .NET programming language), take a look at this sample from MSDN:

Imports System
Imports System.Data
Imports System.Data.OracleClient
Imports Microsoft.VisualBasic

Class Sample

Public Shared Sub Main()

Dim oraConn As OracleConnection = New OracleConnection(”Data Source=MyOracleServer;Integrated Security=yes;”)

Dim oraCMD As OracleCommand = New OracleCommand(”SELECT CUSTOMER_ID, NAME FROM DEMO.CUSTOMER”, oraConn)

oraConn.Open()

Dim myReader As OracleDataReader = oraCMD.ExecuteReader()

Do While (myReader.Read())
Console.WriteLine(vbTab & “{0}” & vbTab & “{1}”, myReader.GetInt32(0), myReader.GetString(1))
Loop

myReader.Close()
oraConn.Close()
End Sub
End Class

Making the Choice

Without assuming you’ve already decided to go with PHP, I’ll conclude that its strengths far outweigh its weaknesses. (See the summary in Table 1.) It boils down to price, speed and efficiency, security, cross-platform applicability, and open-source opportunity. Its only weakness is its lack of a pure and perfect OOP implementation; however, this is a minor drawback. Though language constructs do help, ultimately, good coding is a matter of practice, execution, good habits, and discipline.

Price. Here, we must consider not simply the price tag of the initial investment, which, in the case of PHP, is obviously free, but also the implementation, maintenance, and debugging costs. In the case of PHP, you may invest in the Zend optimization engine. With ASP, however, you’re investing from the very beginning, and you’re spending for add-on technologies—libraries for doing graphics manipulations, for instance. But, in the long term, PHP isn’t going to press you to upgrade and collect more licensing fees. Everyone who has dealt with complex licensing also knows that companies spend time and money just ensuring they are compliant. Furthermore, you have a difference in response when getting bugs fixed. This, of course, translates to time, which translates to cost for overall development.

Speed and efficiency. As I mentioned earlier, ASP.NET is a framework allowing you to use various programming languages. In addition, it is touted as having a great object-oriented model. All this is true, but it becomes a detriment as far as speed is concerned. For all that advantage, there is a lot more code to run through to execute the same ASP page than you have to execute in the PHP engine for an equivalent PHP page. PHP is the quick-and-dirty type of solution, the one to get the job done. And though a lot of robustness has been added to it since its 2.0 and 3.0 days, it still retains that core optimized high-speed approach.

Speed is not the only consideration. Memory usage is also important.

Security. ASP.NET runs on IIS, which has been compromised innumerable times, as evidenced by IT news reports every other week. It has become such a liability, in fact, that in spite of all the marketing dollars spent on it, many IT professionals refuse to have their networks exposed with an IIS Web server. PHP, however, works with Apache, which has a proven track record of speed, reliability, and hardened security. Check www.securityfocus.com for more information.

Cross-platform applicability. ASP.NET runs on IIS and is starting to run on Apache, which can run on a whole host of platforms. PHP has been designed to work with Apache from the beginning, so you have many proven and reliable server platforms to choose from.

Open source opportunity. Open source is not just some philosophical torch idealistic programmers, or companies wanting to save a few bucks on licensing costs, are carrying. When you’re dealing with bugs in the software itself, open source can be a serious godsend.

In either case, with PHP or ASP.NET, you have a large user base using the software and possibly encountering bugs. With ASP.NET, those bugs have to go through a bureaucratic process to get acknowledged, fixed, tested, and rolled out in a new patch or release. PHP fixes, however, can get fixed quickly and rereleased. Anyone who has watched open-source development knows new releases and patches often come out in days rather than in weeks or months, as with commercial software. If that’s not fast enough, you can always fix a problem yourself if you have to.

IBM is porting Lotus Sametime to Linux, the first time this enterprise instant messaging platform will support the popular open source operating system.

IBM will deliver a Linux version of the desktop client before mid-September in Sametime’s version 7.5, which is now in beta testing, an IBM official said.

Meanwhile, the server piece will ship in the first half of next year, IBM’s program director and senior product manager for real-time communication, David Marshak, said. IBM anticipates there will be client and server versions of Sametime for Linux operating systems from Red Hat and Novell.

IM is essential to communications in many workplaces, where IM systems have become communication hubs that provide, beyond text messaging, integration with telephony and email systems, business applications and Web conferencing.

IBM hasn’t estimated what percentage of customers will deploy Sametime on Linux, Marshak said. However, interest is high in industries like education and government, and among small and medium-size companies, because of the generally lower costs of Linux systems and the flexibility of open source, he said.

“The two major themes are choice and cost of ownership,” Marshak said.

Sametime’s biggest competitor is Microsoft’s Live Communications Server, which doesn’t run on Linux. IBM and Microsoft have locked horns recently as they fight for enterprise IM customers.

The porting of Sametime to Linux is yet another swipe IBM is taking at Microsoft, research director with AMR Research, Jim Murphy, said. With this move, IBM offers a lower-cost operating system alternative for Sametime as many customers ponder whether to invest in Vista, Windows’ upcoming upgrade, a likely more expensive and longer-term commitment, he said.

“There’s a lot of competitive positioning in this move,” Murphy said.

In June, IBM announced it was building links between Sametime and Microsoft’s Outlook, Office and SharePoint applications. The links, slated for early next year, will give users access to Sametime 7.5 functionality from within those Microsoft products, an attempt to lure those users to the IBM system.

In January, IBM announced Sametime 7.5 will be linked with the public IM networks from AOL, Yahoo and Google, but it apparently couldn’t reach an agreement with Microsoft over MSN Messenger. Marshak declined to comment on that issue, but the impasse apparently remains.

Other enhancements due in Sametime 7.5 include a refurbished user interface, improved privacy and security capabilities, embedded VoIP functionality and support for the Eclipse open-source application framework.

The Notes-Domino messaging and collaboration system, a sister product to Sametime, already runs on Linux.

Hewlett-Packard is throwing its support behind the Debian Linux distribution, the first major hardware maker to align itself with the noncommercial community-based Linux offering.

HP says it will support Debian Linux on its ProLiant and HP BladeSystem servers, and what it says is the industry’s first Debian Linux customizable thin client from a major vendor, the new HP t5725 Thin Client server.

HP is supporting Debian because it has been shipping Debian Linux servers to customers in the fields of telecommunications and high-performance technical computing, said Jeffrey Wade, open source and Linux marketing manager at HP. HP’s involvement with the Debian dates back to 1995, he said.

Support will be provided directly from HP rather than through a third party service provider as part of the warranty coverage for its Debian Linux servers, better support than customers can expect from other original equipment manufacturers, (OEMs) said Wade.

If you look for Debian Linux support on the Web sites of other OEMs, “you might see a link to some discussion groups or a download of a research paper on Debian,” he said. “But HP takes real phone calls from real customers who need to get their problems solved.”

