You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Fractal is currently responsible for the pagination and serialisation of data, which is continually growing and clouding the core functionality of Fractal.
A lot of the issues and pull requests are to do with a specific serialiser, which means to change something on one serialiser requires a minor version bump which might update the package for people who aren't using the serialiser.
I believe we need to seperate the serialisers into their own self-contained packages:
fractal-jsonapi fractal-csv fractal-{package}
Utilising one of these packages should provide a Fractal manager using the specific serialiser.
The Fractal package would still provide the basic data-array serialiser, and an interface to implement another custom serialiser.
Also, the paginators are not within the scope of the core project, the interfaces would still be provided by the package and the current paginators would be extracted into their own packages.
By doing this and narrowing the scope of the Fractal core, we would create a more maintainable, concise package, focussed on managing the transformation of data. The issues and PR's for the serialisers and pagination would not be part of the Fractal core package, meaning that updating the JSON-API serialiser would not require people using the other serialisers to update the package. It would also enable people to create their own serialisation and pagination, and publish them in a recommended format. The unit tests for the packages and core would be smaller and easier to maintain too.
The text was updated successfully, but these errors were encountered:
This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 4 weeks if no further activity occurs. Thank you for your contributions.
Fractal is currently responsible for the pagination and serialisation of data, which is continually growing and clouding the core functionality of Fractal.
A lot of the issues and pull requests are to do with a specific serialiser, which means to change something on one serialiser requires a minor version bump which might update the package for people who aren't using the serialiser.
I believe we need to seperate the serialisers into their own self-contained packages:
fractal-jsonapi
fractal-csv
fractal-{package}
Utilising one of these packages should provide a Fractal manager using the specific serialiser.
The Fractal package would still provide the basic data-array serialiser, and an interface to implement another custom serialiser.
Also, the paginators are not within the scope of the core project, the interfaces would still be provided by the package and the current paginators would be extracted into their own packages.
By doing this and narrowing the scope of the Fractal core, we would create a more maintainable, concise package, focussed on managing the transformation of data. The issues and PR's for the serialisers and pagination would not be part of the Fractal core package, meaning that updating the JSON-API serialiser would not require people using the other serialisers to update the package. It would also enable people to create their own serialisation and pagination, and publish them in a recommended format. The unit tests for the packages and core would be smaller and easier to maintain too.
The text was updated successfully, but these errors were encountered: