Best Free Internet Safety Check - Supplement

Introduction

This supplement provides useful information on the main article Best Free Internet Safety Check category:

 

Website Raters

Installed website raters

The recommended website raters are:

With one exception, there is a significant gap between the recommended products and the next set of raters:

  • The exception is Trend Micro Web Protection Add-on. It is a time-limited trial so it is our policy not to recommend it. Even so it is worth installing to see how these sorts of products work. There are reports that it continues to work after the trial period has ended but I haven't used it long enough time to confirm this. Why is it so good? In my tests it complemented WOT - between them they only failed to warn me about one bad website. It runs with all the browsers, it identifies a high proportion of dangerous websites, it is simple and easy to use, it blocks dangerous sites and it doesn't let you bypass the block screen even if you request a review of the blocked site. All you can do is lower your chosen level of security: low, medium, or high. The high setting blocks many ads so I'd disable the pop-up warning.
  • BitDefender Traffic Light beta is the outcome of an ambitious project to develop an internet safety check that runs on all the main operating systems with any browser ... all for free. When it is finally working properly it is likely to be recommended because it is multi-platform, has near universal web browser compatibility (at least, under Windows), and uses a simple unobtrusive interface. Users of Firefox, Chrome and Safari also have an option to install TrafficLight add-ons which are limited in features but provide similar protection. I spent far too much time trying to test this but at least that might save you from the same problems. The downsides are very significant: threat detection is only average and the beta version can be unstable - I've had to reinstall it several times. All the features are not fully functioning, e.g. the ad blocker has not been added, and there are known issues with other software such as AVG LinkScanner. Last but not least, the download installer is small but the filtering database is a third of a gigabyte (GB). That's why I suggest that you wait for the completed product.
  • Norton Safe Web is directly comparable with McAfee SiteAdvisor: solid but unexceptional. Both disappointed me because they are not very good at warning of bad websites. Safe Web runs using information from Norton Threat Explorer so it uses the same categories of threats. It also lists trusted third-party sources on site reports. There are three easy to understand categories for issues with websites: computer threats, identity threats, or annoyance factors. Safe Web supports searching with Google, Yahoo! and Ask but not Bing or AOL. You can use it with an existing Norton online account although, in that respect, it could be better integrated. I found the user reviews to be much weaker than WOT because there are no rating categories. You cannot, for example, give a  website a bad comment for excessive ad pop-ups and a good comment for being informative. I saw websites that were being flamed by disgruntled users but, unless you are a website owner, you cannot request a rating review.
  • PC Tools Browser Defender is a another option for 64-bit Windows. It presents little information unless it discovers a problem. It didn't identify enough threats to be recommended here but it usefully scans downloads. It rates searches on Google, Yahoo! and Bing. Like many of these products your support options are limited. There is no online manual and no category for it in the online knowledgebase.
  • G Data Cloud Security adds an icon to the toolbar for Firefox and Internet Explorer. It blocks known malware and phishing websites and works very well for when it does find a website in its database. Overall, too many problem websites were not blocked.
  • Web Security Guard has a database with  less than 150,000 websites. Although it does have a much higher proportion of sites to avoid, 20-30%, that is way short of 10% of WOT's 30 million websites. During my testing it did source blacklists from Google Safe Browsing and Stopbadware.com but for some reason it doesn't perform to the same level as those sources. Another negative is that it installs the less desirable Crawler toolbar which gives access to the Crawler meta-search. On the positive side, it has many useful features that should be in other products: the site report provides links to their third-party sources. There are clear indicators as to which categories the website falls into: Advertising networks; Adware, spyware, viruses; Phishing or scams site; Popup flood; Pornography; Spam; Warez or Ilegal content. It is the only product that allows you to decide the level of danger for each category rather than altering the warning level. This means that if you want to see advertisements or pornography then you can set them to "Safe".
  • BrightCloud provides a simple toolbar (Firefox add-on and a Chrome add-on for Windows XP or later) that displays the reputation rating (0 to 100%) for the current website. The BrightCloud reputation rating system gave me very good estimates of risk but I had to visit the risky site first.
  • Webutation adds an icon to the toolbar and it too only works for pages you browse. It has a Firefox add-on and a Chrome add-on for Windows XP or later. It uses trusted sources like WOT and Google and says that it does real-time scanning. The problem is that it either did not identify known problems or did not score the site low enough to ensure that users would not visit it.

The following raters were not reviewed:

  • Sitehound is being rewritten. I expect a lot from it because It is from the makers of MailWasher, a reputable spam filter for email.
  • Google Toolbar will be reviewed in a later update.

 

Online website raters

Online rating checks are provided for all of the installed raters. I've listed these in the approximate order that I would refer to them so check them out and see what you prefer.

 

Website Scanners

Installed website scanners

 The list of recommended website scanners to install is rather limited:

The other products didn't provide much more protection than the built-in browser safety measures::

  • BitDefender Anti-Phishing just didn't do enough when there are products covering more safety issues. It is compatible with Firefox (but not 3.6.16 or later) and Internet Explorer. Registration is mandatory to run but it took too long to send me an email to activate it. It had to be manually uninstalled.
  • M86 Secure Browsing. M86 bought Finjan in 2010 and appear to have left it largely unchanged. It is simple to use when browsing and its interface is similar to other products like LinkScanner. But it only works with web searchs and some social networking sites. An icon appears in search lists and there are warning pop-ups. It regularly shows as safe websites that have warnings from other raters and scanners. It is also slower than the better products and only works with Internet Explorer and Firefox (beta version).

I curtailed testing of these scanners for various reasons:

  • Dr. Web LinkChecker didn't find any problems during initial testing and wouldn't scan large websites. It was slow and opens a new window for each check. It has add-ons for Firefox and Internet Explorer plus a manual registry install for Opera including editing the .ini file.
  • The NetCraft Toolbar for Firefox and Internet Explorer performed no better than the browsers.
  • Trend Micro Browser Guard 2011 will be reviewed in a later update. It is another beta and only works with Internet Explorer. Like BitDefender TrafficLight, it has much potential. If it works as intended then it mainly protects against the newest threats. It is unobtrusive with only an icon in the system tray.
  • QualysGuard Malware Detection beta will also be reviewed in a later update.
  • Comodo Verification engine will be reviewed in a later update.
  • CallerID will be reviewed in a later update.

 

Online website scanners

Like the online rating checks, I have listed these online scans according to my preferences. Try them out and see what you prefer.

 

Discussion

Issues with the Editor's Choice: Web of Trust (WOT). I have found no reason to alter the previous Editor's Choice. Web of Trust (WOT), is also the Internet Safety Check that we use here at Gizmo's Freeware. Although there will continue to be controversy, WOT has strengthened its position in the year since the previous update. For example, WOT is the only product which has user participation as an essential input. Even so, I was surprised to see that WOT is getting user ratings at forty times the rate of the next best product.

Product Total Ratings Ratings This Week
WOT Community Statistics 32,000,000 rated sites
13,000,000 comments
230,000
McAfee Reviewer Central 6,000,000 ratings 6,000

What are some of the complaints about WOT?

  • "WOT users give their competitors or enemies bad ratings."
    True, this is its biggest weakness particularly if you are an aggrieved site owner. The problem is that all user rating systems can be subverted like this. WOT clearly provides the best system for viewing and managing user ratings. For example, a recent update allows site owners to respond directly to comments on the scorecard.
    WOT also calculates a hidden reliability score for each user. This determines the weight your ratings carry. I've just started rating so my views don't carry much weight at all. The reliability score is not published but a user's activity score is.
  • "WOT Seal of Trust is a scam or protection racket"
    Several reputation services sell seals or certificates that can be used by website owners to vouch for their reputation. Some people take issue with paying for such services and think that it means you can simply pay to get a good reputation. In practice, these products have proactive safety checks to confirm the website owner and to regularly check the website for security issues. So they provide a further level of security for users and website owners.
  • "WOT is censoring."
    Blocking has to be turned on for this to occur and the default is no blocking. The child safety rating can be excluded from the  "optimized" and "lowest rating" options.
    I know that Firefox add-ons are meant to conform to private browsing (leaving no record of your browsing session) but WOT is the only rater that allows me to decide whether I want rating to be on or off while browsing in that mode.
  • "WOT has a lot of user ratings about morality and personal preference."
    True, in areas like pornography, religion, and gambling.  But WOT has appropriate categories to make these views visible and rating reviews can be requested.
    If you don't provide categories for ethical and moral considerations then people will just use another category. I noticed a good example of this sort of behavior while I was testing these products. The free memberships at OpenDNS do not include malware filtering nor do they allow sites to be categorized as malware. That doesn't stop users who will use the first negative category, "Adware", as a catch-all category.

Website raters are similar to parental filters. They often share the same security databases. For example, OpenDNS provides filtering and a website called PhishTank which is used by several website raters to check for phishing websites - those that attempt to defraud by getting you to provide your personal information. Parental filters primarily aim to block access for all web browsers whereas website raters are usually browser add-ons so they are limited to specific web-browsers. What parental filters do better is to categorize the type of website (which is used to block entire categories e.g. gambling websites)  to enforce safe-searching (including image searches), and to better counter the temptation to bypass warnings. They just don't give me any help with a website link until I click on it. While the lines between the categories are blurring - some website raters now provide blocking and new raters are appearing that are not browser add-ons - I don't expect the categories to converge.

Website scanners are basically anti-virus software. So most come from anti-virus vendors and related Internet safety organizations. Just do not use them to replace your anti-virus software which runs full-time on your computer. Continuous monitoring of your system is better than one-off scans using the more limited command-line versions. Be aware that some of the software installs scanning databases which have to be updated just like a normal anti-virus program. I preferred the cloud-based products that didn't install a database unlike the TrafficLight beta which took 325MB of disk space when most are under 30MB.

User rating and comments can be very useful. Look at the extensive categories that are provided for user feedback across all the products I tested. It is true that all but three (in italic) of the sixteen categories are used in WOT which is one reason why it's user ratings are so powerful. I can rate one website in four categories and then leave comments in up to three different categories. I could comment on this site as "Good", "Informative" and "Useful" giving a more detailed comment in each of those categories. Another site I might comment as "Entertaining", "Adult content", and "Ethical issues".

Good Bad Other
Child friendly
Entertaining
Good
Good customer experience
Useful
Adult content
Advertising network
Annoying ads or pop-ups
Bot nets
Browser exploit
Ethical issues
Hateful or questionable content
Malicious content, viruses
Phishing or other scams
Proxy avoiders & anonymizers
Spam
Spyware or adware
Useless
Other

 

Safety checks can slow your browsing. Be warned that browsing will usually be slower if you use a dial-up connection or you run several products at the same time. At one point I had 10 add-ons enabled so web search pages took 15 to 40 seconds longer to completely load, the browsers took longer to do anything, and too many toolbars and icons obscured my view.

Products can be incompatible. They can be incompatible with each other or with your existing security products. This is more likely where your scanner is competing with other security software. As many of these products are add-ons you can have problems with other add-ons already installed for your web browser. The one problem that annoyed me was with Firefox on my test system. At one point I lost the ability to view toolbars or make any changes to the add-ons. In the past I have laughed when I've seen people with many Firefox add-ons having the same sort of problem. So you're welcome to laught at me now. The problem is not pretty and none of the suggestions fixed the problem. So I nuked (completely uninstalled everything to do with) Firefox and reinstalled it. Problem solved but then I had the time-consuming task of reinstalling all the add-ons again.

Don't bypass block screens. There is one particularly dangerous gotcha in some of these products. If you bypass a block screen by adding the website to a safe list, you will find that there is no feature to later remove that bypass. These are simple products without some of the features that you do find in parental filters.

Web browsers have protection too. The effectiveness of built-in safety checks varies - Firefox, Chrome and Safari use Google Safe Browsing, Opera uses AVG, and Internet Explorer uses Microsoft - yet they are all in the middle third of the products reviewed here.