Some enterprises are turning to Debian to avoid the subscription fees required of major commercial Linux distributions such as Red Hat, Novell SUSE and others, Wade said.

Debian is a free operating system that was created by a group of about 1,000 global volunteers who collaborated via the Internet on its development. The Debian Project announced July 24 that it’s next major update, Debian GNU/Linux 4.0, is scheduled for release in December.

Although Debian was not widely embraced at first and users experienced interoperability problems with application software, Debian “has matured as Linux has matured,” said Gordon Haff, principal IT advisor at research firm Illuminata.

“This really is something new where there is a tier-one vendor essentially providing a tier-one level of support for Debian,” said Haff. Although specific market share numbers are hard to come by for noncommercial Linux distributions, “Debian has been quite popular [and] the leading noncommercial distribution,” he said, although only in servers, not desktop computers.

HP’s direct support for Debian may give HP a competitive advantage, Haff said, but there are third party service providers who can support Debian Linux servers from Dell, IBM or other PC makers.

IBM said it takes care of its Debian Linux customers. “IBM works well with Debian in the Linux community and will, and does, support the Debian distribution for our customers,” the company said in a prepared statement. “It’s not a standard offering, but we do it under special bid.”

HP also announced Monday that unit sales of 1.5 million Linux servers generated revenue of close to US$6.2 billion for the 12 months ending in May, 50 percent more revenue than its nearest competitor.

The flavors of Linux are referred to as distributions (often shortened to “distros”). All the Linux distributions released around the same time frame will use the same kernel (the guts of the Operating System). They differ in the add-on software provided, GUI, install process, price, documentation and technical support. Both Linux and Windows come in desktop and server editions.

There may be too many distributions of Linux, it’s possible that this is hurting Linux in the marketplace. It could be that the lack of a Linux distro from a major computer company is also hurting it in the marketplace. Perhaps this will change with Novell’s purchase of SuSE. IBM is a big Linux backer but does not have their own branded distribution.

Linux is customizable in a way that Windows is not. There are many special purpose versions of Linux above and beyond the full blown distributions described above. For example, NASLite is a version of Linux that runs off a single floppy disk and converts an old computer into a file server. This ultra small edition of Linux is capable of networking, file sharing and being a web server.

Graphical User Interface

Both Linux and Windows provide a GUI and a command line interface. The Windows GUI has changed from Windows 3.1 to Windows 95 (drastically) to Windows 2000 (slightly) to Windows XP (fairly large) and is slated to change again with the next version of Windows, the one that will replace XP. Windows XP has a themes feature that offers some customization of the look and feel of the GUI.Linux typically provides two GUIs, KDE and Gnome. See a screen shot of Lycoris and Lindows in action from the Wal-Mart web site. The lynucs.org web site has examples of many substantially different Linux GUIs. Of the major Linux distributions, Lindows has made their user interface look more like Windows than the others. Here is a screen shot of Linux made to look like Windows XP. Then too, there is XPde for Linux which really makes Linux look like Windows. Quoting their web site “It’s a desktop environment (XPde) and a window manager (XPwm) for Linux. It tries to make easier for Windows XP users to use a Linux box.”

Mark Minasi makes the point (Windows and .NET magazine, March 2000) that the Linux GUI is optional while the Windows GUI is an integral component of the OS. He says that speed, efficiency and reliability are all increased by running a server instance of Linux without a GUI, something that server versions of Windows can not do. In the same article he points out that the detached nature of the Linux GUI makes remote control and remote administration of a Linux computer simpler and more natural than a Windows computer.

Is the flexibility of the Linux GUI a good thing? Yes and No. While advanced users can customize things to their liking, it makes things harder on new users for whom every Linux computer they encounter may look and act differently.

Text Mode Interface

This is also known as a command interpreter. Windows users sometimes call it a DOS prompt. Linux users refer to it as a shell. Each version of Windows has a single command interpreter, but the different flavors of Windows have different interpreters. In general, the command interpreters in the Windows 9x series are very similar to each other and the NT class versions of Windows (NT, 2000, XP) also have similar command interpreters. There are however differences between a Windows 9x command interpreter and one in an NT class flavor of Windows. Linux, like all versions of Unix, supports multiple command interpreters, but it usually uses one called BASH (Bourne Again Shell). Others are the Korn shell, the Bourne shell, ash and the C shell (pun, no doubt, intended).

Cost
For desktop or home use, Linux is very cheap or free, Windows is expensive. For server use, Linux is very cheap compared to Windows. Microsoft allows a single copy of Windows to be used on only one computer. Starting with Windows XP, they use software to enforce this rule (Windows Product Activation at first, later Genuine Windows). In contrast, once you have purchased Linux, you can run it on any number of computers for no additional charge.

As of January 2005, the upgrade edition of Windows XP Home Edition sells for about $100, XP Professional is about $200. The “full” version of XP Home is about $200, the full version of XP Professional is $300. Windows Server 2003 Standard Edition with 10 Client licenses is about $1,100. Because they save $100 or so on the cost of Windows, Wal-Mart can sell a Linux based computer for $200 (without a monitor) whereas their cheapest Windows XP computer is $300 (as of January 2005).

The irony here is that Windows rose to dominance, way back when, in large part by undercutting the competition (Macs) on cost. Now Linux may do the same thing to Windows.

You can buy a Linux book and get the operating system included with the book for free. You can also download Linux for free from each of the Linux vendors (assuming your Internet connection is fast enough for a 600 MB file and you have a CD burner) or from www.linuxiso.org. Both these options however, come without technical support. All versions of the Ubuntu distribution are free.

You can purchase assorted distributions of Linux in a box with a CD and manuals and technical support for around $40 to $80 (some distributions may be less, others may be more). Regular updates and ongoing support range from $35 a year for a desktop version of Linux to $1,500 for a high-end server version. August 2004 Red Hat started selling a desktop oriented version of Linux for under $6 per user per year.

After the initial cost (or lack thereof) of obtaining software, there is the ongoing cost of its care and feeding. In October 2002, ComputerWorld magazine quoted the chief technology architect at Merrill Lynch & Co. in New York as saying that “the cost of running Linux is typically a tenth of the cost of Unix and Microsoft alternatives.” The head technician at oil company Amerada Hess manages 400 Linux servers by himself. He was quoted as saying “It takes fewer people to manage the Linux machines than Windows machines.” Microsoft commissioned a study that (no surprise) found it cheaper to maintain Windows than Linux. However, one of the authors of the study accused Microsoft of stacking the deck by selecting scenarios that are more expensive to maintain on Linux.

