Re: Tcl_ObjSetVar2 ...
- From: "miguel sofer" <miguel.sofer@xxxxxxxxx>
- Date: 22 Oct 2005 17:07:51 -0700
In light of Bug #1334947, I think I am reviewing my position. If at all
possible, we should try to:
- insure that "consumers" actually do consume the Tcl_Obj in all paths
(some fail to do so in case of errors), in the sense that the caller
may forget about the Tcl_Obj completely
- shave all hairy monsters
- improve the documentation
Even so, the rule stands: in case of doubt, incr before sending and
decr on return is always the safest thing to do when not calling
mutators.
.
- References:
- Tcl_ObjSetVar2 ...
- From: MartinLemburg@UGS
- Re: Tcl_ObjSetVar2 ...
- From: Kevin Kenny
- Re: Tcl_ObjSetVar2 ...
- From: Eckhard Lehmann
- Re: Tcl_ObjSetVar2 ...
- From: Joe English
- Re: Tcl_ObjSetVar2 ...
- From: Donal K. Fellows
- Tcl_ObjSetVar2 ...
- Prev by Date: Re: string first
- Next by Date: Re: MySQLTCL package and TCLApp issue: 'couldn't load library "[...]\Temp\TCL5E4.tmp"
- Previous by thread: Re: Tcl_ObjSetVar2 ...
- Next by thread: Re: Tcl_ObjSetVar2 ...
- Index(es):
Relevant Pages
|