Closed Bug 12326 Opened 20 years ago Closed 20 years ago

Mork XPCOM factory hare-brained

Categories

(MailNews Core :: Database, defect, P3)

All
Windows NT
defect

Tracking

(Not tracked)

VERIFIED FIXED

People

(Reporter: Bienvenu, Assigned: Bienvenu)

Details

Filing as a reminder to myself to fix the Mork factory, which only works now
because of casting...
Status: NEW → ASSIGNED
Target Milestone: M11
marking m11. I'll wait until the various mork shuffles are done before I do
this, since I'll need to change the consumers as well.
Blocks: 11091
(target milestone is M11 or M12 - add to mail beta tracking bug)
Target Milestone: M11 → M15
Not required for B1
No longer blocks: 11091
Status: ASSIGNED → RESOLVED
Closed: 20 years ago
Resolution: --- → FIXED
fixed a few days ago - Chris, perhaps you could mark this verified?
Status: RESOLVED → VERIFIED
Ok, I'll mark it verified on one condition: you come down and go over some of
the initialization code with me. Poritions of it look like it is going to leak
all over the floor.
sorry, should've been more clear. the _global history_ usage of MDB.
I guess we could be leaking the nsIMdbFactory factory. I'll try adding a
CutStrongRef for that (not that that'll cause it to really get freed yet). Was
there anything else?
well, there's not really a leak, because the factory is really global (and not
meant to be ref-counted) but I'll try to make it play nice.
I have a bunch of changes in my tree, reverting some stuff that davidmc did.
e.g., what is this "oldFileAnchor" business in OpenDB()? anyway, if you could
pop by when you're in and code review them, i'd be much obliged...
David B and I went by to review; the code I added to refcount the new nsIMdbFile
class (which didn't exist before) didn't follow typical style appearance, but
Chris's changes tweak things to match other history usage.
Product: MailNews → Core
Product: Core → MailNews Core
You need to log in before you can comment on or make changes to this bug.