Free antivirus comparison

Introduction

This is a comparison of free antivirus applications. These are all the programs I could dig up on wikipedia, google, and download.com. These tests were all performed within a VirtualBox virtual machine, which has 256 MB of RAM. Unlike my other comparisons, I moved the final score up so that the baseline was 0. These tests were performed on each antivirus application:

time

Time the malware scan takes, in minutes. These will be longer than in normal conditions because the virtual machine is slow, and it has some disk writing quirks. Fastest two get one point, slowest one gets -1 point.

number of items detected

The number of malware files detected. The malware folder is about 12GB large. About 4/10ths of this file is old malware the scanner really should detect. About 1/5th was (painstakingly) collected from the wild. It also includes all the malware in various archive formats, which make up the other half of the file. The program with the least detected gets -2 points, the next best gets -1 points, and so on; the best program gets 4 points.

RAM in use

RAM while the AV program is scanning, in KB. Best gets 1pt, worst -1pt

RAM at rest

RAM while the AV program is protecting the computer, but not scanning. This will not be counted if the program doesn’t include a virus guard. Best gets 1pt, worst -1pt

CPU in use

Percent of used CPU while the AV program is scanning. This will seem larger than normal because the scans are in a virtual machine with very limited CPU. Best gets 1pt, worst -1pt.

number of false positive detected

The number of safe files detected as malware. My false positive folder is about 1GB large. It contains a number of popular programs, and some little-known security programs that I chose because they perform functions similar to malware. If it detects one of these safe files, it gets -1 point.

virus guard

If the program has some sort of virus guard(it checks files as they are written). -2 if it doesn’t, no change if it does.

Clean uninstall

If there are no remnants of the program after I uninstall it and restart. This is important because remnants of an antivirus application can cause problems in the future. -1 point if there are any remnants, no change otherwise

Data

	tests
AV programs	test time	number of malware items detected	number of false positives
AVG	87	11,873 (0 pt)	0
PC tools	86	11,405 (-2 pt)	1
bitdefender	45	20,166 (3 pt)	0
COMODO	34	13,410 (1 pt)	0
clamwin	34	13,629 (2 pt)	0
avast	13	11,828 (-1 pt)	0
antivir	172	37,491 (4 pt)	1

 

	tests
AV programs	RAM while scanning	RAM while idle	CPU while scanning
AVG	50,000	1,348	90
PC tools	25,000	500	85
bitdefender	36,000	N/A	80
COMODO	27,000	20,000	60
clamwin	35,000	N/A	70
avast	21,000	8,000	80
antivir	50,000	5,340	80

 

	info
AV programs	clean uninstall	virus guard	notes	points
AVG	Yes	Claims to, but in my tests it didn’t	I didn’t like the interface. Not enough options. Very slow scan, average detection. It automatically deletes files without prompting the user. I’ll give it the point for virus guard, because it claims to have one, but I couldn’t get it to trigger.	0
PC tools	Yes	Yes	Stylish and easy-to-use interface. Very slow scan, bad detection.	0
bitdefender	No	No	Professional looking, but it took me a few seconds to figure out how to scan. Very nice speed and detection. Automatically deletes infected files without asking. A bit bulky. It comes with a Pro version trial which has a virus guard, but that is disabled after 15 days.	2
COMODO	No	Yes	Simple, not many options. Basic HIPS protection I don’t like much. Very fast, with good detection.	3
clamwin	Yes	No	Plain, the GUI isn’t very user friendly. Many options. Speedy with good detection. Some programs like winpooch provide a virus guard using clamwin.	3
avast	No	Yes, very good	Nice interface, mostly easy to use; some options are hidden deeply. Bulky. Avast has protection features unique to antivirus applications, and it acts almost like a firewall. Fastest, detection isn’t great.	2
antivir	Yes	Yes	I like the clean interface. Not enough options, though. Slowest scans, but the best detection.	3

Summary

These results show COMODO, Clamwin, and Antivir are the best free antivirus applications. Since these ties aren’t very satisfying, here is my opinion on the programs, best to worst:

1. Antivir, for its amazing detection
2. Clamwin, because its an open-source project. Use a program like winpooch with this.
3. COMODO, an all-around good antivirus
4. Avast, which has a fantastic virus guard.
5. Bitdefender, because of its great detection. Unfortunately, you have to pay for a virus guard.
6. AVG, which has better detection than PC tools
7. PC tools, which isn’t very good at all

You should be very secure with Avast, COMODO, Clamwin(with winpooch), or Antivir. I wouldn’t use the others.

