Spec multibid #1055

morlovich · 2024-02-23T18:32:10Z

This permits generateBid() to return (and setBid to take) a sequence of bids as an alternative to a single one.
A k-anon rerun, still limited to producing a single bid, will now happen when none of the returned bids are k-anonymous (which is equivalent to current behavior when one bid is produced).

To make it easier to produce k-anonymous bids when component ads are involved without relying on a re-run,
this also adds optional targetNumAdComponents and numMandatoryAdComponents fields to bid to request implementations to drop some component ads in order to make a bid k-anonymous.

See #595

Preview | Diff

dmdabbs · 2024-02-26T16:50:50Z

With multiple bids and renderURLs to score, sellers' trusted fetches will have more opportunities to hit limits.
Good thing maxTrustedScoringSignalsURLLength (#767) is in the works.

dmdabbs · 2024-02-27T19:55:57Z

Should a bidder exceed the seller's perBuyerMultiBidLimits will it get a "reject reason" or does the buyer's participation silently fail/end unsuccessfully?

morlovich · 2024-02-28T19:21:57Z

Oh, adding a reject reason is a good idea... though not sure what's the story on spec'ing of that.

dmdabbs · 2024-02-28T19:34:23Z

I saw your PR for existing "automatic" rejection reasons: #1064

morlovich · 2024-02-29T19:30:15Z

I saw your PR for existing "automatic" rejection reasons: #1064

I've just submitted the implementation; explainer will need to wait till I get approval to land the above, and as for spec'ing the reject reason... I suppose I will need to patch private aggregation spec around here:
https://patcg-individual-drafts.github.io/private-aggregation-api/#protected-audience-api-monkey-patches:~:text=If%20signalBaseValue%20is%20%22bid%2Dreject%2Dreason%22%3A

... but we probably should have something since forDebuggingOnly keywords can use it, too, and that's in our spec entirely.

qingxinwu · 2024-03-08T15:50:38Z

Oh, adding a reject reason is a good idea... though not sure what's the story on spec'ing of that.

The spec does not have post auction signals yet (so no reject reason). It's planned but just need to find some time to add that. Just created an issue #1079 for that. If you'd like, you can add a todo and link to the issue where you'd like to add the reject reason.

... This is a prereq to specing how it works with targetNumAdComponents.

...into new type =bid debug reporting info=. This is in preparation for multibid, since that would have multiple bids corresponding to a single struct of debug reporting URLs [1]. As a bonus, this permits the removal of slightly awkward "no bid" field on generated bid. [1] So =bid debug reporting info= will ultimately have a /set/ of generated bid ids, rather than id-or-null.

morlovich · 2024-03-14T16:50:33Z

Note: this is now based on #1080 and morlovich#1 so maybe hard to read before that lands.

dmdabbs · 2024-03-18T15:53:24Z

The #1088 discussion noted that browserSignals.renderSize provided to scoreAd as currently described is a single-bid remnant. Size metadata provided by a buyer on a bid is not yet spec'd. Do you think you will simply provide it via the generatedBid(s) and dispense with that browserSignal attribute?

morlovich · 2024-03-18T16:03:34Z

scoreAd is called one at a time even with multi-bid, so I think that issue is orthogonal? I honestly don't know what's up, though it does look like we indeed don't set it, unlike what the explainer says.

dmdabbs · 2024-03-18T16:07:40Z

scoreAd is called one at a time even with multi-bid, so I think that issue is orthogonal?

Ah, ok. Yes, it would be orthogonal. GTK.

...into new type =bid debug reporting info=. This is in preparation for multibid, since that would have multiple bids corresponding to a single struct of debug reporting URLs [1]. As a bonus, this permits the removal of slightly awkward "no bid" field on generated bid. [1] So =bid debug reporting info= will ultimately have a /set/ of generated bid ids, rather than id-or-null.

qingxinwu

Some initial comments (mostly nits for now), still reviewing the logic.
Many steps don't have an ending ".", please add those.

spec.bs

morlovich

Feedback applied...

spec.bs

qingxinwu · 2024-04-10T17:47:23Z

spec.bs

@@ -5326,7 +5382,7 @@ result of [=evaluating a bidding script=], or an [=additional bid=] provided by

 <dl dfn-for="generated bid">
  : <dfn>id</dfn>
-  :: An {{unsigned long}}. Used to identify a [=generated bid=].
+  :: A pair of {{unsigned long}}s. Used to identify a [=generated bid=].


