Skip to content

Audit undeclared marker fields #5281

@mstange

Description

@mstange

Profiles with lots of markers are too big.

A lot of space in the stringified JSON is taken up by the field keys in each marker's data object. Example:
{ ​​"type": "Task", "name": "Document::NotifyIntersectionObservers", "priority": 4, "priorityName": "Normal", "task": 3944 }

It would be nice if we didn't have to repeat those string keys ("name", "priority" etc) for every marker.

#4494 proposes moving the "type": "Task" part outside the data object.

Once that's done, we could change the data object into a data array:
["Document::NotifyIntersectionObservers", 4, "Normal", 3944]

In order to know which value belongs to which field, we would check the marker schema.

However, this can only work if all fields are declared in the schema.

What are the cases in which markers have fields which are not declared in the schema? I'll use this issue to research and write down the answer to this question.

┆Issue is synchronized with this Jira Task

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions