13 | | A good docblock update for the different meta fields (post, comment, user, site, etc.) would include the full description about what's going on, and probably mention that the unexpected types can be avoided at the cost of a tiny bit more processing if the data is serialized. JSON encoding would also work in most cases, but may behave a bit unexpectedly when differentiating between objects and associative arrays. |
| 13 | A good docblock update for the different meta retrieval functions (post, comment, user, site, etc.) would include the full description about what's going on, and probably mention that the unexpected types can be avoided at the cost of a tiny bit more processing if the data is serialized. JSON encoding would also work in most cases, but may behave a bit unexpectedly when differentiating between objects and associative arrays. |