Main EasyList Filter changes .. so keep yours eyes open
Wladimir & I exchanged a little info back & forth today and I am now making changes (slowly) to the main EasyList filter.
You will probably notice in the one to five days it takes to get updated (at least if you update from adblockplus.org) that the EasyList is starting to look a little different .... actually the strings are getting a bit longer.
Wladimir has determined that I should make at least 8 unbroken characters per string. Adblock Plus will create a 'shortcut' with any 'simple expression' filter string using at least 8 unbroken characters ... anything less or using regexps and it will not (the 'patterns.ini' file in the 'adblockplus' directory of your Firefox profile actually tell you which ones have a shortcut and what it is).
The beauty of shortcutting a string is that it only needs to load once (if I'm saying this correctly ... Wladimir can chew me out if I'm wrong). Any time a simple expression filter (which is what I use exclusively and is preferred for use in ADP) meeets this criteria, the filter can parse at a much faster rate when it is needed multiple times. If the expressions are too short ... or if they are regular expressions (regexps), the string will not work in this manner ... the filter string need to be re-read over and over each time.
So, according to Wladimir, by actually making the filter larger and the strings longer to create 8 succesive characters, it will actually work faster.
So, I'm doing what I can and have actually changed, removed, and edited some strings ... but I need people to keep an eye out for anything not working right, false-positives, or missed blocks.
CHANGES: to the standard EasyList Filter
CHANGES: to the standard EasyList Filter
Last edited by rick752 on Fri Sep 08, 2006 4:03 pm, edited 2 times in total.
"Experience is something you don't get until just after you need it"
"Unbroken" means NOT breaking up the characters with a center wildcard like I do with some strings. The 3rd (or even 4th) wildcard is ok as lomg as one of the parts has at least 8 characters.
"Characters" can be anything ... not just letters. They can include punctuations and slashes.
.. would be 8 successive characters
... would not.
But:
... would be ok because 'icedogg1' has 8 unbroken characters
"Characters" can be anything ... not just letters. They can include punctuations and slashes.
Code: Select all
rickad?/
Code: Select all
rickad?*/ads
But:
Code: Select all
*rickad?*/ads*icedogg1*
Last edited by rick752 on Wed Nov 01, 2006 11:45 pm, edited 4 times in total.
"Experience is something you don't get until just after you need it"
While looking thru my filter strings, I notice that some are the name of the site without the 'http' or '.com' or '.net'.
When I first was making the filters, I was trying to keep it as compact as possible because Adblock actually worked better that way.. but Wladimir has since changed all of that ... looks like "longer is better" (don't get excited, ladies )
So I guess I will try to accomodate the program.
When I first was making the filters, I was trying to keep it as compact as possible because Adblock actually worked better that way.. but Wladimir has since changed all of that ... looks like "longer is better" (don't get excited, ladies )
So I guess I will try to accomodate the program.
"Experience is something you don't get until just after you need it"
-
- Guest
-
- Adblock Plus Author
- Posts: 444
- Joined: Thu Mar 09, 2006 1:01 pm
- Location: Cologne, Germany
I still have more to do .... I just have to find out what the "more" IS on other strings. Put in more than a few hours last night. There probably are a few adjustments I'll have to make, but after visiting a few hundred sites last night, I think everything seems to be working as good or even better than before.Wladimir Palant wrote:@Rick: Thanks for doing this, I already see 63ms instead of 98ms in my test - that's quite an improvement. I should have looked at it earlier, would have been much easier for you to fix.
I should be able to get it down to 40ms -ish.
"Experience is something you don't get until just after you need it"
Oops:Vallejo wrote:*http://.ad.*?* doesn't look like it would work and should probably be *http://*.ad.*?* (even though it would have <8 unbroken chars).
Yeah, you're right about that one. I don't know why I added the 'http' to it.
"Experience is something you don't get until just after you need it"
-
- Guest
I'll take care of that later. There will probably be a few days of adjustments. So bear with meAnonymous wrote:You removed *adserv* , thus smartadserver.com is NOT blocked anymore, while it is obviously advertising on many websites.
"Experience is something you don't get until just after you need it"
@ Guest .. it is now fixed (please post in the SAME topic and not in 2 different ones ... thank you).Anonymous wrote:You removed *adserv* , thus smartadserver.com is NOT blocked anymore, while it is obviously advertising on many websites.
"Experience is something you don't get until just after you need it"