I hope you’ve enjoyed this comparison, tell your friends! If you find any errors, please comment. You can link here using antivirus.theymos.com.

Link to my comparisons!

You can now easily link to my comparisons with:
firewall.theymos.com
browser.theymos.com
Please tell your friends; I want to have at least 20 readers before I continue my comparisons.

Vote for my next comparison!

You can vote on what I compare next at this poll

My malware test file is small, with only about 1000 malware files. It might be better to use other comparisons, from companies with larger test files

I’ll start the chosen comparison once I get 20 votes for a particular item. If you’d like to see another comparison, include it in the comments and I’ll add it to the next poll

theymos.com

You might like my new project, theymos.com – providing a free web page for everyone. If you’re skilled enough to make a website, and don’t need our help, you can volunteer to help others!

(I’m going to put links to sites I like to give them better pagerank, they aren’t relevant)
Organize productions

Free Browser Comparison

Introduction

This is a comparison of free browsers. These are all the windows browsers at this wikipedia comparison of browsers that an average user can use. The computer running these tests has 2GB of RAM and 3.4GHz CPU. Multitrack stopwatch was used for all timed tests. These tests were preformed on each browser:

Start time

This includes:

• Cold start, how long it takes for the browser to start after the computer is restarted
• Warm start, how long it takes for the browser to start when the browser has already been run and then closed
• Hot start, how long for another window of the browser to open

Each test is separate. The browser with the lowest time gets 1 point, the one with the highest gets -1.
I invented the term “hot start”

Acid2 test

The acid2 test will test the browsers adherence to standards of HTML and CSS. I measure it on a 1 to 5 scale:

• 1:you can’t tell it’s a face
• 2:you can tell it’s a face
• 3:you can tell it’s a smiling face
• 4:it looks how it should except for scroll bars or a few missing pixels
• 5:it is exactly how it should be

Browsers with 5 get 1 point, Browsers with 1 get -1

Javascript speed test

This is how fast the browser completes this javascript speed test The browser with the lowest time gets 1 point, the one with the highest gets -1.

Table speed test

This is the time it takes the browser to completely show a 27MB table. The browser with the lowest time gets 1 point, the one with the highest gets -1. Some browsers use incremental reflow, where it loads the page as it is downloaded, making it look faster than it is; it will be noted if that is the case.

Memory/CPU test

This measures memory and CPU usage on a blank page, 10 tabs(if applicable) with google in them, and at this GIF-filled page(more than 100 GIFs). The browser with the lowest time gets 1 point, the one with the highest gets -1.

Browser security

This tests the browser’s security, using default settings. I used the Scanit security test. None of the browsers had any vulnerabilities, so points wont change. After it was apparent that none of the browsers would fail the test, I tested the major 3 browsers on 2 more tests; I could not get them to fail.

Other information

I will tell you if a browser is open source, what it’s layout engine is, additional features, and my personal opinion of the browser. None of these things will affect the final score.

Browsers are color-coded by layout engine:

• green:gecko
• blue:trident
• white:presto
• grey:text-only or specialized layout engine. These won’t be compared to the other browsers

Data

	tests
browsers	cold start	warm start	hot start	acid2	javascript speed	table speed
amaya	4.0	2.5	1.0	1	N/A	N/A
AOL	3.8	1.1	0	1	162	110(but it froze shortly afterwards)
Avant	1.6	1.0	.8	1	166	45
ELinks	0	0	0	N/A	N/A	N/A
K-meleon	.6	.6	0	2	172	678, incremental reflow
Lynx	0	0	0	N/A	N/A	N/A
Maxthon	1.9	1.6	1.0	1	164	46
firefox	1.5	1.0	0	2	182, but there was an error on part of the test	608, incremental reflow
Netscape(new)	7	3	0	2	160	783, incremental reflow
off by one	0	0	0	N/A	N/A	more than 15 minutes(when I stopped it)
Opera	.8	.5	0	5	212	36, incremental reflow
seamonkey	1.2	1.0	0	2	169	620, incremental reflow
Internet Explorer 7	1	0	0	1	185	53
flock	2.4	2.1	0	2	192	725, incremental reflow

