Level 1
You're totally unprofessional (as many AV reviewers; and I got my comments blocked, censored, or I'm getting warnings from the moderators for speaking the truth). In this particular case:
1. When Panda blocked a website and it didn't count any blocked threat it was because it did NOT download and scan anything! The URL was blocked just because it was in a list of malevolent URLs (such lists are often outdated and also list safe sites as infected, but they're still used). Either way, counting the blocked URLs as "success" is wrong: only downloaded, scanned and detected FILES should be counted. Blocking sites based on blacklists is stupid.
2. When IE seemed stuck because it loaded a binary (it was an EXE, not a script; the reviewer can't stop saying BS), why was the reviewer thinking something wrong was happening in the background? It wasn't a malevolent JS. It was just a binary served as in-line, and the dumb IE tried to find some HTML to display — this is why it took over the CPU.
3. Finally, when a bunch of files were left undetected by repeated on-demand scans, the reviewer didn't have the minimum number of neurons to try to execute each and every of them! IN SO MANY CASES, an AV product detects such malware ON EXECUTION, and blocks or deletes the file. Maybe Panda would have detected half of them, or maybe not. We'll never know.

As a general rule, people who make such quick YouTube reviews don't bother to come with a fully functional list of malware. Here, at least one URL was not working. Each and every malware should be tested as existing prior to such a test — how much time would this need? 15 minutes?

I don't know why I'm wasting my time on such a forum. Have you noticed that employees at the major security vendors usually don't bother to monitor this forum and to answer with regards to their products?


Level 85
@Ludditus: I can see valid points on the information.

Well its already a part of product to test from immediate protection so of course different side of story test without web filter. *

* Which sometimes the result are different because of detection/signatures it can provide


Then to the second one, yes it can be an isolated issue however sometimes the samples taken causes to influence the performance. But usually the system is indeed not infected (e.g malicious script, or exploit trying to execute)


I understand the sentiments, of course that type of test is pretty common in order to assess, how the product last against those undetected threats. Because many programs can manage to prevent infection even though bypass from on-demand test.


Overall, many programs have flaws; of course its a right to defend from any side because of possible discrepancy and irregularities. :)