It'll be useful to add a brief explaination about what each element of the pair is about

maybe use [=tuple=]? A [=tuple=] consisting of two {{unsigned longs}}, one for xxx and one for xxx
See the example in https://infra.spec.whatwg.org/#tuples.

spec.bs

morlovich

Hmm, I think something is broken in compiling it (it complains about not finding chrome?), so I might need to do something to poke it to get it to try again...

spec.bs

qingxinwu · 2024-04-11T15:14:21Z

Hmm, I think something is broken in compiling it (it complains about not finding chrome?), so I might need to do something to poke it to get it to try again...

Yeah I got the same error on my local main branch check as well after I synced.

qingxinwu · 2024-04-11T21:15:01Z

@domfarolino I think it's ready for you to review. Thanks!

domfarolino · 2024-04-12T13:23:51Z

It's a pretty big PR! Can you please update the OP way from:

Another one that's very much in progress and not ready for review --- it's completely missing the limit stuff

... to something more descriptive that can help get up to speed.

morlovich · 2024-04-12T14:37:31Z

Description updated

This reverts commit 850b2ff.

SHA: 850b2ff Reason: push, by JensenPaul Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

JensenPaul · 2024-04-16T01:33:04Z

Drats, I accidentally merged this PR when I meant to merge #1048. I reverted this PR, but I'm not sure how to reopen this one :/

SHA: 9ffcb43 Reason: push, by JensenPaul Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

domfarolino · 2024-04-16T01:42:48Z

You cannot; you'll have to open it again fresh.

JensenPaul · 2024-04-16T11:48:05Z

Probably easiest to hit the revert button on my #1137 to essentially reopen a similar PR. Again, I apologize for merging the wrong PR.

morlovich · 2024-04-16T13:44:23Z

Created #1138 using sophisticated high-reliability technology (diff/patch)

morlovich changed the title ~~Start on multibid spec, missing limit stuff for now.~~ Start on multibid spec Feb 23, 2024

qingxinwu added the spec Relates to the spec label Feb 23, 2024

Maks Orlovich added 7 commits March 8, 2024 13:11

Explicitly spec out how non-k-anon auction works.

7148d90

... This is a prereq to specing how it works with targetNumAdComponents.

Fix various formatting mishaps

740a93d

Some easy parts of feedback

5b6b0c1

non-k-anon => non-k-anon-enforced

c5404fc

Link update algo

e7263f9

Start on multibid spec.

8a9e916

morlovich force-pushed the spec-multi-bid branch from 3b9c625 to 8a9e916 Compare March 14, 2024 13:24

Maks Orlovich added 2 commits March 14, 2024 12:33

Merge branch 'refs/heads/spec-split-reporting-info' into spec-multi-bid

bef6463

Actually adjust algorithms for reporting for multibid.

ab8d8e9

Maks Orlovich added 3 commits March 14, 2024 14:05

Unbreak logic around collecting bids + clones

ddd778f

Merge branch 'main' into spec-k-anon

1030d77

Don't double-count duplicates.

9340c7d

Maks Orlovich added 5 commits March 18, 2024 13:10

Merge branch 'spec-k-anon' into spec-split-reporting-info

db5abd0

Merge branch 'spec-split-reporting-info' into spec-multi-bid

7f0014b

Merge branch 'refs/heads/spec-split-reporting-info' into spec-multi-bid

54dc765

IDL.