I don’t know if there will ever be an objective measure of the ongoing care and feeding costs for Linux vs. Windows. If there were however, it would have to consider:

• Dealing with bugs in the operating system
• Dealing with bugs in application software
• Dealing with viruses, worms, Spyware, etc.(big advantage to Linux here
• Dealing with software upgrades to new versions (both the OS and applications)

Getting the Operating System

Nothing need be said about getting Windows. As for Linux, you have to look around to buy a new computer with it pre-installed. The major PC vendors sell Linux based machines only as servers, not to consumers.In August 2004 HP said it would be an exception to this rule and will soon offer a business notebook computer, the Compaq nx5000, with SUSE Linux pre-installed. A low end model was expected to sell for $1,140 — about $60 less than a comparable model running Windows XP. HP was also scheduled to sell Linux PCs for consumers in Asia as of June 2004.

Wal-Mart sells new PCs with Linux pre-installed. In fact, if you need a new computer, the cheapest ones, bar none, sell for $200 at Wal-Mart. Wal-Mart sells different Linux distributions: Lycoris, Lindows and SuSE. In March 2004, they started selling Sun Microsystems’ version of Linux, the Sun Java Desktop, starting at $298. In December 2004, Wal-Mart started selling a $498 laptop computer running Linspire. As of November 2005, MicroCenter sells Linux desktops starting at $250.

Lindows (now Linspire) maintains a web page listing vendors that sell computers with Lindows pre-installed. Emperor Linux loads Linux distributions on new laptops from Dell, IBM, Sony and others. Pogo Linux sells custom-built desktops loaded with both Linux and Windows. Their main customers are universities and government agencies. A handful of companies sell notebooks with Lindows/Linspire pre-installed.

If you are up to installing Linux yourself, you can buy a new computer without any operating system from Dell or Wal-Mart, perfect for installing your favorite flavor of Linux. Dell sells their PowerEdge servers without an OS, Wal-Mart sells some of their Microtel computers without an OS.

You can buy a 3GB MicroDrive (1 inch hard disk) with Ubuntu Linux. The drive plugs into a USB port and is fully powered by the USB port. As of mid-November 2005, it sold for $132. See Taking Linux On The Road With Ubuntu at Toms Hardware.

And what about installing Windows and Linux? Installing Windows from scratch is much easier than installing Linux from scratch, in my opinion. If nothing else, installing Windows is always the same whereas the different distributions of Linux have their own installation programs (these may even change with different versions of the same distribution). You can’t read an article on this however, without it saying how installing Linux is getting easier all the time.

Installing Linux on a computer without an OS is much easier than installing it on a machine with an existing OS that you want to preserve. The later should only be attempted by experts, it is all too easy to lose the pre-existing OS and setting up a dual-boot environment is tricky. In his Linux book, Mark Minasi said that installing Linux on a desktop computer was more likely to be successful than on a laptop computer.

Part of the difficulty in installing Linux is terminology and documentation. The install process is designed by Linux people for Linux people. A Windows-only person is likely to encounter terminology and concepts they are unfamiliar with. With Red Hat Linux 8, the booklet on how to install the OS is over a hundred pages.

October 2005 saw a new development that made buying a computer with Linux pre-installed or installing Linux yourself, unnecessary and undesirable. This new development is also preferable to the next topic, running Linux from a CD. I’m speaking of the VMware Player. This free software from VMware lets you run a virtual machine created by someone else. As of October 28, 2005 you could freely download pre-built virtual machines for three distros: SUSE Linux Enterprise Server, Novell Linux Desktop and Red Hat Enterprise Linux. Virtual machines let you run multiple operating systems on one computer at the same time. You can’t beat it.

Running from CD

One thing that Linux can do that Windows can not, is run from a CD. To run Windows, it has to first be installed to your hard disk. That said, if Windows is broken to the point that it can’t start up, there is a free program called Bart’s Preinstalled Environment (BartPE) that can run a few Windows programs from a bootable CD. However, this is not from Microsoft and it can only run a handful of programs that have been set up ahead of time for use with BartPE. BartPE fills an important need, but creating the CD is not trivial and it only works with Windows XP and 2003. The main point stands, in and of itself, Windows can not run from a CD.

Normally Linux also runs from a hard disk, but there are quite a few versions of Linux that run completely from a CD without having to be installed to a hard disk (the term for this is a “Live” CD). This is a great way for Windows users to experience Linux for the first time. Among the Linux distros that have a CD-only version are Knoppix, SuSE (called Live-Eval), Linspire (formerly Lindows) and Slackware. The version of Linspire that runs from a CD used to be called Lindows CD, then it was called Lindows Live and now it is called Linspire Live. It’s available for download, but only with BitTorrent (see Lindows offers software for free over P2P from CNET News.com, January 2004). SuSE’s Live-Eval is free, I haven’t checked the other distros. I tried SuSE Live Eval version 9 in October 2003 and had some gripes. The SUSE LINUX Professional 9.3 Live version (as of August 2005) runs off a DVD rather than a CD – the ISO file is 1.5GB. FreeBSD, a version of Unix (rather than Linux), also has LiveCD to run it from a bootable CD.

The CD based versions of Linux differ in their use of the hard disk. Some, such as Lindows, do not write anything at all to your hard disk, making it the safest and easiest way to experience Linux. The downside of this is speed (CDs are much slower than hard disks) and continuity (being able to save data between uses). Other versions, such as SuSE, do use your hard disk (SuSE creates over 200 MB worth of files). What you give up in safety, you gain in speed. For reviews of Linux distributions that run from a CD see A Taste of Linux by Jim Lynch at ExtremeTech January 23, 2004 and A Taste of Linux, Part Two By Jim Lynch March 5, 2004.

In addition to kicking the tires, a CD based version of Linux can also be used to insure your hardware is supported by Linux and possibly to recover files. If Windows is not able to boot up and there are files you need on your computer, booting Linux from a CD may offer the opportunity to copy files off the computer. This is problematic however with the NTFS file system. I found that the CD based versions of Lindows and SuSE could not read files stored in an NTFS partition. I am told that Knoppix can.

Application software: There is more application software available for Windows.

Obtaining application software: If you buy a copy of Windows on a CD-ROM, you get no application software with it. If you buy a copy of Linux on a CD-ROM (or two or three) it typically comes with gobs of free application software. A new computer with Windows pre-installed may have additional application software, this is up to the PC vendor. I have seen a new Sharp laptop machine that came with no software other than Windows itself. In contrast, Sony VAIOs, for example, come with a lot of software. Often however, necessary software, such as the Adobe Acrobat reader, is not pre-installed by the PC vendor. Each Linux distribution comes in multiple flavors, the more expensive versions come with more application software.

Application software installation: The installation of applications under Windows, while not standardized, is generally consistent. Installing software under Linux varies with each distribution and is not nearly as simple, easy or obvious as Windows. Lindows lets users install software in a manner somewhat like Windows update – applications are downloaded and automatically installed. They charge for this service however. The May 20, 2004 issue of the Langa list newsletter had an interesting take on the difficulty installing software under Linux. In a July 4, 2004 review of Linux in the Washington Post (Linux, Still an Awkward Alternative), Rob Pegoraro called application software installation “Linux’s biggest embarrassment” and I have no doubt that he is correct.

