Set optInPartitioning = true #140

AnotherUser321 commented 2026年06月27日 09:23:50 +02:00

Originally added in #87, then accidentally labeled as redundant in #97.

Originally added in https://codeberg.org/librewolf/settings/pulls/87, then accidentally labeled as redundant in https://codeberg.org/librewolf/settings/pulls/97.

any1here commented 2026年06月27日 19:41:48 +02:00

Could you also add network.cookie.cookieBehavior.optInPartitioning.pbmode

Could you also add `network.cookie.cookieBehavior.optInPartitioning.pbmode`

AnotherUser321 commented 2026年06月28日 10:19:04 +02:00

optInPartitioning.pbmode only seems to matter when the normal pref is false. I can still add it but that's why I intentionally left it out

github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieService.cpp (L1072-L1080)
github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieParser.cpp (L698-L706)
github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieCommons.cpp (L494-L501)

`optInPartitioning.pbmode` only seems to matter when the normal pref is `false`. I can still add it but that's why I intentionally left it out https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieService.cpp#L1072-L1080 https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieParser.cpp#L698-L706 https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieCommons.cpp#L494-L501

any1here commented 2026年06月28日 19:47:28 +02:00

@AnotherUser321 wrote in #140 (comment):

optInPartitioning.pbmode only seems to matter when the normal pref is false. I can still add it but that's why I intentionally left it out

github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieService.cpp (L1072-L1080) github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieParser.cpp (L698-L706) github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieCommons.cpp (L494-L501)

I would still add it in case the behavior would change upstream

@AnotherUser321 wrote in https://codeberg.org/librewolf/settings/pulls/140#issuecomment-18201170: > `optInPartitioning.pbmode` only seems to matter when the normal pref is `false`. I can still add it but that's why I intentionally left it out > > [`github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieService.cpp (L1072-L1080)`](https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieService.cpp#L1072-L1080) [`github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieParser.cpp (L698-L706)`](https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieParser.cpp#L698-L706) [`github.com/mozilla-firefox/firefox@cea5f8cacb/netwerk/cookie/CookieCommons.cpp (L494-L501)`](https://github.com/mozilla-firefox/firefox/blob/cea5f8cacbe1ba3c8c95a7b17a309ce6704a4c7d/netwerk/cookie/CookieCommons.cpp#L494-L501) I would still add it in case the behavior would change upstream

ltguillaume commented 2026年06月29日 17:15:11 +02:00

I'd like to know why this is necessary. According to https://searchfox.org/firefox-main/source/modules/libpref/init/StaticPrefList.yaml#14302 the only thing this does is that it Changes cookieBehavior=5 to block third-party cookies by default.

network.cookie.cookieBehavior is already set to 5 by default. I may have looked over something, so in this case please do educate me, but otherwise it looks to me like it's yet another unnecessary change to explicitly set something that doesn't need to be set.

Edit
Ah, it actually makes sense now I read Changes cookieBehavior=5 from dynamic partitioning to block-by-default with opt-in. (https://searchfox.org/firefox-main/source/docs/_addons/etp_matrix.py#84).

The description from the first comment was ambiguous (when reading to as in order to).

That said, won't this break a lot of stuff, which is exactly why dynamic partitioning was made?

I'd like to know why this is necessary. According to https://searchfox.org/firefox-main/source/modules/libpref/init/StaticPrefList.yaml#14302 the only thing this does is that it `Changes cookieBehavior=5 to block third-party cookies by default`. `network.cookie.cookieBehavior` is already set to `5` by default. I may have looked over something, so in this case please do educate me, but otherwise it looks to me like it's yet another unnecessary change to explicitly set something that doesn't need to be set. **Edit** Ah, it actually makes sense now I read `Changes cookieBehavior=5 from dynamic partitioning to block-by-default with opt-in.` (https://searchfox.org/firefox-main/source/docs/_addons/etp_matrix.py#84). The description from the first comment was ambiguous (when reading `to` as `in order to`). **That said, won't this break a lot of stuff, which is exactly why dynamic partitioning was made?**

any1here commented 2026年06月29日 20:25:23 +02:00

@ltguillaume wrote in #140 (comment):

That said, won't this break a lot of stuff, which is exactly why dynamic partitioning was made?

We used to have this set until a certain someone removed it. It would depend on how far sites have gotten in supporting CHIPs, which I have no idea about.

https://developer.mozilla.org/en-US/docs/Web/Privacy/Guides/State_Partitioning#storage_access_heuristics is still enabled, which would help reduce breakage.

@ltguillaume wrote in https://codeberg.org/librewolf/settings/pulls/140#issuecomment-18293189: > That said, won't this break a lot of stuff, which is exactly why dynamic partitioning was made? We used to have this set until _a certain someone_ removed it. It would depend on how far sites have gotten in supporting CHIPs, which I have no idea about. https://developer.mozilla.org/en-US/docs/Web/Privacy/Guides/State_Partitioning#storage_access_heuristics is still enabled, which would help reduce breakage.