Re: [bug-gv] [bug #19632] quiet and .gv watch settings are not respected

[Markus Steinborn]

Hi John,

> If you want to disable the quiet setting again I think all
> you need to do is remove from info.c the line
>
> if(gv_gs_quiet) return;

> Better yet, this conditional should be disabled only if gs returns with
> exit code 1.

Scheduled for 3.6.5, but not promised.

> The entry for the quiet option in the man page should then be updated to 
> explain this more sensible behaviour.

Clearly.

> Or, if we really want to restore the old and rather annoying behaviour 
> of the quiet option,

Ypou find the "quiet"-Option under ghostscript settings, not under gv 
settings. So it has to be an option that influences the ghostscript 
invocation, I think. I was very surprised to realize that gs is invoked 
with "-dQUIET" even so I switched "quiet" off in the ghostscript options 
dialog.

> we should at least add a "silent" option so that gs popups about (in 
> practice very common) noncompliant PostScript can be avoided.

> The silent setting would be even stronger than quiet in that it will 
> suppress all warnings and errors.

I am not sure. I currently have two options in mind:

--- first possibility ---
I think, the silent option is good, but if getting the return code will 
work, then it would be better to have three levels of silence:

1st  suppressing only outputs where ghostscript does not terminate (so 
these are no errors) -- perheps considering the return code

2nd suppressing all messages not containing the word "Error:" (at the 
beginning of a line!?), from the word "Error:" to infinty not supressing 
anything, as this should be the complete error message.

3rd  supressing all messages of ghostscript

--- second possibility ---
Another option is, not supressing anything, but showing the popup window 
only in case of an abortion of the ghostscript process (it normally won't 
terminate during viewing a single file).




This change would be too late for 3.6.4 (testing time would be too short, 
the patches I have accepted has been tested by me for five month, debian 
has tested them, too. So I can assume them to be stable and relaase them 
soon).

So lets schedule it for 3.6.5.


Greetings from Germany

Markus Steinborn