Viruses and Spyware: (Updated March 2005) There are many types of malicious software programs. The most common types are referred to as Viruses and Spyware. Spyware has become a generic term, much like “Xerox machine” (which is taken to mean any copying machine, not just those made by the Xerox corporation). The term “Spyware” now refers to a whole host of malicious software such as worms, Trojans, dialers, keystroke loggers, browser hijackers and, of course, actual Spyware. The vast majority of all malicious software (of all types) runs on Windows. I don’t know the actual percentages, but it wouldn’t surprise me if it was 98% or so.

Spyware on Windows has become such a problem that Microsoft purchased an anti-Spyware software company and released their product as the Microsoft Anti-Spyware program in early 2005. As this is written the product is still in beta form, but Microsoft has stated that it will be free even when complete. In my opinion, Spyware is the worst problem effecting Windows based computers. In addition to running an anti-virus program constantly, Windows users also need an anti-Spyware program constantly running in the background to protect them.

Users and Passwords: (updated August 2005) Both Linux and Windows 2000/XP require a userid and password and boot time. That said, Windows XP allows the omission of the password (a very bad idea), I’m not sure if Linux does. All can be configured to either ask for the userid/password at startup time or defaults can be set instead. In Windows 2000 it is very easy to set a default userid/password, in Windows XP the method varies between the Home and Pro versions – in one it is straightforward, in the other it’s a pain. I’ve been told that in Linux the KDM and GDM login managers support automatic login.

A new Windows XP machine used by a home user is likely to not ask for a userid/password at start-up. However, this depends on the number of users defined to Windows. When you create a new user in XP the default is not to require a password (user friendly triumphs over security – the Microsoft way). Windows 98, never mind.

Windows XP, 2000 and Linux all support different types or classes of users. Windows XP Home Edition supports Administrator class users that have full and total access to the system and restricted users that, among other restrictions, can’t install software. Windows XP Pro and Windows 2000 support additional levels of users.

Both Linux and Windows can group users into groups (finally, something well named) and assign privileges to the group rather than to each individual user. Windows XP and 2000 come with some pre-defined user groups (such as Power Users), I don’t think Linux does (but I’m not sure). In XP and 2000 user “Administrator” is a member of the Administrators group.

Linux privileges are basically whether you can read, modify or execute a file. Files in Linux are always owned by a specific user and group. Windows has similar file-related privileges but only when using the NTFS file system. The earlier FAT and FAT32 file systems had no file level security.

I’m told that Windows NTFS permissions are a bit more functional than those in Linux, but that Linux distributions are starting to use extended Access Control Lists as a part of the file system, bringing them more on par with NTFS. (I won’t swear by this)

Very often Windows users use an Administrator class userid which gives viruses total access to their system (see Why you should not run your computer as an administrator from Microsoft). This obviates the security rules discussed above. In contrast, Linux users often run as regular non-root users which not only means better security it also means that, if they get a virus, the operating system greatly restricts what the virus can do.

I have tried a couple times to set up a Windows XP computer for use by a family, creating Administrator class userids for the parents and restricted userids for children. Both cases failed because there were too many programs that did not function correctly when run by a restricted Windows user (more details here). While Linux has supported the concept of root and restricted users from the get-go, this is a relatively new thing to Windows. It will be a long time before all Windows software is designed to be used by a restricted user. Until then, viruses and malware will have free reign on Windows machines. Certainly software written for Windows 95, 98 and Me expects total system access and may not work when run from a restricted userid.

See also Linux vs. Windows: Which Is More Secure? from eWeek March 30, 2004 and

Windows v Linux security: the real facts by John Lettice October 22, 2004 in The Register

Security Report: Windows vs Linux by Nicholas Petreley October 22, 2004

Bugs: All software has and will have bugs (programming mistakes). Linux has a reputation for fewer bugs than Windows, but it certainly has its fair share. This is a difficult thing to judge and finding an impartial source on this subject is also difficult. Fred Langa wrote an interesting article on whether Linux or Windows has fewer bugs in Information Week magazine January 27, 2003. The article also addressed whether known bugs are fixed faster with Linux or Windows. In brief, he felt that bugs used to be fixed faster in Linux, but that things have slowed down. See this article too Security research suggests Linux has fewer flaws December 13, 2004 from CNET News.com.

In March 28, 2003, Microsoft decreed that it will not issue a Windows NT4 bug fix for a security problem that effects Windows 2000, XP and NT4. They would prefer customers to move off of NT4, thus making Microsoft more money. It is their ball, their bat and their field. This is not true with an open source operating system such as Linux. No one can decree that a bug will not be fixed in a specific version of Linux.

The difference in OS development methodologies may explain why Linux is considered more stable. Windows is developed by faceless programmers whose mistakes are hidden from the outside world because Microsoft does not publish the underlying code for Windows. They consider it a trade secret. In contrast, Linux is developed by hundreds of programmers all over the world. They publish the source code for the operating system and any interested programmer, anywhere in the world can review it. Besides the wide audience for peer review, there is likely to be pride of ownership on the part of the developers of Linux that can not exist with Windows. The official term is “egoboo” which refers to the rush a programmer gets from public recognition, especially for something done for free. Bruce Perens (Business Week magazine, March 3, 2003 issue) commented on why open source software works well:

…it taps into the true motivation of programmers in a way that corporations often don’t. Programmers are like artists … They like to showcase their best stuff for their peers. In open source, they can. But at most corporations, their best work is hidden behind locked and guarded doors.

As to Microsoft’s OS development methodology, Michael Miller of PC Magazine reported that Chris Jones, Vice President of the Windows Client team, “stressed” that Microsoft’s priority when developing Vista “was to change the process of developing Windows to ensure that the underlying code was tighter and more secure.” Vista will be the tenth version of Windows produced by Microsoft and they admit that the development process needs to be improved. See Vista: I Can See Clearly Now July 25, 2005.

Note: Shortly after I revised this section, all heck broke loose on the Internet due to a bug in the Plug-and-Play portion of Windows (see Fast-Moving Worms Slam Media, Enterprise Networks eWeek August 17, 2005). On August 9, 2005 Microsoft issued a fix for this bug. Within a week there were a dozen different malicious programs actively attacking computers without the bug fix. Many large companies were hit. This is par for the course, but it in light of this topic, it occurred to me that no story on the problem looked to find/blame/name the programmer that wrote the buggy Plug-and-Play code. No doubt, the world will never who caused all this grief. When faceless programmers can hide behind a corporate wall, these things are more likely.

