Re: status of local sh variables

[David A. Wheeler via austin-group-l at The Open Group]

> On Nov 3, 2024, at 1:40 PM, Christoph Anton Mitterer via austin-group-l at 
> The Open Group <austin-group-l@opengroup.org> wrote:
> 
> On Sun, 2024-11-03 at 18:34 +0000, Stephane Chazelas via austin-group-l
> at The Open Group wrote:
>> nothing I can think of sound really appealing.
> 
> I found the idea with options to `local`, where only these option make
> it truly portable intriguing.
> 
> Portability aside, it might allow to exactly control what should
> happen, like how the masking works of unset variables, etc..

Using "local" also:
* makes the final result more like existing practice. People *currently* use
  the keyword "local" for local variables.
* Avoids the problem that someone might use another keyword for a command.
  E.g., there's probably someone out there with functions or commands named 
"my" or "ours".
  However, it's widely expected that "local" is a shell built-in for local 
variables.
  Adding an option for portability on an existing use of "local" is easy to 
understand
  and doesn't risk interfering with anything else.

--- David A. Wheeler