Notices by :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site), page 38
-
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:48:29 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@espectalll it's a known issue that telecoms historically do not care about implementing BCP38 because they like the profit driven from DDoS-related overages.
but with bandwidth pricing now being extremely low (some carriers now quoting less than a penny per megabit!!!) there really is no financial incentive to not implement BCP38 at this point. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:46:22 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
my network implements BCP38. if i can do it, so can you. i don't have any fancy Juniper or Cisco certifications. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:45:03 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
do you hate stuff like Cloudflare? why not make DDoS obsolete? demand your ISP implement BCP38 today. the answer to these problems has been known for literally decades at this point. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:41:30 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@march exactly -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:40:51 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
there really is no reason not to do it either. you just restrict ingress traffic from IPs that could not possibly be routed to an interface at the customer-facing side of your network. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:38:59 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
imagine how much better the internet would be if people actually enforced BCP38 on their networks -
Oneesan succubus (lain@pleroma.soykaf.com)'s status on Thursday, 18-Jul-2019 09:38:19 EDT Oneesan succubus
Mexicans practicing for Trump's wall
Tusky_1563457076875_C6VPTUED2K.… -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:29:12 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
my intention is for Pleroma to take a secure-by-default stance. this is congruent with our historical security posture.
it will be possible to disable capability enforcement and such, but that won't be the default, for good reason. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:23:32 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@succfemboi no, they are going to be in pleroma core. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:20:39 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
the path to adopting bearcaps then becomes:
- wait for bearcap spec to come out
- adapt bearcaps to internal capabilities using a plug to map the bearcap to identity + capability set
- implement bearcap negotiation strategy
- get other implementations to switch to bearcaps
- eventually stop using signatures altogether -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:18:19 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
basic plan:
- new plug: MappedSignatureToIdentityPlug. maps the valid signature to the identity of the signature based on it's KeyID.
- new plug: MappedIdentityToCapabilityPlug. maps the valid identity or null identity to a bounded set of capabilities according to the configured policy hooks. default policy grants access to anyone, like MRF.
- control access based on mapped capabilities (effectively noop by default policy since it's accept-all)
- rework non federation traffic to also leverage mapped capabilities -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 10:00:47 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
now to work on mapping these signatures to virtual capabilities.
things we will need:
- policy hook (to determine what capabilities a signed request may or may not be granted)
- access control of fetches based on possessed capabilities
- modification of elixir plugs to handle the actual capability mapping (and structured in a way where it can work with bearcap invocations later) -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 09:53:54 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@bn4t soon(tm) -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 08:06:50 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@pony why the fuck would they need the X resource database -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 07:22:29 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@glitch @absturztaube @lucy
not for now -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Thursday, 18-Jul-2019 07:14:21 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
"blocks are forbidden in the AP S2S spec" is a maliciously bad take
the actual reality is "blocks aren't discussed in the AP S2S spec" -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Wednesday, 17-Jul-2019 19:44:29 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@flussence if you think supervisor is a shitshow, look at gaffer -- the same idea, but made by benoitc (therefore you know it's awful) -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Wednesday, 17-Jul-2019 17:04:35 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@static @1iceloops123 @ffs @fluffy @lnxw48a1 @roka
no, that was something else. -
:abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy: (kaniini@pleroma.site)'s status on Wednesday, 17-Jul-2019 16:56:59 EDT :abunhdhappyhop: :abunhdhappy: :abunhdhop: :abunhd: :abunhdhappyhop: :abunhdhappy:
@waifu weh