Microsoft claimed that the Plug-and-Play bug only affected Windows 2000. Anti-virus companies claimed to have seen many other versions of Windows get infected. Was Microsoft lying to minimize the bad publicity? Were anti-virus companies trying to make their products seem more useful? Without access to the source code for Windows, no one can know for sure. I wonder if Microsoft fired the person(s) who wrote the buggy code or if they even know who did it?

Is It Soup Yet?

When is a new version of an Operating System done? Hard to say. All software has bugs and OSs, being very large software are very likely to have many bugs. When a new version of an Operating System is being developed, someone, at some point, has to say “enough is enough”. That is, someone decides that a reasonable number of bugs have been found and fixed and the OS can now be considered finished. Linux and Windows differ greatly in how this decision is made.

With Linux, the decision is made by a computer nerd with a public reputation to protect. With Windows, the decision is made by business people with billions of dollars in profit at stake. Speaks for itself.

Be sure to read the excellent How Microsoft’s Misunderstanding of Open Source Hurts Us All by Robert X. Cringely (October 23, 2003). The article argues in favor of Linux and ridicules comments by Steve Ballmer. Addressing the “Is it soup yet?” issue, this article includes a quote from Mr. Linux, Linus Torvalds (the nerd with the reputation to protect):

“Because the software is free, there is no pressure to release it before it is really ready just to achieve some sales target. Every version of Linux is declared to be finished only when it is actually finished, which explains why it is so solid. The other reason why free software is better is because the personal reputation of the developer is attached to every release.”

In contrast, Microsoft makes huge profits on a new version of Windows. Whoever decides when a new version of Windows (currently Longhorn/Vista) is ready to ship, has a lot more on their plate than just bug fixes. No doubt, Microsoft defenders will cite the many months, if not years, that a new version of Windows is beta tested by thousands of people before it is released. It’s not enough. Anyone who recalls the bragging done about stability when both Windows 2000 and XP were released must wince with the hundreds and hundreds of bug fixes both versions have required over the years.

Microsoft defenders may also note that Windows is a frequent target of attack by the bad guys just because it is so widely employed. True. But this just magnifies the fallout from any corners that may get cut on the way to releasing a new version of Windows.

He’s Dead Jim: On the other end of the life cycle, just as a particular version of Windows gets reliable, and everyone gets used to its quirks and other software interfaces with it well, Microsoft walks away from it. Not enough profit in it.

I am not familiar with the retirement policies of the various Linux vendors regarding old versions of their Linux distributions. But at least there is choice in the Linux marketplace.

Software restrictions: A program written for Linux will not run under Windows and vice versa. For example, Microsoft makes a version of Office for Windows and another version for the Mac. They are two different products, each capable of only running on the operating system it was designed for. There is no version of Microsoft Office for Linux. Some programs, such as Firefox, are available for multiple operating systems (Firefox runs on Linux, Windows, Macs and more). Vendors of such software go to the trouble of making different versions for each supported operating system.

This is the rule, but there are a fair number of exceptions.

The most ambitious exceptions allow for installing one operating system under another. For example, on a computer running Linux (referred to as the host or native OS), you can install a copy of Windows (referred to, in this case, as the guest OS). In the Windows OS running under Linux, you can install any and all Windows programs. Somewhat like a split personality, one computer can run two operating systems at the same time. The guest operating system runs in an environment referred to as a virtual machine (VM). A company called VMware was the first to market with a virtual machine product (also called VMware) for personal computers. Virtual PC, competes with VMware. The two products differ in their supported host and guest operating systems. Virtual PC used to be from Connectix but Microsoft purchased them in 2003. The first release from Microsoft, Virtual PC 2004 will run Linux, but Microsoft does not provide support. Microsoft also lowered the price to $129.

Win4Lin, by NeTraverse is also a virtual machine product, but not as full featured as Virtual PC or VMware. It provides a virtual Windows environment under Linux for Windows 95/98/Me. At $89 though, it is much cheaper than the more feature laden virtual machine products and still lets you run two operating systems concurrently. (read a review). There are two flavors of Win4Lin, a standard desktop version and a terminal server version. For more see Software maker weds Linux PCs, Windows applications from CNET News.com April 22, 2004.

Less ambitious than virtual machine products (which let you run any application on the guest OS) are Linux distributions that attempt to run a handful of Windows programs directly. The goal here is enable migration of desktop users from Windows to Linux while still being able to run some Windows applications under Linux. The three Linux distributions that do this, all include a product called CrossOver Office from CodeWeavers that allows running Microsoft Word, Excel, PowerPoint, Outlook (from Office 97 and Office 2000), Visio 2000, Intuit’s Quicken and Lotus Notes directly under Linux. The first Linux distribution to offer this feature was Lindows. Xandros Linux (previously Corel Linux) released their Desktop 1.0 distribution in October 2002 with better Windows compatibility than Lindows (see ExtremeTech review). SuSE Linux Office Desktop was released in January 2003 and can run some Microsoft Office programs (read more about this from ExtremeTech, USA Today, ZDNet).

For more on the subject of running selected Windows applications directly under Linux read: CrossOver Office Professional 3.0.1 by Jim Lynch at ExtremeTech (June 2004), Run Microsoft Office Without Windows from PC Magazine (April 2002), Breaking Windows: CodeWeavers and NeTraverse Bring Office to Linux from Open for Business (May 2002) and Linux Offers Better Windows Apps Without the Wait by Steven J. Vaughan-Nichols in eWeek (May 2004).

CrossOver also can be purchased separately (CodeWeavers sells it for $50) and installed on other Linux distributions. Support for Office XP is expected in the future. Front Page and Access are not supported. This approach does less but costs less – there is no need for a Windows license and no need for a VM product.

CrossOver is based on the Wine project, which attempts to allow a handful of important Windows applications to be more or less usable in a Linux environment. Wine is free software and is included with many GNU/Linux distributions. CrossOver Office is a commercial version of Wine that you have to buy. Wine does not, and will not, provide 100% compatibility; it will never run all Windows applications perfectly. Even applications that its runs well may not have all their features and functions available. Mary Jo Foley wrote that the Microsoft WGA (Windows Genuine Advantage) program looks for Windows programs running with Wine and purposely generates an error. See Microsoft Seeks to Bottle Up Open Source Wine (February 18, 2005).

TransGaming makes WineX which allows Linux users to run various Windows games that require OpenGL and DirectX. The combination of CrossOver Office and WineX should let you run many Windows programs under Linux and is likely to be cheaper than VMware or Virtual PC. In July 2004 TransGaming gave WineX a new name, Cedega, and announced support some DirectX 9 games. (Linux Takes on Windows Gaming from ExtremeTech July 28, 2004)

On yet another front, Ximian (owned by Novel) is developing Project Mono, which will allow Microsoft .Net applications to run on Linux. This should be very very exciting (at least to us nerds).

Microsoft has no interest in supporting Linux applications under Windows. The virtual machine products just mentioned however can do this.

