Changes between Initial Version and Version 1 of Ticket #23381, comment 4
- Timestamp:
- 02/22/2013 01:15:56 AM (11 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Ticket #23381, comment 4
initial v1 2 2 > As per rboren on IRC, changed the patch to switch `add_option()` to match `update_option()`. Will require fixing APC to handle the specific use-case. 3 3 4 rmccue, could you explain therationale for switching add_option() to match update_option(), rather that the other way around. This would leave an inconsistent pattern of 'alloptions' being either a serialized string or an array depending on the existence of an object cache, like it it now with the r13673 defect. I am not liking the idea of having mixed variables with a context dependency added. Update_option() is the one that should match add_option() for consistency in my view. That's where it was broken in the first place.4 rmccue, could you explain rboren's rationale for switching add_option() to match update_option(), rather that the other way around. This would leave an inconsistent pattern of 'alloptions' being either a serialized string or an array depending on the existence of an object cache, like it it now with the r13673 defect. I am not liking the idea of having mixed variables with a context dependency added. Update_option() is the one that should match add_option() for consistency in my view. That's where it was broken in the first place.