	tests
browser	RAM blank page	RAM 10 tabs	RAM GIF page	CPU blank page	CPU 10 tabs	CPU GIF page	Security
amaya	44,344	69,820	N/A(doesn’t animate GIFs)	0	4	N/A	N/A
AOL	33,736	35,000	45,000	2	6	9	0
Avant	42,876	43,800	100,000	0	8	50	0
ELinks	2,256	2,256	N/A	0	0	N/A	N/A
K-meleon	28,000	28,752	161,000	0	0	7	0
Lynx	2,256	N/A	N/A	0	N/A	N/A	N/A
Maxthon	22,468	24,628	67,000	1	1	5	0
firefox	24,620	26,840	94,000	0	1	4	0
Netscape(new)	53,600	54,784	104,900	0	1	7	0
off by one	8,120	N/A	30,000	0	N/A	2	0
Opera	17,604	21,920	27,080	0	0	0	0
seamonkey	29,336	29,928	55,000	0	0	4	0
Internet Explorer 7	40,410	50,160	85,620	0	0	7	0
flock	32,552	33,752	72,000	1	1	3	0

	info
browser	open source?	layout engine	features	notes	points
amaya	yes	specialized	WYSIWYG editor	This is a web page editor built by the W3C. It can function as a limited browser, but it doesn’t support javascript and it thinks you’re trying to edit the page instead of navigate it.	0
AOL	no	trident	AOL integration RSS reader	AOL explorer is a good looking internet explorer adaptation that supports integration with your AOL account.	-2
Avant	yes	trident	online account storage customizable look autofill RSS reader gestures	Avant is a feature-packed but bulky IE adaptation. The “features” get kind of annoying; I had to restart it 4 times before it would start without asking me something.	-2
ELinks	yes	text-only	tabs	Elinks is an easy-to-use text-only browser. It displays tabs along the bottom of the text window.	0
K-meleon	yes	gecko		K-Meleon is a firefox adaptation, with a few trivial built-in extras. The biggest change is that it is slightly faster.	3
Lynx	yes	text-based		Lynx is the text-based browser. It’s absolutely incompatible with any new web technologies.	0
Maxthon	no	trident(by default, can also use gecko)	customizable	Maxthon is a bulky adaptation of IE with more customization.	-2
firefox	yes	gecko	many ways to customize	Firefox is slow and can get pretty bulky, but it makes up for it with it’s extensions, which can greatly expand it’s capabilities. Firefox doesn’t have many built-in features, and someone who expects it to be perfect “out of the box” will be disappointed. Out of the three major browsers, firefox is the most customizable.	2
Netscape(new)	no	gecko(can switch to trident)	integrated security AOL integration autoswitches to trident layout engine	The netscape browser is firefox version of the AOL browser. Security center is very integrated, and it autoswitches to the trident layout engine if it detects the site needs that.	0
off by one	no	unique(minimalistic)		Off by one is a minimalistic browser that only displays text and images. You’d think it would be fast, but it’s slower than normal browsers. It has nice memory usage, but I still wouldn’t use it.	0
Opera	no	presto	bittorrent gestures widgets	Opera is the speediest browser with regular pages, beating trident-based browsers, and it includes incremental reflow. While not as powerful has firefox’s extensions, opera allows you to customize with widgets. It is also perfectly compliant with standards. The only downside to opera is it’s slow javascript time	8
seamonkey	yes	gecko	mail IRC RSS reader HTML editing	This is a suite of applications, including email, IRC, and an HTML editor. The browser doesn’t really stand out.	3
Internet Explorer 7	no	trident	tabs phishing filter	Internet explorer 7, while an improvement over 6, doesn’t stand out. It was just updated to include the things people expect from a browser nowadays.	4
flock	yes	gecko	linking to various online accounts	Flock is basically firefox with built-in extensions.	0

Summary

It’s clear from these results that Opera was the best browser, followed by Internet explorer 7. Then came the gecko-based browsers, and finally the other trident-based browsers. Which browser you choose depends on your taste, if you must have extreme customization, go with a gecko-based browser like k-meleon or firefox. If you’re looking for the absolute fastest, smallest browser, go with opera. IE 7 is a fine, slim browser, but it’s speed doesn’t stand out. Other than IE 7, you should avoid trident-based browsers. I currently use firefox, but I’m not at all impressed with it’s speed or resource usage. The only reason I keep using it is that I’m so used to my extensions; it would be a major change to live without them. If firefox 3 isn’t a major improvement, I will seriously consider switching to opera.

