Add data point for IndexedDB improved error reporting for large value read failures

[chrisdavidmills]

Summary

Chrome has updated its error reporting for large value read failures with transient causes such as low memory and unrecoverable causes such as source blob files being deleted.

Chrome 130 updates the DOMException types to be more appropriate, and Chrome 132 further updates them and improves the associated error messages.

See https://chromestatus.com/feature/5140210640486400 for the data source.

Test results and supporting details

Related issues

[caugner]

Can you point me to the part of the spec that mentions this, and/or the spec PR that added this behavior to the spec?

[chrisdavidmills]

@caugner I'm pretty sure this is the one: w3c/IndexedDB#430.

The exceptions are listed in the spec at https://www.w3.org/TR/IndexedDB/#exceptions.

[caugner]

Based on w3c/IndexedDB#423, I would suggest to rename the feature as follows:

Suggested change

	"improved_reporting_large_value_read_failures": {
	"__compat": {
	"description": "Improved reporting for large value read failures",
	"throws_NotReadableError": {
	"__compat": {
	"description": "Throws `NotReadableError` on storage read failures",

My impression is that "large value" is Chrome terminology, not part of the spec.

/cc @Elchi3 for a second opinion, as we may see more of these very specific behavioral features in the future.

[caugner]

PS: We could add this spec_url: https://www.w3.org/TR/IndexedDB/#:~:text=NotReadableError

[Elchi3]

I like @caugner's description and ID a lot more but I also find it strange to use both, sub features and array support statements here. Is this the summary?

"chrome": [
  {
   "version_added": "132",
   "notes": "Throws `NotReadableError` and `UnknownError`"
  }, 
 {
   "version_added": "130",
   "notes": "Throws `NotFoundError` and `DataError`"
  }, 
 {
   "version_added": "48",
   "notes": "Throws `DOMExceptions`"
  },
  {
   "version_added": "23",
   "notes": "Throws `DOMError`"
  }
],

[Elchi3]

From reading the content discussion, I think "Transient and unrecoverable read errors" (transient_unrecoverable_read_errors) is a good feature name and id.

"Improved reporting for read errors" and then only setting Chrome to support and other browsers to false seems wrong given this is about Chrome catching up and not providing "improvements" here, right? If the other browser support it, then we should say so.

[chrisdavidmills]

I've updated the feature name as suggested.

I agree it would be nice to include other browser support information, but I've got no idea how to find that out. I can't find a mention of NotReadableError in the Gecko source code so, although they do support the same kind of error handling for read errors, I'm not sure if it is using the same exception names.

From the spec bug, I am wondering whether this has bene agreed as the error to put in the spec, which matches Fx behavior, but Fx hasn't actually implemented the official specified error yet. In which case, could we call it something like "Specified reporting for read errors", and it would be accurate to say that only Chrome has implemented it so far?

[caugner]

I'm not 100% sure, but most likely it was Firefox 14 that implemented NotReadableError in https://bugzilla.mozilla.org/show_bug.cgi?id=730161, cf. mozilla/gecko-dev@1ce6cdf#diff-38a23a8920ed3a6e05068b536b2ba2f20b2ef6449a7a084ee0bc76faeca8e6d5R45

[caugner]

Webkit most likely implemented it in https://bugs.webkit.org/show_bug.cgi?id=102137 / WebKit/WebKit@b368491, i.e. Safari 537.43.1 before Safari 7 (537.71).

[chrisdavidmills]

That sounds good enough for me ;-)

[caugner]

Based on w3c/IndexedDB#423, I would suggest to rename the feature as follows:

Suggested change

	"improved_reporting_large_value_read_failures": {
	"__compat": {
	"description": "Improved reporting for large value read failures",
	"throws_NotReadableError": {
	"__compat": {
	"description": "Throws `NotReadableError` on storage read failures",

My impression is that "large value" is Chrome terminology, not part of the spec.

/cc @Elchi3 for a second opinion, as we may see more of these very specific behavioral features in the future.

[caugner]

Note that the NotFoundError has been in the spec since at least 2011 (before it was referred to as NOT_FOUND_ERR).

[chrisdavidmills]

That's true. It is only mentioned here because Chrome reported that error type for read errors for a couple of versions (incorrectly, I assume), before it was then updated to NotReadableError.

[chrisdavidmills]

@caugner @Elchi3 I had to make some further updates due to linting failures. I noticed that the new data point had been put at the wrong nesting level (oops!), and I had to adjust the support versions for FxA and Saf because the specified values were earlier than the support values for the parent feature.