LGTM1
On Wed, Oct 28, 2026 at 11:20 PM 'Monica Chintala' via blink-dev <
[email protected]> wrote:
> Thanks, Alex. We cannot measure the exact intersection, but existing
> UseCounters provide an upper bound. Access to WebTransport.datagrams
> peaked at 0.0126% of page loads over the last 30 days. For the common
> getReader({mode: "byob"}) pattern, the upper bound is 0.00232%, fewer than
> 1 in 43,000 page loads.
>
> The actual intersection may be lower because the BYOB counter covers all
> APIs. Affected applications can preserve the current behavior with
> datagramsReadableType: "bytes", and the change remains Finch-controlled
> for rollback.
>
> On Wednesday, October 28, 2026 at 8:32:01 AM UTC-7 Alex Russell wrote:
>
>> Is there an upper bound for potential breakage? Do we have metrics for
>> total use of datagram readable streams w/ BYOB behaviours?
>>
>> On Wednesday, October 28, 2026 at 7:35:15 AM UTC-7 Monica Chintala wrote:
>>
>>> We don’t have a metric for this exact usage pattern. The affected case
>>> is limited to applications requesting a BYOB reader without opting into
>>> "bytes".
>>>
>>> Default readers continue to work unchanged. Applications using BYOB can
>>> preserve the existing Chromium behavior by adding: {datagramsReadableType:
>>> "bytes"}
>>>
>>> This option can be added before rollout and works with both the old and
>>> new behavior. Firefox and Safari stable already use the new default. The
>>> rollout is Finch-controlled, so we can restore the previous default without
>>> a new release if compatibility problems are reported.
>>>
>>> On Tuesday, October 27, 2026 at 12:02:26 PM UTC-7 Alex Russell wrote:
>>>
>>>> What's the potential for breakage, given that we're changing default
>>>> behavior? Do our histograms tell us the fraction of pageloads that would be
>>>> affected?
>>>>
>>>> Best,
>>>>
>>>> Alex
>>>>
>>>> On Wednesday, October 21, 2026 at 1:52:53 PM UTC-7 Chromestatus wrote:
>>>>
>>> *Contact emails*
>>>>> [email protected], [email protected]
>>>>
>>>>
>>>>>
>>>>> *Explainer*
>>>>> https://github.com/w3c/webtransport/blob/main/explainer.md
>>>>>
>>>>> *Specification*
>>>>>
>>>>> https://www.w3.org/TR/webtransport/#dom-webtransportoptions-datagramsreadabletype
>>>>>
>>>>> *Summary*
>>>>> Adds the WebTransportOptions.datagramsReadableType constructor
>>>>> option. By default, incoming WebTransport Datagrams are exposed through a
>>>>> default ReadableStream . Applications that require BYOB reading can
>>>>> request a readable byte stream by setting datagramsReadableType: "bytes"
>>>>> .
>>>>>
>>>>> *Blink component*
>>>>> Blink>Network>WebTransport
>>>>> <https://issues.chromium.org/issues?q=customfield1222907:%22Blink%3ENetwork%3EWebTransport%22>
>>>>>
>>>>> *Web Feature ID*
>>>>> webtransport <https://webstatus.dev/features/webtransport>
>>>>>
>>>>> *Motivation*
>>>>> WebTransport Datagrams are discrete messages rather than an
>>>>> undifferentiated byte sequence. A readable byte stream can lose empty
>>>>> Datagrams and can lose message boundaries when BYOB reads use a minimum
>>>>> byte count spanning multiple Datagrams. Making a default ReadableStream
>>>>> the default preserves Datagram boundaries and empty Datagrams.
>>>>> Applications
>>>>> that benefit from BYOB allocation control can explicitly opt into the
>>>>> byte-stream behavior.
>>>>>
>>>>> *Initial public proposal*
>>>>> *No information provided*
>>>>>
>>>>> *TAG review*
>>>>> Not applicable — this is a small incremental addition within the
>>>>> existing WebTransport specification.
>>>>>
>>>>> *TAG review status*
>>>>> Not applicable
>>>>>
>>>>> *Goals for experimentation*
>>>>> None
>>>>>
>>>>> *Risks*
>>>>>
>>>>>
>>>>> *Interoperability and Compatibility*
>>>>> Interoperability risk is low to medium. The API and default-stream
>>>>> behavior are defined in the WebTransport specification and covered by WPT.
>>>>> Firefox already uses a default non-byte stream when the option is omitted,
>>>>> but does not yet support the explicit "bytes" mode in current public
>>>>> test
>>>>> runs. Compatibility risk is mitigated by runtime gating. Chromium
>>>>> historically exposed the Datagram readable as a byte stream, so enabling
>>>>> the feature changes the default behavior: existing applications using a
>>>>> BYOB reader without specifying datagramsReadableType: "bytes" would fail.
>>>>> Applications using ordinary readers continue to receive Uint8Array
>>>>> chunks
>>>>>
>>>>> *Gecko*: Positive (
>>>>> https://bugzilla.mozilla.org/show_bug.cgi?id=2007755) Firefox is
>>>>> implementing datagramsReadableType: "bytes" and BYOB Datagram reads
>>>>>
>>>>> *WebKit*: In development (
>>>>> https://bugs.webkit.org/show_bug.cgi?id=319830) WebKit main
>>>>> implements the standardized datagramsReadableType option
>>>>>
>>>>> *Web developers*: Positive (
>>>>> https://github.com/web-platform-tests/interop/issues/1121) The
>>>>> Interop 2026 WebTransport proposal received 28 👍 reactions. Developers
>>>>> also reported active WebTransport use and significant cross-engine
>>>>> differences.
