marmoute
marmoute
Could we get to a middle ground where nextcloud detect this is 2.1 and issue an appropriate error message. The current error reporting is quite opaque. With an accurate message...
> What we **will/can** do > > * Add proper warning for 2.1 vcards Yes please !
For the record, here is the result of the "simplified" example: ``` all copies are linear hashmap_linear: 5.492893306s persistant_map_linear: 5.271898ms copies on both side of the merge hashmap_merge_both: 3.977184043s persistant_map_merge_both:...
Let me dump my brain about the issue here. **About the Hashmap** If I understood the data structure right, it is defined as follow. - The HashMap use a root...
Status update: - I have a prototype, it gives good result on the example usage I attached. - My prototype needs a proper rewrite since I discovered various things. -...
… and last point, If I understand your profile right, you are London based. We will be in London from Thursday Nov 28th nigh to Tuesday Dec 3rd, it might...
Someone pointed me at OrdMap.diff and I was able to build some efficient code using that. I'll focus on that first.
I've a "working" implementation using OrdMap relying on OrdMap::diff to speed things up. And a small patch to OrdMap::diff to be even faster. Howerever, on larger data set, I get...
Eventually using this code (combined with my open MR) I get the massive performance boost I was looking for (the map union get from 75% of total time to "entirely...
Not strictly a dupe, as mine is not narrowed to aarch64, but very similar yes.