1095602 - remove the STATE_SECURE_HIGH, STATE_SECURE_MEDIUM, STATE_SECURE_LOW distinction from nsIWebProgressListener

Status: RESOLVED FIXED

(Core :: DOM: Navigation, defect, P2)

Description

[Dana Keeler (she/her) (use needinfo) [:keeler]]

According to the documentation in nsIWebProgressListener.idl, if the state is STATE_IS_SECURE, the level of security is described by one of STATE_SECURE_HIGH, STATE_SECURE_MEDIUM, or STATE_SECURE_LOW. However, in the actual implementation, we only ever use STATE_SECURE_HIGH. In any case, the distinction is unhelpful; either you're using strong, known-secure encryption and we can reasonably say a connection is secure, or you're not, and we can't say it's secure at all. We should just remove any classification of this sort and rely on STATE_IS_SECURE vs. STATE_IS_INSECURE vs. STATE_IS_BROKEN (which, again, the implementation actually already does).

Comment 1

[(no longer active)]

I think we're almost out of these state flag values when bug 1504728 lands, so fixing this could be a nice way to get three more flag values back...

Comment 2

[Kyle Machulis [:qdot] [:kmachulis] (INACTIVE)]

Attachment: Bug 1095602 - Remove STATE_SECURE flags from nsIWebProgressListener;

Only STATE_SECURE_HIGH is used, and that's only in instances where
STATE_IS_SECURE is also used, so we can remove the security level
flags and just assume STATE_IS_SECURE is also STATE_SECURE_HIGH.

Comment 3

[Kyle Machulis [:qdot] [:kmachulis] (INACTIVE)]

Was doing some docshell triage and saw this, and it fits in with my work on bug 1503630, so figured I'd take care of it real quick.

Comment 4

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/b83cac4f938a
Remove STATE_SECURE flags from nsIWebProgressListener; r=Ehsan

Comment 5

[Andrei Ciure[:aciure]]

https://hg.mozilla.org/mozilla-central/rev/b83cac4f938a

Comment 6

[:aceman]

Hi, did this miss to remove the mention of the flags at https://dxr.mozilla.org/comm-central/source/uriloader/base/nsIWebProgressListener.idl#171 ?

Comment 7

[Dana Keeler (she/her) (use needinfo) [:keeler]]

comm-central is based on an earlier import of mozilla-central, so I don't think it has this change yet.

Comment 8

[:aceman]

Sorry the link was mentioning comm-central, but the file is in mozilla-central (and c-c does always contain an up to date copy of m-c). Anyway, the question was purely for m-c as the comment still mentions the now removed constants.
https://searchfox.org/comm-central/source/mozilla/uriloader/base/nsIWebProgressListener.idl#171

Comment 9

[:aceman]

https://searchfox.org/mozilla-central/source/uriloader/base/nsIWebProgressListener.idl#171

Comment 10

[Dana Keeler (she/her) (use needinfo) [:keeler]]

Ah, I see. Thanks.