Page 1 of 1

Input of cookie exceptions handled differently in 28

Posted: 2018-11-05, 18:01
by yellowfinch
On my old machine I had PM 27.4.2 and the cookie exceptions were recorded the same as they had always been in previous versions as well as in Firefox. For example, if I typed just reddit.com, it would be recorded exactly as entered.

On my new machine I have PM 28.0.1. Since I didn't find a simple way to export/import these exceptions, I started entering then manually and soon noticed that if I type just reddit.com, it results in http://reddit.com and https://reddit.com being added to the list. These addresses don't exist as such and this is confirmed by the fact that, in about:permissions, they show "0 visits" even though I have since logged in to Reddit.

Having both http and https is fine, but the reason why I entered only reddit.com is that I wanted the exception to cover both www.reddit.com and old.reddit.com.

For many of the exceptions on my old machine, I don't know if the actual, complete address is prefixed with "www.", or something else, or nothing at all. So I would have to access all of these Web sites individually to find out. Until then, it's like going back in time several years and having to fiddle with the cookie settings each time I access something for the first time on this machine.

Is there a known bug in the mechanism that handles the input of cookie exceptions?

Re: Input of cookie exceptions handled differently in 28

Posted: 2018-11-05, 19:19
by JustOff

Re: Input of cookie exceptions handled differently in 28

Posted: 2018-11-05, 19:47
by yellowfinch
Thanks, but the first one is about the inclusion (or not) of the protocol and the second one is about the use of wildcards, neither of which I questioned.

Except...

In the mean time, I experimented further. It turns out that my rule to accept cookies from https://reddit.com does work for both www.reddit.com and old.reddit.com, which reveals that the rule is applied as if there were a wildcard between the protocol and the domain name, even when there isn't. At least, that's the observable outcome.