bretg

Based on the discussion over in PBJS, I've updated the proposal here to support an array of adjustments. I think this will work for the new use case of ```...

Ok, we're locking the protocol into exactly 3 dimensions: mediatype, bidder, and deal. Updated the description's example.

1. yes. 2. Added `Merge request-level ext.prebid.bidadjustments with account-level bidadjustments in the normal way: deep merge with preference for the request-level with array overwrite.` 3. Always mediaType, bidder, and deal...

Brian and I discussed > I assume the wildcard is simply a catch-all but need not be specified right? The order in which dimension values are specified does not matter....

@bsardo - fixed! We can make the config be explicit: video-instream and video-outstream.

Discussed in committee. A limit=0 will mean max-limit. We'll do this in 3.0.

Discussed in committee. There was concern that this may increase the prevalence of cross datacenter reads. In the end the prevalence of that scenario probably wouldn't increase as a result...

Close - we discussed the [GET endpoint](https://github.com/prebid/prebid-server/issues/3629) in the CTV committee, not the /vtrack endpoint.

Comments: 1. It's unclear to me that most Prebid Server host companies care about dchain, since the PBS company most likely isn't paying the publisher. 2. The PBJS field `bid.meta.dchain`...

Discussed in committee. @pm-saurabh-narkhede - can you help us understand the Prebid Server use case here? Few PBS host companies are payors as described by the IAB.