007cf6f

morlovich changed the title ~~Start on multibid spec~~ Spec multibid Apr 4, 2024

Maks Orlovich added 2 commits April 4, 2024 09:03

Cosmetics from proofread

89bf79d

Fix merge mixup

b218d14

morlovich requested a review from qingxinwu April 8, 2024 14:07

qingxinwu reviewed Apr 9, 2024

View reviewed changes

Apply feedback

837a6f9

morlovich commented Apr 10, 2024

View reviewed changes

qingxinwu reviewed Apr 10, 2024

View reviewed changes

Some more feedback

809b34b

Liggbjjk approved these changes Apr 10, 2024

View reviewed changes

qingxinwu reviewed Apr 10, 2024

View reviewed changes

spec.bs Outdated Show resolved Hide resolved

spec.bs Outdated Show resolved Hide resolved

More suggestions applied

78f03c5

morlovich commented Apr 11, 2024

View reviewed changes

qingxinwu requested a review from domfarolino April 11, 2024 21:14

JensenPaul merged commit 850b2ff into WICG:main Apr 16, 2024
1 check passed

JensenPaul added a commit that referenced this pull request Apr 16, 2024

Revert "Spec multibid (#1055)"

aaba74d

This reverts commit 850b2ff.

JensenPaul mentioned this pull request Apr 16, 2024

Revert "Spec multibid" #1137

Merged

JensenPaul added a commit that referenced this pull request Apr 16, 2024

Revert "Spec multibid (#1055)" (#1137)

9ffcb43

This reverts commit 850b2ff.

github-actions bot added a commit that referenced this pull request Apr 16, 2024

Spec multibid (#1055)

79d92df

SHA: 850b2ff Reason: push, by JensenPaul Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

github-actions bot added a commit that referenced this pull request Apr 16, 2024

Revert "Spec multibid (#1055)" (#1137)

30e66ba

SHA: 9ffcb43 Reason: push, by JensenPaul Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

morlovich mentioned this pull request Apr 17, 2024

Spec Multibid (reland) #1138

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Spec multibid #1055

Spec multibid #1055

morlovich commented Feb 23, 2024 •

edited by pr-preview bot

Loading

dmdabbs commented Feb 26, 2024

dmdabbs commented Feb 27, 2024

morlovich commented Feb 28, 2024

dmdabbs commented Feb 28, 2024 •

edited

Loading

morlovich commented Feb 29, 2024

qingxinwu commented Mar 8, 2024

morlovich commented Mar 14, 2024

dmdabbs commented Mar 18, 2024

morlovich commented Mar 18, 2024

dmdabbs commented Mar 18, 2024

qingxinwu left a comment •

edited

Loading

morlovich left a comment

qingxinwu Apr 10, 2024

qingxinwu Apr 10, 2024

morlovich left a comment

qingxinwu commented Apr 11, 2024

qingxinwu commented Apr 11, 2024

domfarolino commented Apr 12, 2024

morlovich commented Apr 12, 2024

JensenPaul commented Apr 16, 2024

domfarolino commented Apr 16, 2024

JensenPaul commented Apr 16, 2024

morlovich commented Apr 16, 2024

Spec multibid #1055

Spec multibid #1055

Conversation

morlovich commented Feb 23, 2024 • edited by pr-preview bot Loading

dmdabbs commented Feb 26, 2024

dmdabbs commented Feb 27, 2024

morlovich commented Feb 28, 2024

dmdabbs commented Feb 28, 2024 • edited Loading

morlovich commented Feb 29, 2024

qingxinwu commented Mar 8, 2024

morlovich commented Mar 14, 2024

dmdabbs commented Mar 18, 2024

morlovich commented Mar 18, 2024

dmdabbs commented Mar 18, 2024

qingxinwu left a comment • edited Loading

Choose a reason for hiding this comment

morlovich left a comment

Choose a reason for hiding this comment

qingxinwu Apr 10, 2024

Choose a reason for hiding this comment

qingxinwu Apr 10, 2024

Choose a reason for hiding this comment

morlovich left a comment

Choose a reason for hiding this comment

qingxinwu commented Apr 11, 2024

qingxinwu commented Apr 11, 2024

domfarolino commented Apr 12, 2024

morlovich commented Apr 12, 2024

JensenPaul commented Apr 16, 2024

domfarolino commented Apr 16, 2024

JensenPaul commented Apr 16, 2024

morlovich commented Apr 16, 2024

morlovich commented Feb 23, 2024 •

edited by pr-preview bot

Loading

dmdabbs commented Feb 28, 2024 •

edited

Loading

qingxinwu left a comment •

edited

Loading