Feature #6908
Make Adblock Plus filter lists lighter
100%
Description
Our filter lists for Adblock are redundant. EasyList exists 3 times: 1st list is EasyList, 2nd list is ListFR+EASYLIST, 3rd list is EasyList Germany+EASYLIST.
* both easylist+fr and easylist+de are larger than the regular
easylist
* I’ve grepped a few patterns from easylist and found them in the two
other lists
So it looks like we’re shipping the same patterns three times.
Subtasks
Feature #6909: Investigate Adblock pattern duplication | Resolved | 100 |
|||
Feature #6910: Drop the regular easylist from our list of subscriptions | Rejected | 0 |
History
#1 Updated by spriver 2014-07-11 17:00:07
I would like to do the task.
#3 Updated by spriver 2014-07-15 08:42:13
I was thinking about merging the lists together into one custom list but this would lead to a security issue regarding to anonymization (the user could be tracked by monitoring which elements are getting blocked on a webpage). Or is that a minor issue? Has anyone got ideas on that?
-idea no 1: Only use one public avialable list (“standard EasyList”)
-idea no 2: Merge all three lists into one custom and removing redundances
-idea no 3: keep current setup
Absolutely other idea: we could use privoxy rules to block trackers and ads by default (by importing easylist as blocking rules)
#4 Updated by intrigeri 2014-07-15 18:06:20
> I was thinking about merging the lists together into one custom list but this would lead to a security issue regarding to anonymization (the user could be tracked by monitoring which elements are getting blocked on a webpage).
I think that’s orthogonal to the problem at hand, which is only about making ABP lists lighter.
> Or is that a minor issue?
We discussed this two years ago, and decided that it was acceptable to have Tails be fingerprintable at the exit node level (e.g. due to shipping AdBlock), as long as it shows the same fingerprint as the Tor Browser on the webmaster’s side.
If you’re interested in rebooting this discussion, please start a thread on tails-dev@ :)
#5 Updated by spriver 2014-07-16 07:23:29
> We discussed this two years ago, and decided that it was acceptable to have Tails be fingerprintable at the exit node level (e.g. due to shipping AdBlock), as long as it shows the same fingerprint as the Tor Browser on the webmaster’s side.
>
> If you’re interested in rebooting this discussion, please start a thread on tails-dev@ :)
So do you mean the fingerprintig without AdBlock then? (e.g. fonts, window size)? I’m asking because I think that our browser diffs much from the TBB. Maybe it would be also useful to add a plugin like Disconnect or Ghostery.
#6 Updated by intrigeri 2014-07-16 10:34:02
> So do you mean the fingerprintig without AdBlock then? (e.g. fonts, window size)? I’m asking because I think that our browser diffs much from the TBB. Maybe it would be also useful to add a plugin like Disconnect or Ghostery.
I’m glad you’re interested in researching this topic further. However, may we please have this discussion on tails-dev@, instead of burying it in comments on an unrelated ticket where basically nobody will see it (not mentioning it adds off-topic noise to the ticket discussion)? Thanks in advance :)
#7 Updated by spriver 2017-01-09 15:10:15
As I reported on Feature #6909 the regular EasyList subscription can be dropped from ABP lists’ subscriptions. The issue does not exist with uBlock (Feature #9833), as uBlock is only subscribing to the additional EasyListFR and EasyListDE (ABP ships EasyList+FR resp- EasyList+DE).
#8 Updated by anonym 2017-01-12 14:16:04
- Status changed from Confirmed to Fix committed
- Assignee deleted (
spriver) - Target version set to Tails 2.10
spriver wrote:
> As I reported on Feature #6909 the regular EasyList subscription can be dropped from ABP lists’ subscriptions. The issue does not exist with uBlock (Feature #9833), as uBlock is only subscribing to the additional EasyListFR and EasyListDE (ABP ships EasyList+FR resp- EasyList+DE).
Now that we ship uBlock Origin (Feature #9833) we have gotten this for free!
#9 Updated by anonym 2017-01-24 20:47:57
- Status changed from Fix committed to Resolved