Experimental Physics and
Industrial Control System

Benjamin Franksen <[email protected]> · Mon, 9 Apr 2018 15:11:32 +0200

On 03/15/2018 05:15 PM, Laznovsky Michael (PSI) wrote:
> Hi Ben- just a belated comment... I took a look at linenoise, and apparently it does 
> malloc/memmove/realloc (and expects the caller to free() each line of input)...is that 
> something you want in EPICS?  Hasn't that been a problem in the past in other areas?
> 
> Perhaps not a big deal for relatively low-traffic console input, but may be one of 
> those things that can creep up on you later, if eg. a large amount of input is piped 
> in or something.
AFAIU, when input is redirected (in or out), EPICS automatically
switches to a "dumb mode" that circumvents any EPICS_READLINE
functionality, however implemented.

Anyway, it turned out that using linenoise has another disadvantage, at
least with the way we set up our IOCs: The log files stored by the
conserver daemon now contain all the control codes and repeated
rendering of the command line, which makes them nearly unreadable for
us. This is why I have, for the moment, abandoned the idea. It might be
possible to work around this issue with a filter somewhere that
understands and interprets all the control codes, too. This requires
duplicating the work that is done by (and the code implementing) the
linenoise (or any other) command line library, which is why I don't
particularly like the idea.

Cheers
Ben
-- 
"Make it so they have to reboot after every typo." ― Scott Adams

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System