4 ms·

I'm sure they do feel they're working hard to compromise on this issue, but their decisions for this will break sites with very careful styling/designs (e.g. ch

by thorax 17y ago

I'm sure they do feel they're working hard to compromise on this issue, but their decisions for this will break sites with very careful styling/designs (e.g. checkmarks, font styling, etc). It just doesn't sit well with me that Mozilla feels so strongly (or caves to pressure) on issues like this that they're willing to break so much backwards compatibility. I would love to see them make this optional and turn it off by default. It's a paranoia thing that people have lived with forever, and the most concerned people could turn on this mode if they ever need to do so.

tptacek 17y ago
It only seems like paranoia because widespread, highly-publicized abuse of history sniffing hasn't started yet. I'm happy they've jumped onto this grenade before it exploded. The most valuable thing Mozilla just did was to send a signal to the market that history sniffing is going to be unreliable, and that ad networks shouldn't invest time and money into exploiting it. The scariest thing about history sniffing was the idea that some large company would adopt it and create pressure to keep the vulnerability there forever, as part of how the Internet works.
- thorax 17y ago
  Yeah, I know it's valuable to put out some solution to this, but I would have preferred more options discussed on how to prevent this rather than choosing an option that breaks presentation backwards compatibility. I know I have the minority opinion on this, so I have to speak up or that opinion will go unheard. I prefer to avoid breaking backwards compatibility for presentation as much as possible. I'd much rather some approach be tried/found that isn't quite so inelegant as this one.
  - briansmith 17y ago
    Many, many options were proposed. Look in the Bugzilla entry (#147777) for it. There were over 241 comments in that bug report alone, not counting dupes. Many of those comments were proposing alternatives.
_delirium 17y ago
I guess I'm struggling to think of a case where it's particularly important that visited/unvisited links differ by complex styling. If leaking history is a paranoia thing, that's a useless fluff thing. The checkmarked visited links thing seems completely wrong to me, to take that example. Something like a checkmark next to a link feels like part of site UI rather than the browser's, and a site's UI shouldn't have access to what I've done on other sites! If they want to keep track of what links I click on the site and maintain checkmarks in their UI accordingly, they can still do that, of course.
- thorax 17y ago
  All presentation could be described as "useless fluff". If you wanted to change the font for a visited link aside from color, this will eat that. Note that I take the word "paranoia" from Mozilla's own blog discussing the topic. I do feel that it is a real privacy issue, but it isn't new and I'd much prefer something like what SafeHistory does than this presentation level hack they're planning.

by thorax 17y ago

tptacek 17y ago

It only seems like paranoia because widespread, highly-publicized abuse of history sniffing hasn't started yet. I'm happy they've jumped onto this grenade before it exploded. The most valuable thing Mozilla just did was to send a signal to the market that history sniffing is going to be unreliable, and that ad networks shouldn't invest time and money into exploiting it. The scariest thing about history sniffing was the idea that some large company would adopt it and create pressure to keep the vulnerability there forever, as part of how the Internet works.

thorax 17y ago
Yeah, I know it's valuable to put out some solution to this, but I would have preferred more options discussed on how to prevent this rather than choosing an option that breaks presentation backwards compatibility. I know I have the minority opinion on this, so I have to speak up or that opinion will go unheard. I prefer to avoid breaking backwards compatibility for presentation as much as possible. I'd much rather some approach be tried/found that isn't quite so inelegant as this one.
- briansmith 17y ago
  Many, many options were proposed. Look in the Bugzilla entry (#147777) for it. There were over 241 comments in that bug report alone, not counting dupes. Many of those comments were proposing alternatives.

_delirium 17y ago

I guess I'm struggling to think of a case where it's particularly important that visited/unvisited links differ by complex styling. If leaking history is a paranoia thing, that's a useless fluff thing. The checkmarked visited links thing seems completely wrong to me, to take that example. Something like a checkmark next to a link feels like part of site UI rather than the browser's, and a site's UI shouldn't have access to what I've done on other sites! If they want to keep track of what links I click on the site and maintain checkmarks in their UI accordingly, they can still do that, of course.

thorax 17y ago
All presentation could be described as "useless fluff". If you wanted to change the font for a visited link aside from color, this will eat that. Note that I take the word "paranoia" from Mozilla's own blog discussing the topic. I do feel that it is a real privacy issue, but it isn't new and I'd much prefer something like what SafeHistory does than this presentation level hack they're planning.