There is yet another approach to crossing the OS divide that goes by the names Terminal Services, Server Based Computing and Thin Client Computing. On the Windows side, the Terminal Services feature in Windows 2000 Server and Windows 2003 server can be supplemented with software from other companies (Citrix being the market leader) to provide server-based Windows applications to non-Windows machines. I used an earlier version of the Citrix software, then known as WinFrame, to provide Windows NT server-based applications to Windows 95 and 98 clients. Quite cool. On the Linux/Unix side, Tarantella supports the display of Unix server applications on non-Unix, non-Linux machines.

Supported Hardware Devices: More hardware works with Windows than works with Linux. This is because hardware vendors write drivers for Windows more often than they do for Linux. When Windows XP came out however, many existing peripherals would not work with it because XP required new drivers and the vendors had little motivation to write drivers for old hardware.

The poor hardware support in Linux is drastically illustrated in an article by Fred Langa. He wrote about problems getting nine different Linux distributions (versions) to work correctly with two different sound cards (one real, one virtual) that all versions of Windows, even back to Windows 95, dealt with perfectly (see Linux’s Achilles’ Heel, Information Week magazine. April 19, 2004).

Hardware the OS runs on: Linux runs on many different hardware platforms, not so with Windows. For example, Windows NT used to run on MIPS CPUs until Microsoft changed their mind. It also used to run on Alpha CPUs, again, until Microsoft changed their mind. No one gets to change their mind with Linux. It runs on a very wide range of computers, from the lowest of the low to the highest of the high. The supported range of computers is all but stunning.

Because of its ability to run without a GUI, and thus need less hardware horsepower than Windows, Linux can run on very old personal computers such as 486 based machines. I took a Linux class where the server the students used was a 100 MHz Pentium. To get more mileage out of old hardware, Papa John’s converted 2,900 pizzerias to Linux. On the high end, Linux runs natively on IBM mainframes (the Z series) and on other high end IBM servers. eBay runs their web site on Linux as does Google. IBM’s family of “Blue Gene” supercomputers, used by Lawrence Livermore National Laboratory for nuclear weapons simulations, run Linux. NASA uses it on supercomputers that run space-shuttle simulations. On the small side, the iPodLinux Project created a version of Linux that runs on Apple iPods. NEC is working on Linux-based cell phones and Motorola is going to make Linux its primary operating system for smart cell phones. Debian Linux can run on on a computer the size of a deck of playing cards (100mm by 55mm) with an ARM cpu. In the home, Sony and Matsushita (parent company of Panasonic) will use Linux to build increasingly ’smart’ microwave ovens, TVs and other consumer gizmos. Likewise MontaVista Software will release a version of its embedded Linux for use in consumer electronics devices. Web site linuxdevices.com seems to track this.

Clustering: Linux has an edge here. It has been used to make enormous clusters of computers. In October 2002, ComputerWorld magazine said: “Linux clusters provide supercomputer-type performance at a fraction of the expense.” The same article reported that Merrill Lynch runs a cluster of 50 Linux computers. In March 2005, Forbes magazine reported that: “Linux now has become so technically powerful that it lays claim to a prestigious title–it runs more of the world’s top supercomputers than any other operating system.” (see Linux Rules Supercomputers by Daniel Lyons March 15, 2005).

Multiple Users: Linux is a multi-user system, Windows is not. That is, Windows is designed to be used by one person at a time. Databases running under Windows allow concurrent access by multiple users, but the Operating System itself is designed to deal with a single human being at a time. Linux, like all Unix variants, is designed to handle multiple concurrent users. Windows, of course, can run many programs concurrently, as can Linux. There is a multi-user version of Windows called Terminal Server but this is not the Windows pre-installed on personal computers.

Networking: They both do TCP/IP. Linux can do Windows networking, which means that a Linux computer can appear on a network of Windows computers and share its files and printers. Linux machines can participate on a Windows based network and vice versa. See Mixing Unix and Windows By Larry Seltzer (July 2002).

Hard disk partitions: Windows must boot from a primary partition. Linux can boot from either a primary partition or a logical partition inside an extended partition. Windows must boot from the first hard disk. Linux can boot from any hard disk in the computer.

Swap files: Windows uses a hidden file for its swap file. Typically this file resides in the same partition as the OS (advanced users can opt to put the file in another partition). Linux uses a dedicated partition for its swap file (advanced users can opt to implement the swap file as a file in the same partition as the OS).

File Systems: (updated September 2005) Windows uses FAT12, FAT16, FAT32 and/or NTFS with NTFS almost always being the best choice. The FATx file systems are older and have assorted limitations on file and partition size that make them problematical in the current environment. Linux also has a number of its own native file systems. The default file system for Linux used to be ext2, now it is typically ext3.

File systems can be either journaled or not. Non-journaled systems are subject to problems when stopped abruptly. All the FAT variants and ext2 are non-journaled. After a crash, they should be examined by their respective health check utilities (Scan Disk or Check Disk or fsck). In contrast, when a journaled file system is stopped abruptly, recovery is automatic at the next reboot. NTFS is journaled. Linux supports several journaled file systems: “ext3?, “reiserfs” and “jfs”.

All the file systems use directories and subdirectories. Windows separates directories with a back slash, Linux uses a normal forward slash. Windows file names are not case sensitive. Linux file names are. For example “abc” and “aBC” are different files in Linux, whereas in Windows it would refer to the same file. Case sensitivity has been a problem for this very web page, the name of which is “Linux.vs.Windows.html”. At times, people have tried to get to this page using “linux.vs.windows.html” (all lower case) which resulted in a Page Not Found error. Eventually, I created a new web page with the name in all lower case and this new page simply re-directs you to the real page, the one you are reading now (with a capital L and W).

As for crossing over, Linux can read/write FAT16 and FAT32. Some Linux distributions can read NTFS partitions, others can not. Captive-NTFS can be used to give Linux read/write access to NTFS partitions. I don’t know much about it, but it does not seem to come pre-installed. Also, Windows XP SP2 caused it problems. In September 2005, PC Magazine reviewed a $70 product called NTFS for Linux from the Paragon Software Group that gives Linux the ability to read/write to NTFS partitions. For information on Linux and NTFS, see the Linux NTFS Project.

On its own, Windows can not read partitions formatted with any Linux file system. However, Explore2fs by John Newbigin can be used to read Linux ext2 and ext3 partitions from Windows. Ext2Fsd claims to be an Ext2 File System Driver for Windows and there is another open source project for an Ext2 File System Driver for Windows. I have not tried these.

