urlclassifier3.sqlite

[couldabeen]

What is it? and WHY is it so damned big (Approx 12,000KB and growing)?
And can I delete it, or use the SQL database browser to edit or remove a bunch of what looks like encrypted domains.

I'm up to the Beta 5pre version of the nightlies.(sorry, posted this with a BonEcho nightly)

[Littlemutt]

Its the Phishing filter data-base from Google. Its a list of sites known to be phishing sites and will throw up the new Malicious site warning if the URL your visiting is in the database. I think last time I looked at mine it was around 19meg or just a tad over.

Its now all downloaded at once, but in chunks to reduce the impact to dial-up users.

[BenBasson]

12,000KB is big?

[couldabeen]

To me it is. Oh, I have plenty of disk space. That's not the problem.
I don't know how big the comparable file would be in BonEcho. I don't have that option set there, AND I don't have it set in Minefield. It seems I got that 'big' database it without asking for it.

[Bluefang]

Yeah, mine is 14MB.

Do you have the "suspected attack" and "forgery" checking turned off? If you do, then that might be a bug.

But there really is no reason not to use those features. I don't think it uses live data anymore, only the data in the urlclassifier3.sqlite, which I think gets rid of most privacy concerns.
https://bugzilla.mozilla.org/show_bug.cgi?id=388652

[the_dees]

If both options are disables the file remains. However, if you delete the file,it should be recreated as a 0-Byte file and stay that way.

[alterna]

I have a urlclassifierkey3.txt (154 bytes) but no urlclassifier3.sqlite.
I have both forgery and attack checked in options.
How can I download the database?

Thanks.


Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9b5pre) Gecko/2008032106 Minefield/3.0b5pre ID:2008032106

[couldabeen]

Yeah, mine is 14MB.

Do you have the "suspected attack" and "forgery" checking turned off? If you do, then that might be a bug.

But there really is no reason not to use those features. I don't think it uses live data anymore, only the data in the urlclassifier3.sqlite, which I think gets rid of most privacy concerns.
https://bugzilla.mozilla.org/show_bug.cgi?id=388652

I know it's not using live data any longer, but it jumped nearly 4MB in the past two days, and Yes I have both options turned off. So it's not live, but IT IS updated, and it looks to me like it is updated whether it is used or not. I'll have to do some more digging I guess.

@alterna I don't know where the DB is located for download, but PM me with an email address, and I'll send you mine, it's not doing me any good except to take up space I didn't ask it to.

[Bluefang]

The file should be automatically downloaded/updated so you shouldn't have to do anything special.

[Bluefang]

couldabeen, I turned off those 2 settings and deleted the urlclassifierkey3.txt and urlclassifier3.sqlite. After restarting, they were recreated and the DB has stayed below 9KB (which I would assume is the empty DB structure).

[couldabeen]

^Thanks for the information. I'll go ahead and delete and see what happens.

[alan_x]

Another thing this file does is screw up Sandboxie, which throws an error because the file exceeds Sandboxie’s file copy size limit, set to keep Sanboxie’s startup time within reasonable limits.

Given that, according to the zdnet.au article linked to above, 68 percent of all internet-based malware is now being hosted on legitimate sites, we need a better way to do this.

BTW, it wouldn't hurt if Webmasters (including Mozillazine) would refrain from using active content wherever possible. Javascript is fun, but, given the risks it exposes users to, it's about way overused.

[Bluefang]

1. Well then don't use Sandboxie. Firefox only touches 3 folders (profile, cache, and install) so there is no real need to sandbox it. Plus, most malicious browser attacks go through memory, not files. Or turn off the safe-browsing functionality.

2. JS is the direction the internet is moving (Web 2.0 anyone?) so the unfortunate fact is that you're going to have to get used to it or turn it off and deal with limited page functionality. Plus, I think you're way over-exaggerating the security problems. You can use AdBlock Plus or NoScript to block 3rd party scripts (possible XSS attacks).

3. Of that 68%, how much of that only targets IE? Probably most of it. And a majority of that probably doesn't act through JS.

[rickst29]

Given that, according to the zdnet.au article linked to above, 68 percent of all internet-based malware is now being hosted on legitimate sites, we need a better way to do this.

The database is HUGE because there really ARE lots of "completely bad" URLs out there. But you're right, a careful Firefox user needs even more protection. And we've got it! Use the "noscript" extension.