Yesterday I finished my mythology formal model research for a some old sci-fi TV series "The Lost Room" (2006). Formal model is expressed in a form of axioms, theorems, geometry, etc. that make the physics more plausible. The paper is published on gist: https://t.co/jbRFnyZVl3
@awakecoding I would say..forget it. Based on my past experience with their proprietary stuff -- no chance. At least not within X.509 ecosystem. Maybe in MTC or whatever they come with for signing.
@awakecoding I don't believe it will happen any time soon. And I believe that in MSFT there are people who share your wish, but business reality is as this. AAS even can't afford their custom extension, so they have to reuse EKU to include signer ID info.
@awakecoding Even more: GUI doesn't do anything, it queries CryptFormatObject and render whatever the function returns. And it is a breaking change I'm talking about.
@awakecoding "The certificate viewer is long overdue for improvements anyway" -- right, because it is too expensive from support and backward compatibility perspective. You need to realize how much 3rd party code rely on current renderer.
@awakecoding And just add: EKU and X.509 profile is very generic and isn't tied exclusively to authenticode, TLS, S/MIME or whatever. Leaking context there would be the worst thing they did to CryptoAPI.
@awakecoding But "whatever they want" would cost a hell of money. Rewriting CryptoAPI for something that can change any time is expensive given how deep it is integrated and what is core part of OS.
@awakecoding Except that this doesn't scale in code level, because RFC 5280 doesn't support such contextual parsing not to say that it is a breaking change in UI. Moreover, no APIs would support this.
Hey Twitter fellows! I've got some updates, I'm now available in Gopher Hole: `gopher://hole.sysadmins.lv`. It is my new space for random thoughts and PKI rants.
@awakecoding OID resolution is based on local OID mapping database, OIDs alone don't carry this info. And I'm not sure how local OID db could be updated with dynamic data. In AD you can centralize this, but not in workgroups, it will be individual process.
@richardhicks It is by design. And this is why I wrote SID Extension handling policy module: https://t.co/74nt4g9EGu. Offline templates disable deep extension check on CA side.