File Hierarchy: Windows and Linux use different concepts for their file hierarchy. Windows uses a volume-based file hierarchy, Linux uses a unified scheme. Windows uses letters of the alphabet to represent different devices and different hard disk partitions. Under Windows, you need to know what volume (C:, D:,…) a file resides on to select it, the file’s physical location is part of it’s name. In Linux all directories are attached to the root directory, which is identified by a forward-slash, “/”. For example, below are some second-level directories:

/bin/ —- system binaries, user programs with normal user permissions

/sbin — executables that need root permission

/data/ — a user defined directory

/dev/ —- system device tree

/etc/ —- system configuration

/home/ — users’ subdirectories

/home/{username} akin to the Windows My Documents folder

/tmp/ —- system temporary files

/usr/ —- applications software

/usr/bin – executables for programs with user permission

/var/ —- system variables

/lib — libraries needed for installed programs to run

Every device and hard disk partition is represented in the Linux file system as a subdirectory of the lone root directory. For example, the floppy disk drive in Linux might be /etc/floppy. The root directory lives in the root partition, but other directories (and the devices they represent) can reside anywhere. Removable devices and hard disk partitions other than the root are attached (i.e., “mounted”) to subdirectories in the directory tree. This is done either at system initialization or in response to a mount command.

There are no standards in Linux for which subdirectories are used for which devices. This contrasts with Windows where the A disk is always the floppy drive and the C disk is almost always the boot partition.

Hidden Files: Both support the concept of hidden files, which are files that, by default, are not shown to the user when listing files in a directory. Linux implements this with a filename that starts with a period. Windows tracks this as a file attribute in the file metadata (along with things like the last update date). In both OSs the user can over-ride the default behavior and force the system to list hidden files.

Case: Case sensitivity is the same with commands as with file names. When entering commands in a DOS/command window under any version of Windows, “dir” is the same as “DIR”. In Linux “dir” is a different command than “DIR”.

Modems: To save money, many internal modems do not include their own processor, instead they rely on the main cpu in the computer for their processing needs. These cheap modems are referred to as “winmodems” because they typically include low level calls to the Windows operating system that can not be replicated in Linux. For the most part, winmodems will not work under Linux. If you need to run a winmodem under Linux see Linmodems.org and www.linuxant.com/drivers.

Scripting: Windows started with BAT files (a combination of OS commands and optionally its own language) and then progressed to Windows Scripting Host (WSH) which supports two languages, JavaScript and VB Script. Windows also has WMI, the Windows Management Interface which makes many functions available to scripts.

Linux, like all Unix variants, provides multiple scripting languages, referred to as shell scripts. Among them are the Bourne shell, the C shell, Perl, Python and Ruby. I have used the first three, but not Python and Ruby. I find the Linux scripting languages cruder than WSH but much more powerful than BAT files. They tend to use special characters instead of English commands and don’t support objects (this only matters to programmers).

One scripting language that can run on both Linux and Windows is PHP. It always has to be installed under Windows, it may have to be installed under Linux. PHP is typically found running on Linux based web servers in combination with Apache, but it is capable of running “client side” (on your computer).

Printer Drivers: Every computer printer ships with drivers for last last few versions of Windows (at the time it was manufactured). Running the printer on a very old or too new version of Windows may or may not work. Still, this a far better situation than with Linux which does not support as many printers as Windows. In an environment with many Linux users, shared network printers a tech support staff, this should not be an issue as you can limit yourself to well supported printers. Home users of Linux however, will no doubt suffer from the relatively poor support for printers. Stephen Manes wrotein Forbes magazine (October 28, 2002) that “Linux printer drivers tend to handle the simplest of printing functions and leave out the rest”.

Help: Stephen Manes wrote in Forbes magazine (October 28, 2002) that Linux “Help screens often omit everything but the most rudimentary information and the most arcane.” For commands, both OSs offer help on the syntax and options of individual commands, providing you know the command. In Linux it is often done by appending –help or -h to the command. In Windows it’s often done by appending /? to the command.

Owing to it’s command line history, Linux also offers help via the man (short for manual) and info commands. The man documentation for the ls command, for example, is referred to as the man page for ls. From the Linux command line you can search the man database with “man -k somestring” which produces a list of manual pages containing “somestring”.

Windows offers help in the GUI interface for the GUI interface. On the Linux side, both KDE and Gnome also have built-in help about using themselves. I’m told that the Windows help is better for newcomers (nerds hate writing documentation, especially on an introductory level, and Linux is done by volunteers).

User Data: Windows allows programs to store user information (files and settings) anywhere. This makes it impossibly hard to backup user data files and settings and to switch to a new computer. In contrast, Linux stores all user data in the home directory making it much easier to migrate from an old computer to a new one. If home directories are segregated in their own partition, you can even upgrade from one version of Linux to another without having to migrate user data and settings.

Shutting Down: Both have to be told to shut down and for the same reason, to quiesce in-flight activity in an orderly manner. You shut down Windows thru the Start button, then select Shutdown. In both the KDE and Gnome GUIs for Linux, you shut the system down by first logging out (equivalent to logging off in Windows). In Gnome, you select the Halt option, in KDE, the shutdown option. Linux can also be shut down from a command prompt using the shutdown command which can either shut the system down immediately or be told to shut it down at some time in the future. Windows XP also has a shutdown command, earlier versions of Windows did not.

The scheduler built into Windows XP is not able to schedule programs to run at shutdown time. I find this annoying, because its an excellent time to schedule backups. Linux can schedule programs to run automatically at shutdown time. I haven’t tried this but a quick glance at How Linux boots makes it seem complicated. I’m told to follow the instructions in this article but replace runlevel 5 with runlevel zero, which represents shutdown time.

Choosing Linux vs. Windows: On the Personal Computer show in December 2003 John C. Dvorak predicted a bright future for Linux. His main points being: it’s free, the applications are getting more mainstream, Open Office is a “fabulous” product, the GUI is pretty much like Windows, it’s high quality, bullet proof and resistant to the thousands of Windows viruses and worms. If he owned a company with thousands of PCs, he would put everyone on Linux.

To date the only organizations (that I’ve heard about) using hundreds or thousands of Linux based computers are government agencies in countries all over the world. They may be driven by cost and/or security concerns (practically speaking there are no Linux viruses). Some countries also may not like being beholden to a U.S. based company for so much of their software.

See More Balls Through Windows from The Economist April 20, 2004 on cfo.com. Is Microsoft finally about to face real competition in desktop-computer software? This article raises a point I agree with. The next version of Windows, the one that will replace XP, is not going to be delivered for a long time opening up a window of opportunity for Linux. In addition, the upcoming version of Windows is likely to be expensive and require new hardware, two other areas where Linux competes well. It may also involve too many changes. By the time it’s delivered, more and more Linux distributions will look more like Windows, just as Microsoft rolls out a new user interface. People accustomed to the current Windows UI may resist the change.

