[opensource-dev] Inventory Patch you should integrate

[Armin Weatherwax]

Am Wednesday 11 January 2012 10:07:55 schrieb Henri Beauchamp:
> Oops...
>
> There was a bug in the patch I sent for v1 viewers (I forgot to replace
> mFolderAdded(FALSE) with mFolderAdded(folder_added) in the constructor
> of LLInventoryCopyAndWearObserver).
>
> Attached to this email is the fixed patch.
>
> Henri.

Thank you for porting, Henri!

And yes, it would be very good to have somwhere to test in the betagrid, there 
are a couple of things I think it would good to have an eye on:

* in LLFloaterOpenObject::moveToInventory it looks to me that at least a 
LLCategoryCreate object gets leaked if a null category_id is returned, 
however all the 
	// If we get a null category ID, we are using a capability in
	// createNewCategory and we will handle the following in the
	// callbackCreateInventoryCategory routine.
	if (category_id.notNull())
seems to contradict whats actally happening in 
LLInventoryModel::createNewCategory: null id returned on 2 possible failures, 
non-null  in all other cases including success.
* LLInventoryModel::createNewCategory returns a non-null id, even though 
requesting the cap failed if :
 - one or both of callback or user_data == NULL (ok, that is both rather 
unlikely) 
 - the cap url is empty (e.g. the capability response didn't arrive (yet)).
   About that Im not so sure that it can not happen: buy new avatar, tp to 
busy sandbox on a busy weekend  and have a bad network route anyway, rezz - 
open - copy and wear directly after teleport ... not sure yet how to simulate 
that in the betagrid, but I want to see what happens before it does to users.

Armin