On 01/11/2014 06:18 PM, Chet Ramey wrote:
You don't want the value generated in all cases. This has an effect on
some variables: RANDOM, for example. I believe the current behavior is
preferable.
I guess I don't understand why returning stale values for the dynamic variables
under any circum
On 1/11/14, 5:54 PM, John R. Graham wrote:
> On 01/11/2014 04:55 PM, Chet Ramey wrote:
>> On 1/10/14, 6:06 PM, John R. Graham wrote:
>>> Some of the automagically created special array variables (GROUPS and
>>> DIRSTACK for soer; perhaps others) appear to not be fully initialized.
>>> Their values
On 01/11/2014 04:55 PM, Chet Ramey wrote:
On 1/10/14, 6:06 PM, John R. Graham wrote:
Some of the automagically created special array variables (GROUPS and DIRSTACK
for soer; perhaps others) appear to not be fully initialized. Their values
don't appear correctly in some corner cases until they'
On 1/10/14, 6:06 PM, John R. Graham wrote:
> Some of the automagically created special array variables (GROUPS and
> DIRSTACK for soer; perhaps others) appear to not be fully initialized. Their
> values don't appear correctly in some corner cases until they've been
> referenced with parameter ex
Some of the automagically created special array variables (GROUPS and DIRSTACK
for soer; perhaps others) appear to not be fully initialized. Their values
don't appear correctly in some corner cases until they've been referenced with
parameter expansion (I think that's the triggering criteria, an
