234
edits
User:Sspitzer/GlobalFrecency: Difference between revisions
no edit summary
No edit summary |
No edit summary |
||
| Line 87: | Line 87: | ||
14) on delete visits, see nsNavHistory::RemovePage() and nsNavHistory::RemovePagesFromHost(), set frecency = -1 | 14) on delete visits, see nsNavHistory::RemovePage() and nsNavHistory::RemovePagesFromHost(), set frecency = -1 | ||
15) | |||
add frecency <> 0 to ac queries | |||
16) | |||
add code so that for all non place: and unvisisted children of livemarks we set frecency to 0, so that we can use it after migration, clear all private data, etc | |||
'''todo''' | '''todo''' | ||
| Line 98: | Line 106: | ||
x) | x) | ||
for visit count re-calculation, visit_type NOT in (0,4), is that correct? | |||
x) | x) | ||
add | // should we add "AND visit_type NOT IN (0,4)" | ||
// XXX aren't embed's affecting visit count already in ::AddVisit()? | |||
// I think we have a bug about not letting them do that. | |||
// XXX note, should we fix AddVisit() to recalc visit count by using | |||
// CalculateVisitCount()? | |||
x) | |||
// should we recalc frecency for all non-livemark item, non place: query | |||
// bookmark items? What if there are a lot? Better to let it happen on idle? | |||
// if we add a frecency > 0 check to autocomplete, nothing will show up | |||
// we could have visited a livemark item, cleared history, which marked it | |||
// as non hidden, then, until recalculation on idle, it wills how up | |||
// in autocomplete. we could recalc frecency for (or just make 0) | |||
// all place: uri and livemark items, and then do frecency <> 0 | |||
// and keep hidden <> 1. this way, we will show everything in ac | |||
// after clear all data that we've got, except for place: and livemark items | |||
x) | x) | ||
| Line 124: | Line 148: | ||
x) | x) | ||
do "on idle" work in "on visit"? | do forced "on idle" work in "on visit"? | ||
x) | x) | ||