View Single Post
Old 06-22-2017, 03:13 PM   #14
chaley
Grand Sorcerer
chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.chaley ought to be getting tired of karma fortunes by now.
 
Posts: 12,450
Karma: 8012886
Join Date: Jan 2010
Location: Notts, England
Device: Kobo Libra 2
Quote:
Originally Posted by kovidgoyal View Post
@chaley: Yeah it does sound a lot like the wireless device The only question is to what extent do we allow two way data flow. Metadata updates from device->calibre? How does one handle conflicts -- all the usual distributed data reconciliation headaches...
The wireless device is in the end a calibre device. It doesn't do much more than the standard device interface, which as you know is mostly one way: calibre to the device. The plumbing is there to update local metadata using information from the device; it is currently used to sync a bool "read" column and a "date" date-read column. I haven't extended it any further for exactly the reason you suggest: conflicts. IMO that problem is close to irreconcilable without human intervention in the conflict reconciliation process, especially for is_multiple columns.
chaley is offline   Reply With Quote