Data backing#302
Conversation
|
Hi @johankool, thanks for exploring this. We will chat about this approach, but by far the most dangerous part as it stands right now is that if you switch a column from |
|
@johankool We chatted about this today and still think the automatic approach mentioned in #281 is the way to go. We could start with an arbitrary, hard-coded threshold (10kb? 16?) and then could explore further configuration/behavior in the future. This approach would hopefully minimize the additional API surface area and folks could benefit with minimal configuration. Are you open to updating this PR with this alternate approach? |
c03e962 to
57dfa6a
Compare
|
Still needs some more testing, both writing unit tests and actually trying it out in a real app. |
ef27e1e to
7556ea9
Compare
| ┌──────────────────────────────────┐ | ||
| │ RemindersListAsset( │ | ||
| │ remindersListID: 1, │ | ||
| │ coverImage: Data(16,811 bytes) │ |
There was a problem hiding this comment.
Need to fix this in CustomDump: on machines with a thousands separator that is not a comma this test fails.
There was a problem hiding this comment.
A fix is proposed here: pointfreeco/swift-custom-dump#163
3 participants
Regarding this message on Slack and this discussion: could this be an approach you would consider to be able to customize wether a column is stored as
Datain CloudKit or as aCKAsset?