The installed safety checks differ in their support for different browsers. Firefox (18 products work with it) and Internet Explorer (17 programs) are the best supported browsers although Firefox gets the best features overall. For example, WOT only provides all its features in Firefox. Further back are Chrome (8), Opera (3), and Safari (2 programs). This is not as bad as it looks because the best products support all the main web browsers.

Why are these products free? The vendors can sell upgrades to paid versions and security suites with better protection but there are a few other reasons why you help them when you use their products. First, they get more input on safety issues when they have their software on more computers. Secondly, they can sell their seal or symbol of trust and good reputation to website owners. The more users that rely on these products then the greater the momentum behind these seals of approval. Finally, users will put up with more problems in beta software when they don't have to pay for the final version.

How I tested the detection of threats. I tested all the products by sampling 43 websites what were newly identified as dangerous and added a further 5 long-term offenders. I selected websites from various security databases. The breakdown was 35 malware, 8 phishing and 5 spam websites and included 12 downloads. In general, the sites had more than one type of problem. I emphasised those sites that appear quickly, do their damage, get discovered, and then disappear just as quickly. Therefore I was not surprised to have hacked websites cleaned up and malicious websites go off-line during my testing. The main test output is the proportion of websites identified with known active threats (31 web pages and downloads that had threats confirmed by scans).

A summary of threat detection tests. My reviews and tests have led me to draw some tentative conclusions but left me with some questions I can't answer. More testing could settle most of these issues except that I don't have the time to do this.

  • Why aren't these checks more effective? These products are not as good as I expected. The best only identified two-thirds of the active threats. The worst didn't find any. On average, I found the following:
    • Raters (detect 27% of websites with active threats) are much more effective than Scanners (16%).
    • Web Browsers, which provide security in the same way as Raters, are also better than Scanners.
    • Meta-raters and Meta-scanners are better than Raters and Scanners respectively. The best Meta-scanner detected another 50% more than the best Scanner.
  • Overall, there was no significant difference between Installed Software and Online Services. But comparing the categories some significant differences are clear:
    • On average, Installed Raters performed better than Online Raters. This is because Installed Raters tend to come from vendors of security software while many Online Raters do not. They are more likely to be from small voluntary organizations or to be a specialized service. For example, ISECLab.org has several specialized services including FIRE, ANUBIS, and Wepawet.
    • Online Meta-Raters were nearly twice as effective as Installed Meta-Raters because they run a more comprehensive set of scans. As a result, they are less convenient and do take much longer to provide a result.
  • Why is there variation in performance for products from the same vendor particularly where I'd expect the ratings to come from the same database? Where an organization publishes both an installed product and an online service, the online service will often perform worse. But where the installed product has paid upgrades the online service often performs better. This suggests to me that free installed products are restricted in some way to make the paid versions more attractive.
Type of Product Best Average
Meta-rater 65% 43%
Rater 61% 27%
Meta-scanner 55% 49%
Browser 43% 34%
Scanner 34% 16%

Meta-raters and meta-scanners are not always better. Using the best safety checks will not remove the need to use your own good sense. Single reports of a problem often don't affect the rating enough to warn users. The meta-rater's quandary is like our own: whose report do we believe. If we knew which was best we could ignore the others. Choosing to use the lowest rating has the risk of flagging a good website as bad, i.e. a false negative. Choosing the highest rating has greater risks for us because it could mean we visit a bad website thinking it was safe, i.e. a false positive. To get the best out of these products you should ignore any meta-score and assess the risk by looking at the individual scores. This is where the online services excel because they use the most comprehensive set of safety checks.

 

Editor

This software category is in need of an editor. If you would like to give something back to the freeware community by taking it over, check out this page for more details. You can then contact us from that page or by clicking here

Back to the top of the article.

 

Share this
4.916665
Average: 4.9 (12 votes)
Your rating: None

Gizmo's Freeware is Recruiting!

Gizmos Needs YouShare your knowledge of free software with millions of Gizmo's readers by joining our editing team.  Details here.