TPAC 2023 status report

[inexorabletash]

See also:

• Specification status report for TPAC 2020 #338
• Status report and planning for TPAC 2021 #364
• TPAC 2022 status report #389

Progress in the last 12 months

In the spec itself, only editorial changes have been made in the last 12 months. These include:

• Editorial: Link fixes/simplifications #395
• Editorial: Remove redundant ticks from <dfn event> text #398
• Adjust links to ECMAScript specifications #401
• Disambiguate "running" for transactions #409
• Editorial: Update note markup/markdown and styling  #404
• Editorial: Move Ali to Former Editor, note affiliation #407
• Rename "open a database" algorithm to "open a database connection" #411
• ... plus a handful of Bikeshed syntax updates, typos/grammar fixes, etc.

The set of new APIs (relative to 2.0) remains unchanged:

• databases() - implemented in Chromium and WebKit, but not Gecko
• durability - implemented in Chromium and WebKit, but not Gecko
• request - implemented in Chromium, WebKit and Gecko
• commit() - implemented in Chromium, WebKit and Gecko

No new PRs that haven't merged.

New unresolved issues since last TPAC:

• Could we drop the spec versioning? #402 - process
• Consider adding a simple(ish) query engine #403 - feature request; a great idea, but unlikely w/o strong consensus with both library authors and implementers
• Specify behavior when document is not fully active #412 - unspecified behavior

Moving to CR/Rec

Relevant:

• Could we drop the spec versioning? #402

See comments there, namely: we'll need to be more rigorous about getting consensus before landing any normative changes. But the change rate is low.

I believe these are the highest priority issues to tackle:

• Normative changes/corrections:
  • Specify behavior when document is not fully active #412 - some discussion, needs PR, tests, and implementer commitment to align
  • How should IDB connections and outstanding IDB transactions behave when a page enters back/forward cache? #381 - some discussion, needs PR, tests, and surveying implementations to see if we're aligned
  • Clarify the handling of +0 and -0 in keys #375 - has a PR and tests, just needs implementer commitment to align
• Cross-spec integration
  • Specifying "indexedDB" storage bottle deletion when data is evicted by quota #400
  • Is it really true that StructuredDeserialize can't throw in this spec? #391
  • Define Indexed DB as a storage endpoint, use hooks #334

There are plenty of other ways the spec could be made more rigorous tracked as issues, but they would just be editorial. PRs welcome!

What is your plan to unblock it and do you need any help?

Bit of a broken record here, but additional editors would be helpful. @aliams stepped down as editor leaving just @inexorabletash. Reviews for editorial-type changes can get done by the community, but actively pushing workstreams forward would strongly benefit from other implementers stepping up to help as editors.

cc: @marcoscaceres @LJWatson @siusin

[inexorabletash]

Note that I won't be attending TPAC 2023 in person, and the timing is pretty bad for the spec status reports session, so I'm not planning to participate live. I'm not expecting much discussion about IndexedDB, but thanks in advance for capturing any conversation in the minutes.

Extra big thanks to all the people that volunteer to be editors. 😉