However, Microsoft fights the spread of Linux at all costs. For example, when Thailand was going to make government subsidized Linux based computers available throughout the country, Microsoft cut a deal with the Thai Information, Communication and Technology Ministry. People in Thailand can buy Windows XP and Office XP (without Outlook) pre-installed, activated, and ready to run for the equivalent of $37 US dollars. See Microsoft offers cut-rate Windows from CNET News.com February 9, 2004. Then again, the governments of China, Japan and South Korea are teaming up to create their own Asian flavor of Linux.

My 2 Cents

The above are facts, to the best of my knowledge, not opinion. This is opinion.

Will Linux spread to the point of becoming a serious competitor to Windows on the desktop? No. And this has nothing to do with which is better, no matter how you judge “better”. Consider the keyboard in front of you. The key arrangement was made long ago when keeping the metal wires connected to the keys from hitting each other was the big consideration. Now that that no longer applies, does anybody switch over to a keyboard with a better design? No. All of us who know how to type, are used to the current arrangement of keys. Switching to a new layout would be a major disruption and thus have to pay back in a major way. I don’t think desktop versions of Linux will pay back enough to encourage people to switch. My Linux experience is not extensive, but coming to it with a Windows background, as most people will, it is often frustrating to figure out how to do ordinary everyday things. Servers are another story. So too are call centers and other single-use environments where Linux makes a lot of sense.

Related Links

Newer links are at the bottom In November 2002, a security problem in a Microsoft FTP server resulted in the exposure of internal Microsoft documents. Among these was a white paper discussing the approach they used to convert Hotmail from UNIX to Windows 2000. The white paper includes these topics: Advantages of UNIX, Problems of Windows, Strengths of Windows.

Linux vs. Windows debate / discussion at DesktopLinux.com.

An article on the Wine project from PC World Magazine, February 2002.

An article on Lindows, including comments on Wine, by Fred Langa in Information Week, December 2002.

Can I run Windows software on UNIX? AnandTech. July 14th, 2002

Linux runs Office XP, but not Outlook. April 23, 2003. Cross Over Office (about $55) by CodeWeavers enables Linux users to run Microsoft Office XP, but it is not possible to use Outlook XP or Access XP. Outlook 2000 and Access 2000 are supported under Linux.

A Beginner’s Guide to Linux by David Pogue in the New York Times. December 5, 2002

Windows is cheaper than Linux By Oliver Rist of ZDNet. December 9, 2002. The article says the cost of owning a server OS is mostly due to the man-hours spent on maintenance. For basic file and print serving, Windows is plug-and-play, meaning less time spent tweaking configurations. The author says that Linux servers take longer to configure and troubleshoot than Windows servers but notes that once this is done, Linux requires significantly less troubleshooting than Windows. However, this only applies to standard Linux installations. Any modifications though, and you are own your own. As to security, the author says that only someone smart, and therefore expensive, can manage Linux security. In contrast, Windows 2000 automatic security updates can be overseen by anyone. Often Windows security patches cause something else to drop dead, a huge waste of time, effort and expense. The author has almost never seen this happen with Linux.

Linux continues desktop march. ZDNet (UK). January 3, 2003. It was predicted that Linux will surpass the Mac OS as the number two desktop operating system. However, its growth is being held back by a lack of well-established applications. Despite the fact that there are Linux programs that compete reasonably well with Microsoft Office and Outlook, the article quotes someone as saying that companies often prefer to go with the most popular software. The fuzzy situation with technical support when running Microsoft Office directly under Linux (via Crossover Office) is also considered a hindrance. On the other hand, major Linux vendors are just starting to pay attention to the desktop market. HP will support desktops with Mandrake Linux and the governments of assorted countries are interested in open source software.Friend or foe? April 10, 2003. The Economist magazine. The rise of Linux is dividing the computer industry into winners and losers. Linux is changing the dynamics of the computer business. Some of the industry’s titans benefit from its advance, while others lose.

Why You Won’t Be Getting A Linux PC by Lisa DiCarlo, Forbes magazine. June 17,2003

The Limitations Of Linux by Lisa DiCarlo, Forbes magazine. June 16, 2003The November 25, 2003 issue of PC Magazine has an article comparing Linux and Windows. Michael Miller said “The complexity of Linux has kept people away, but that’s all changing. Many current distributions look a lot like Windows”. The article is If You Don’t Do Windows Linux developers have polished their desktops, and those at Apple have created their richest OS to date, making both OSs more appealing than ever to the mainstream PC owner.

Crossover Office 2.1 Runs MS Windows Software on GNU-Linux Systems An Implementation of MS Windows APIs for Linux providing a Linux-based, Windows-software compatible environment by Mike Angelo November 10, 2003Learn about Linux from the Linux Documentation Project. Much of it is for techies, but it includes Introduction to Linux – A Hands on GuideUnix for Windows By Tom Yager in InfoWorld January 23, 2004. About running Windows applications on Linux and vice versa.Linux and Windows Many articles about making the two Operating Systems work together.

Linux Loyalists Leery Forbes March 31, 2004. About the selling of Linux. How cheap does it need to be to compete with Windows?

Plugging the Linux holes April 30, 2004 CNET News.com The lack of some familiar applications is slowing Linux adoption. Still another way to run Windows programs under Linux? It’s called David from a company called SpecOps. Too early to tell if this is real or a hoax. David is not a Virtual Machine product, it is “system service”. Install David, and then install your Windows applications (or so they say).

Finding the missing link in Linux-Windows compatibility April 25, 2004 By Erwin Lemuel G. Oliva on INQ7.net

SpecOpS Labs Analysis: Is it real? on Linuxelectrons.com

Linux vs Windows: Another Great OS Leap Forward On the Way? April 21, 2004 on LinuxWorld.com

Why Windows Beats Linux by Scot Finnie in PC Today May 2004

CrossOver Office Professional 3.0.1 reviewed by Jim Lynch at ExtremeTech.com June 30, 2004. Included is a list of the Windows applications that it can run under Linux.

Desktop Linux a vehicle for pirating Windows CNET News.com September 29, 2004

Comparing Windows with Linux and UNIX from Steve Ballmer, head honcho at Microsoft. Consider the source. October 27, 2004

Linux/Unix vs. Windows Hosting from SiteMount a web site hosting company. Added July 2005.

Is Linux For Losers? by Daniel Lyons in forbes.com. June 16, 2005. Linux is not the only free clone/copy/edition of Unix, there are three open source versions of BSD Unix (FreeBSD, NetBSD and OpenBSD). In this article, the lead developer of OpenBSD brags about how much better it is than Linux.

Windows vs. Linux/OSS today, part 2 by Jan Stafford SearchEnterpriseLinux.com July 27, 2005

Linux on a ThinkPad which models are certified for which Linux distributions. Last modified December 1, 2005.

Desktop Linux versus Windows XP shootout by George Ou December 20, 2005. A ZDnet blog.

This article has been taken from http://www.michaelhorowitz.com/Linux.vs.Windows.html