I hope you’ve enjoyed this comparison, tell your friends. If you find any errors, please comment. You can link here using browser.theymos.com

Free firewall Comparison

Introduction

This is a comparison of free firewalls. These are the top user rated free firewalls on download.com. The computer running these firewalls has 1GB of RAM and 3.6Ghz CPU. The same applications were running during all firewall tests. MultiTrack Stopwatch was used for the usability test. These tests were performed on each firewall:

port scan

This will measure if the firewall is properly covering the computer’s ports. Full stealth is best, closed is ok, open is really bad. If it has full stealth, it gets one point, if it has open ports, it gets -1 points.

memory and CPU used

The amount of RAM and CPU the application uses. I used task manager to test this. I also tested how it changed while I was doing the port scan. The application with the least memory gets one point, the one with the most gets -1; the same for CPU.

size

The size of the program on the disk after default installation. The least gets one point, the most gets one subtracted.

ease of use

This is measured on how long it takes me to figure out how to make an application exception for the firewall. N/A indicates either that feature isn’t included, or I couldn’t figure out how to do it in 5 minutes. The least gets one point, any with N/A get one subtracted.

proper uninstall

After uninstalling and rebooting if it tells me to, I’ll search the computer for signs of it. If yes(it was properly uninstalled), it gets one point; points will not be subtracted.

blocks outgoing?

I will run a batch file that runs a tracert(trace route). It should block it, to prevent malware from communicating with the outside world. If yes(it blocks it), it gets one point; points will not be subtracted.

Data

	tests
firewalls	port scan	memory(idle/”under attack” in KB)	CPU usage %(idle/attacked)	size	ease of use	clean unistall?	block outgoing?	notes	points
filseclab	5 ports were closed, but not stealthed	17,428/17,428	0/2-4	about 21MB	20	.lnk file remained	yes	Simple firewall, not the best or worst.	0
GeSwall	closed only	6,372/6,372	0/0	about 6MB	25	no	no	Not really a firewall, it puts your applications in a “sandbox” where they can’t do any harm. Great memory and CPU usage, but not the best security.	1
C.O.M.O.D.O	full stealth	20,000/20,300	0/1	about 16MB	18	no	no	Comodo is bulky, but good for new users. Easy configuration and good security.	0
safety.net	closed only	15,000/15,000	1/1	about 9MB	47	no	no	Poor firewall. It has high resource usage, bad security, and a hard to understand user interface. I don’t recommend using this firewall.	0
Jetico	full stealth	8100/8300	0/1	about 3MB	50	yes	yes	Low resource usage, perfect security, and many options make this one of the best firewalls. This one is not for inexperienced users and the user interface is hard to navigate.	3
Ashampoo	closed only	12300/12300	1/2	about 11MB	20	no	no	This one has the best looking user interface of all of them. A good one for inexperienced users. It has bad resource usage and security, so I wouldn’t recommend this to experienced users.	0
AS3	closed only	9656/9656	2/4	about 636KB	N/A	no	no	Worst.Firewall.Ever. It dosn’t ask you whether it should accept or deny programs it hasn’t seen before. It just blocks them. On top of that, there is no way to allow all connections from a certain application; ports only. At least it’s small.	0
hurricanesoft	closed only	15,100/15,700	2/4-10	about 2MB	N/A	No, the uninstaller dosn’t work	no	Very simple. You can’t add your own rules in this application, and almost everyone needs to do that at some point. I don’t recommend this one.	-2
ZoneAlarm	full stealth	4,500/4,500	0/0	about 8.5MB	N/A	yes	no	Great resource usage and security, but I couldn’t figure out how to add an application exception.	3
WyvernWorks	Some ports were open!	10,000/10,000	0/1	about 2.4MB	12	no	no	Stylish and easy to use, but some ports were open with this! It looks like the software was made in 2004 but the company’s website, wyvernworks.com dosn’t seem to exist any more.	0
Windows firewall	closed only	N/A	N/A	N/A	15	N/A	no	Not actually too bad, it’s easy to use and security is ok, but not great. Very inexperienced users with XP SP2 should probobly stick with the windows firewall.	0

Summary

The best two firewalls were Jetico Personal Firewall and ZoneAlarm. If you can handle the interface, I recommend Jetico. Inexperienced users should use ZoneAlarm or COMODO. If you want a really nice looking application, go with Ashampoo. If you want extra security, you can run GeSwall along with your firewall. Avoid AS3, Hurricanesoft, and WyvernWorks.