-
Duplicate: #1787911.
This way since 7.6 AFAIU, hard to define proper semantics. Solution might be to only use [array names] on env.
2009-10-30 10:27:02 UTC in Tcl
-
The BC implementation has to mimic exactly the non-compiled implementation.
As [info exists] triggers read traces on the array and the traces could well create new elements, the effect on searches is inevitable (or at least very hard to avoid).
The comments at the start of TclVarTraceExists attempt an explanation:
/*
* The choice of "create" flag values is delicate here, and...
2009-10-30 10:15:54 UTC in Tcl
-
msofer committed patchset 11421 of module tcl to the Tcl CVS repository, changing 2 files.
2009-09-23 23:36:50 UTC in Tcl
-
I confirmed the leak reported by echm [http://groups.google.com/group/comp.lang.tcl/browse_thread/thread/a873bbc1ee78d776/b5ca31cdab5af4d6?lnk=gst&q=md5#b5ca31cdab5af4d6]
The following script consumes ever increasing amounts of memory:
% package require md5
2.0.7
% parray md5::accel
md5::accel(critcl) = 1
md5::accel(cryptkit) = 0
md5::accel(trf) = 0
% while 1 {md5::md5...
2009-08-30 16:46:58 UTC in tcllib
-
This cannot be fixed easily without deep changes in the exception range mechanism, or possibly some costly extra functionality in INST_START_CMD.
The problem is that the exec stack depth is not recorded on entering an exception range: neither at compile time, which would be iffy, nor at run time (catch does it with INST_BEGIN_CATCH).
Given this, there is no way to restore the stack when...
2009-08-21 21:42:07 UTC in Tcl
-
msofer committed patchset 11318 of module tcl to the Tcl CVS repository, changing 2 files.
2009-08-02 14:26:07 UTC in Tcl
-
8.6 seems to be miscounting recursion levels from [unknown]
proc history args {}
rename set xx
set x 1
enters a busy loop and never hits the recursion limit.
2009-07-21 23:47:10 UTC in Tcl
-
The 8.5 behaviour is correct - it is a consequence of your breaking [unknown] (which uses [set]) by not providing a reasonable implementation of [set]. This causes [unknown] to call itself recursively.
Not sure yet about what is happening with 8.6.
2009-07-21 21:24:36 UTC in Tcl
-
non-bcc'ed [switch] is not nre-enabled! Make it so ...
2009-07-14 15:37:51 UTC in Tcl
-
Coroutines are currently not NRE-enabled; the naive attempt to enable them segfaults on [yield].
The problem is that TEBC does not know how to:
(1) resume a coro in the same TEBC instance
(2) yield into the same instance
The problem seems to be in the management coordination of the different stacks on swapping the execEnv, especially 'bottomPtr' (?).
See also 2037611.
2009-06-15 04:00:10 UTC in Tcl
