Changes between Initial Version and Version 1 of Ticket #4888, comment 4


Ignore:
Timestamp:
May 19, 2012, 11:05:14 PM (10 years ago)
Author:
jordan
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #4888, comment 4

    initial v1  
    11x190, thanks for reporting this and for tracking down the broken code.
    22
    3 Regarding the proposed change to tr_cpLeftUntilComplete()... it feels a little wrong to add that logic. IMO the problem is that tr_cpLeftUntilComplete() is based on assumptions that aren't always right. This isn't necessarily a showstopper, but happily for us it looks like nobody else uses this function. So if we move the logic into announcer.c we can make the assumptions a little more explicit.
     3Regarding the proposed change to tr_cpLeftUntilComplete()... it feels a little wrong to add that logic. IMO the problem is that tr_cpLeftUntilComplete() is based on assumptions that aren't always right. This isn't necessarily a showstopper, but happily for us it looks like we can do better this time because nobody else uses this function. So we can move it from completion.h to be private inside announcer.c and make the assumptions more explicit.
    44
    55What do you think of magnet-left-until-complete.patch ?