Palant rants about Adblock (without Plus)

Discussion in 'other software & services' started by tlu, Jul 29, 2014.

Thread Status:
Not open for further replies.
  1. tlu

    tlu Guest

  2. HAN

    HAN Registered Member

    Joined:
    Feb 24, 2005
    Posts:
    2,080
    Location:
    USA
    I'm not in the mood to verify what he's written but it is interesting to read. I used Adblock for a while a year or two ago. I've ran 2 or 3 other ad blockers since and never went back. Doubt if I will as I'm happy with my current choice...
     
  3. gorhill

    gorhill Developer

    Joined:
    Nov 12, 2013
    Posts:
    745
    Location:
    Canada
    I don't doubt his findings. I personally trust ABP way more than AdBlock.

    A while ago ([here](https://news.ycombinator.com/item?id=7704523)) I found that AdBlock was connecting to goldenticket.disconnect.me (through HTTPSB's behind-the-scene matrix), and nowhere could I find what was that about. Maybe the same day I saw Martin Brinkmann inquires AdBlock about this (here: http://support.getadblock.com/discussions/questions/501-what-is-the-deal-with-disconnect).

    The answer was less than satisfying, as I never saw that "survey-like UI " which was supposed to appear. I remembered trying AdBlock recently and goldenticket.disconnect.me was still there.
     
  4. Nanobot

    Nanobot Registered Member

    Joined:
    Jun 23, 2010
    Posts:
    237
    Location:
    Neo Tokyo
    I did what others didn't and I passed the message you left on HN as a reply to one of his comment regarding the ublock (https://news.ycombinator.com/item?id=8104651)

    and here's what Palant answered me back, directly in my mail (you can have a copy if you want). So don't look on his blog post to read it, you're not gonna find it there.

    Reply text:

    contents of private email removed

    There you are, Play nice and eventually you'll get what you want, Nobody claimed you aren't a capable developer but having this arrogant attitude (yes it's apparent to everyone) it's not gonna help you not only in this case but in everything you do in your life.
     
    Last edited by a moderator: Jul 30, 2014
  5. TheWindBringeth

    TheWindBringeth Registered Member

    Joined:
    Feb 29, 2012
    Posts:
    2,084
    This, together with other examples where browser and/or extension behavior was changed in a way that would concern some users, makes me contemplate the possible benefits of automated background communications monitoring. The rough idea:

    Some component constantly inspects the background communications and extracts what information it can from the requests it sees to create a presumably tiny background request database. The user could pull up a display for this, showing the background requests that their browser has made... and the background requests that each of their extensions has made... during a period (since the last time things were cleared or some moderately long rolling period). At a minimum this could include origin type information to see what hosts have been contacted via what protocols. At a maximum it could include all requests. Maybe there is a way to boil things down to something in between those two extremes. This feature might help users keep an eye on those requests without having to perform long-running captures via other means and to distill the results on their own. They could simply pull up some history/summary info whenever they wanted to.

    If the component knows what the browser and extensions have done in the past, it has the potential to alert the user when it sees a change in behavior. For example, if it hasn't seen extension X communicate with example.com before, it could immediately alert the user when that happens and even block it until the user has approved. Such a feature might keep users actively/promptly informed of changes while also reducing the amount of work they need to do to be so informed.

    Conceptually, such a component could have alerted users if/when Adblock started to communicate with goldenticket.disconnect.me... and maybe even recognized (if this is what actually happened) that a new unique identifier param had been added to background requests. It could alert users if/when their browser started to communicate with tiles.up.mozillalabs.com and/or some other new privacy reducing feature or experiment server.

    Frankly, I'm not sure how well the idea... particularly the automated recognition of changed behaviours part... would pan out in practice. However, I figured it was worth sharing.

    Edit: After some waking up, it occurs to me that the "automated recognition of changed behaviours" aspect may reduce to setting up a default action (block, block and alert, alert) for not yet whitelisted background requests.
     
    Last edited: Jul 30, 2014
  6. gorhill

    gorhill Developer

    Joined:
    Nov 12, 2013
    Posts:
    745
    Location:
    Canada
    Is that comment meant as an advice for me?
     
  7. Nanobot

    Nanobot Registered Member

    Joined:
    Jun 23, 2010
    Posts:
    237
    Location:
    Neo Tokyo
    Focus on what's important here i.e. Palant's desire to start a communication with you, not if my post was meant as advice or not, it's not personal.
     
  8. harsha_mic

    harsha_mic Registered Member

    Joined:
    Mar 11, 2009
    Posts:
    791
    Location:
    India
    nanobot, i am not sure where palant replied to gorhil asking about his inputs. And what exactly was broken in uBlock's implentation? I am just curious and then gorhill could focus on the these important things :)

    i am also curious to know what is that gorhill traded performance for memory. As far as i see HTTPSB+uBlock does not hang the webpages or cause memory spikes, unlike ABP which is true for memory.. And interested to see which are the sites that hangs due to uBlock/HTTPSB ( i dont see any examples anywhere), so, we could raise a ticket in github and gorhill could eventually fix it ;). He is amazingly super fast in fixing the issues...
     
  9. gorhill

    gorhill Developer

    Joined:
    Nov 12, 2013
    Posts:
    745
    Location:
    Canada
    Look I asked whether this was addressed to me because I don't even know what all that is about, given the content of the email you posted was removed, so it wasn't clear.

    Just to be clear re. "you'll get what you want": I want nothing, I don't know what you had in mind when you wrote that.

    The point is rather what I don't want: intellectually dishonest characterizations regarding the technical aspects of the work I did: "intentionally broken", "only the simplest filters are supported", "trading memory use for CPU" are all untrue statements. Lecturing me about life is not going to magically make these statements become true and sound.
     
  10. Wladimir Palant

    Wladimir Palant Registered Member

    Joined:
    Mar 11, 2007
    Posts:
    25
    Looks like I still have an account here from ancient times... There was nothing private about the reply I sent to Nanobot, it was simply extremely off topic:
    I don't have the time to analyze a large codebase every day, my comments were based on an older state that I saw. Maybe things improved since then, I would have to check that. I explicitly said that my reply wasn't about µBlock.
     
Loading...
Thread Status:
Not open for further replies.