Apparently ads and trackers are regularly not blocked by ABP for Chrome and IE for technical reasons related to the limited "power" that add-ons have in these browsers. I can't check first hand how much of a fact this is since I only use Firefox.
But that does NOT make failing to block network requests for which rules exist acceptable on Firefox! Really, please don't lower your quality requirements for Adblock Plus on Firefox just because such quality cannot be reached on other browsers. That would be such a messed up reasoning.
Not to mention Firefox is the reason for Adblock Plus' success. We want a reliable ad and tracker blocker that honors lists perfectly by blocking them before a network request occurs.
I read Wladimir say that in rare cases, ABP on Firefox prior to 2.6 already failed to block stuff. Well, first of all I would like to hear of such situations, like when do they trigger and at what frequency. But more importantly let us NOT add more of those situations, especially a leak as serious as we have on 2.6 !
If that's the direction we are heading, where ABP starts to compromise with the very purpose of its existence, its future is pretty gloom. I won't be able to trust it any more. I mean, where did you get the idea that it would be acceptable to ship 2.6 as it is, with the many cases where this would lead to tons of network requests passing through when they should be blocked ?
Please note that I am not the type of guy to freak out and panic and imagine stupid things. I'm saying this in a very down to earth way, I know how you intend to minimize the problem by speeding up ABP's startup, but:
1/ If at all possible, it will take time to reach a speed high enough to make the amount of unblocked requests negligible on startup. In the mean time we are left with no response to our various arguments and we have no time frame.
2/ This sets a worrisome precedent that basically says Adblock Plus will merely attempt to honor lists given to it. It's even worse because your success on IE and Chrome may let you think that it is acceptable to have a product that only partially fulfils its promise to the end user.
We are on Firefox, you can fulfil the promise for real, so you have to do it. Adblock Plus has become a profitable business with several millions of users and several millions of dollars to act with, so you kind of have a duty now. :/
3/ We are ill informed about when requests pass through and the likeliness of such an event. I had to dig out the seriousness of 2.6 issue, and on top of that your communication really minimized it. So in the future, how can I trust that no other serious losses in reliability will occur without me noticing ? I cannot. For reasons given above, I'm forced to think that it will occur again, that you will try to minimize it but won't succeed enough, and that the amount of unblocked network requests is unlikely to stay negligible as Adblock Plus evolves further.
I think a mix of communication and real changes in the development plan are necessary :/
I understand the pressure you may have from Firefox developers who have to deal with serious competition from Chrome regarding speed, but your product's promise goes first. You can eventually let the ultimate choice to the end user, speed vs full protection, but it's not for you to decide unilaterally is it ?