>>>>>
>>>>> *Other signals*:
>>>>>
>>>>> *Ergonomics*
>>>>> Low. The default stream better matches Datagram message semantics.
>>>>> Applications needing BYOB allocation control must explicitly opt into
>>>>> "bytes" , making the semantic tradeoff visible in application code.
>>>>>
>>>>> *Activation*
>>>>> The main activation risk is existing Chromium-specific code assuming
>>>>> that transport.datagrams.readable always supports BYOB readers. Such
>>>>> applications must explicitly request datagramsReadableType: "bytes" . The
>>>>> feature remains experimentally gated while compatibility and
>>>>> interoperability are evaluated.
>>>>>
>>>>> *Security*
>>>>> Low. The option changes only the JavaScript stream adapter used to
>>>>> expose already-received Datagrams. It does not provide additional network
>>>>> access, expose new information, or alter WebTransport’s secure-context and
>>>>> connection security requirements.
>>>>>
>>>>> *WebView application risks*
>>>>>
>>>>> Does this intent deprecate or change behavior of existing APIs, such
>>>>> that it has potentially high risk for Android WebView-based applications?
>>>>> Low. The implementation is cross-platform Blink code and requires no
>>>>> WebView-specific integration. The primary compatibility consideration is
>>>>> the same as Chrome: applications using a BYOB reader must explicitly
>>>>> request datagramsReadableType: "bytes" .
>>>>>
>>>>>
>>>>> *Debuggability*
>>>>> No feature-specific DevTools UI is required. Applications can inspect
>>>>> whether datagrams.readable supports BYOB by attempting to acquire a BYOB
>>>>> reader. Existing WebTransport logging, NetLog, and DevTools network
>>>>> diagnostics remain available for connection and Datagram issues.
>>>>>
>>>>> *Will this feature be supported on all six Blink platforms (Windows,
>>>>> Mac, Linux, ChromeOS, Android, and Android WebView)?*
>>>>> Yes
>>>>> Yes. Implemented in cross-platform Blink code without
>>>>> platform-specific dependencies.
>>>>>
>>>>> *Is this feature fully tested by web-platform-tests
>>>>> <https://chromium.googlesource.com/chromium/src/+/main/docs/testing/web_platform_tests.md>?*
>>>>> Yes
>>>>>
>>>>> https://wpt.fyi/results/webtransport/datagrams.https.any.html?label=experimental&label=master&aligned
>>>>>
>>>>> *Flag name on about://flags*
>>>>> chrome://flags/#enable-experimental-web-platform-features
>>>>>
>>>>> *Finch feature name*
>>>>> WebTransportDatagramsReadableType
>>>>>
>>>>> *Rollout plan*
>>>>> Will ship enabled for all users
>>>>>
>>>>> *Requires code in //chrome?*
>>>>> False
>>>>>
>>>>> *Tracking bug*
>>>>> https://issues.chromium.org/issues/547572475
>>>>>
>>>>> *Availability expectation*
>>>>> WebKit has implemented the option and corresponding default-stream
>>>>> behavior in its main branch. Firefox implements the default non-byte
>>>>> stream
>>>>> behavior, although explicit "bytes" /BYOB mode does not yet pass current
>>>>> public WPT runs. Broad availability is expected as implementations
>>>>> converge
>>>>> on the WebTransport specification.
>>>>>
>>>>> *Adoption expectation*
>>>>> Adoption is expected mainly among WebTransport applications that use
>>>>> BYOB readers or need reliable preservation of Datagram message boundaries.
>>>>> Most applications using ordinary readers require no code changes.
>>>>>
>>>>> *Adoption plan*
>>>>> Adoption is expected through WebTransport documentation and Interop
>>>>> 2026 visibility. Existing Chromium applications using BYOB readers will
>>>>> need to explicitly set datagramsReadableType: "bytes" when the feature
>>>>> is
>>>>> enabled.
>>>>>
>>>>> *Estimated milestones*
>>>>> Shipping on desktop 159
>>>>> DevTrial on desktop 155
>>>>> Shipping on Android 159
>>>>> DevTrial on Android 155
>>>>> Shipping on WebView 159
>>>>>
>>>>> *Anticipated spec changes*
>>>>>
>>>>> Open questions about a feature may be a source of future web compat or
>>>>> interop issues. Please list open issues (e.g. links to known github issues
>>>>> in the project for the feature specification) whose resolution may
>>>>> introduce web compat/interop risk (e.g., changing to naming or structure
>>>>> of
>>>>> the API in a non-backward-compatible way).
>>>>> None currently anticipated. The option and default-stream behavior are
>>>>> already defined in the WebTransport specification.
>>>>>
>>>>> *Link to entry on the Chrome Platform Status*
>>>>> https://chromestatus.com/feature/5148774971342848?gate=5517621058469888
>>>>>
>>>>> *Links to previous Intent discussions*
>>>>> Intent to Prototype:
>>>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/6aa86417.edf999b6.71d4e.0a07.GAE%40google.com
>>>>>
>>>>>
>>>>> This intent message was generated by Chrome Platform Status
>>>>> <https://chromestatus.com>.
>>>>>
>>>> --
> You received this message because you are subscribed to the Google Groups
> "blink-dev" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To view this discussion visit
> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/2684f886-eeec-4900-b634-8cf8d497ca1an%40chromium.org
> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/2684f886-eeec-4900-b634-8cf8d497ca1an%40chromium.org?utm_medium=email&utm_source=footer>
> .
>
--
You received this message because you are subscribed to the Google Groups
"blink-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion visit
https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAOMQ%2Bw_HHFqMLJKZfK%2BnUqOpz0bA4MUDm4TQ9o7e3oaYh-vsgw%40mail.gmail.com.
