#25418 closed defect (bug) (fixed)
bloginfo for pingback_url doesn't respect SSL
Reported by: | technosailor | Owned by: | nacin |
---|---|---|---|
Milestone: | 3.7 | Priority: | normal |
Severity: | normal | Version: | 2.6 |
Component: | XML-RPC | Keywords: | has-patch commit |
Focuses: | Cc: |
Description
Attached patch moves from using get_option(), which is not SSL aware, to using site_url() which is.
Attachments (2)
Change History (13)
#1
@
11 years ago
- Component changed from General to XML-RPC
- Milestone changed from Awaiting Review to 3.7
- Version changed from trunk to 2.6
#2
@
11 years ago
- Milestone changed from 3.7 to Awaiting Review
- Severity changed from major to normal
#4
in reply to:
↑ 3
;
follow-ups:
↓ 5
↓ 6
@
11 years ago
- Milestone changed from Awaiting Review to 3.7
Replying to technosailor:
Kind of a security issue, so I'd say it's major.
I could see how this could be a problem, sure. Can anyone make a case that this could (or would not) break something?
#5
in reply to:
↑ 4
@
11 years ago
Replying to nacin:
I could see how this could be a problem, sure. Can anyone make a case that this could (or would not) break something?
All you need is a CSRF attack, now or in the future, to corrupt pingback_url and then encrypted traffic would leak. Theoretically.
#6
in reply to:
↑ 4
@
11 years ago
Replying to nacin:
I could see how this could be a problem, sure. Can anyone make a case that this could (or would not) break something?
As for breaking things, we're not changing bloginfo()/get_bloginfo() and site_url() already wraps, intelligently, around the siteurl option so... I can't foresee how this would possibly break anything. Eating our own dogfood and whatnot. If site_url() broke something, we'd already see it elsewhere... Plus, we have time for this to be baked in during the beta period
@
11 years ago
Here's a broader patch that ensures we eat our dogfood throughout core, but I don't feel as strongly about it and would be okay pushing to 3.8 for the second patch. Still think original patch needs to be in 3.7
#9
@
11 years ago
Each instance in siteurl2.diff needs to be reviewed to make sure the current protocol's page is proper for that result. I see a few that could cause problems, a few that are unnecessary, and a few that are definitely correct.
Existing issue, not major.