Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-26 Thread Corey Bonnell
-Original Message- From: Acme On Behalf Of Andrew Ayer Sent: Wednesday, March 22, 2023 4:30 PM To: Corey Bonnell Cc: acme@ietf.org Subject: Re: [Acme] ARI: Indication if certificate will be revoked Hi Corey, On Wed, 22 Mar 2023 17:55:59 + Corey Bonnell wrote: > Hi Andrew, > Is th

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-23 Thread J.C. Jones
This might be a little orthogonal, but what about the signal being some manner of "last updated" or "last modified" time for the ARI response itself? Generally that could be the issuance time of the certificate. That doesn't even have to be a field in the response document, we could utilize the

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Andrew Ayer
On Wed, 22 Mar 2023 14:16:40 -0700 Aaron Gable wrote: > I'm not totally sold on the utility of including extra information in > the ARI response, if that extra information will not modify client > behavior. If the purpose is to modify human behavior, then I believe > the current explanationURL

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Aaron Gable
I'm not totally sold on the utility of including extra information in the ARI response, if that extra information will not modify client behavior. If the purpose is to modify human behavior, then I believe the current explanationURL is sufficient. Adding a machine-readable problem document that

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Andrew Ayer
On Thu, 23 Mar 2023 01:55:06 +0900 Seo Suchan wrote: > I think it's pretty safe to say IFF ARI time changes from what it's > set just after certificate creation, you could guess there will be > revocation for that leaf certificate. I don't think that's a safe assumption - the CA could be

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Andrew Ayer
Hi Corey, On Wed, 22 Mar 2023 17:55:59 + Corey Bonnell wrote: > Hi Andrew, > Is the purpose of the "revocationTime" field such that ACME client > behavior would be different than the recommended replacement > time-selection algorithm in section 4.1, or is it to provide richer > metadata

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Andrew Ayer
On Wed, 22 Mar 2023 12:46:46 -0400 Amir Omidi wrote: > My concern with this is that it creates a bit of a requirement to > revoke by/on that time, which doesn't seem to be the intent of ARI I > think? > > Also what should the precision of this time field be? day/hour/etc? The same as the

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Corey Bonnell
e@ietf.org Subject: [Acme] ARI: Indication if certificate will be revoked I'm working on adding an ARI client to a certificate monitoring service to notify users when one of their certificates is scheduled to be revoked. Unfortunately, ARI doesn't currently convey whether the suggestedWindow is mand

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Seo Suchan
IIRC it was dual purpose: state some randomish time to reduce load spike at 12:00AM or mass renewal after mass revocation event, and order renew when revocation is imminent. I think it's pretty safe to say IFF ARI time changes from what it's set just after certificate creation, you could

Re: [Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Amir Omidi
My concern with this is that it creates a bit of a requirement to revoke by/on that time, which doesn't seem to be the intent of ARI I think? Also what should the precision of this time field be? day/hour/etc? On Wed, Mar 22, 2023 at 10:35 AM Andrew Ayer wrote: > I'm working on adding an ARI

[Acme] ARI: Indication if certificate will be revoked

2023-03-22 Thread Andrew Ayer
I'm working on adding an ARI client to a certificate monitoring service to notify users when one of their certificates is scheduled to be revoked. Unfortunately, ARI doesn't currently convey whether the suggestedWindow is mandatory (because the certificate is going to be revoked) or merely