On Wed, Dec 19, 2012 at 02:57:09PM -0800, James Paige wrote: > On Wed, Dec 19, 2012 at 02:36:40PM -0800, [email protected] > wrote: > > james > > 2012-12-19 14:36:40 -0800 (Wed, 19 Dec 2012) > > 387 > > Various cleanup in for adding a reload node to HeroDef > > Convert subs.bas to subs.rbas > > Add a global reload doc for transient in-memory reload nodes > > Add leaked-node-check for the global reload doc > > Add a destructor to the HeroDef type > > Do a little Forward Pointer Voodoo to avoid including reload.bi in udts.bi > > (which would not have worked anyway because of some namespace-related > > troubles) > > Making a destructor seems nice and easy, but I am pretty unhappy with > how I have to declare the NodePtr member of the HeroDef type. > > I can't do: > > #INCLUDE "reload.bi" > > TYPE whatever > .foo as reload.NodePtr > END TYPE > > because it doesn't like the "reload." > > I can't do: > > #INCLUDE "reload.bi" > USING reload > > TYPE whatever > .foo as NodePtr > END TYPE > > because it doesn't seem to allow USING in udts.bi > > The workaround I have is to add in reload.bi > > TYPE reload_Node as Node > > And then in udts.bi I do: > > TYPE NodeFwd as reload_Node > > TYPE whatever > .foo as NodeFwd ptr > END TYPE > > And it works, but Then the compiler keeps complaining: > warning 3(1): Passing different pointer types, at parameter 1 of FREENODE() > > and I feel like there has to be a better way :P > > --- > James
Examining this a little further, it seems that if i do this the nice simple way with no forward pointer nonsense, it doesn't choke until it gets to the first .rbas file but looking at savegame.rbas.bas has not elicidated why this might be :( --- James _______________________________________________ Ohrrpgce mailing list [email protected] http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
