Bill Allombert on Thu, 01 Nov 2007 22:39:51 +0100

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: Patch: user warnings in GP

On Thu, Nov 01, 2007 at 03:18:29PM +0100, Jeroen Demeyer wrote:
> Index: src/functions/programming/error
> ===================================================================
> RCS file: /home/cvs/pari/src/functions/programming/error,v
> retrieving revision 1.7
> diff -u -r1.7 error
> --- src/functions/programming/error	31 Mar 2007 13:49:48 -0000	1.7
> +++ src/functions/programming/error	1 Nov 2007 14:17:44 -0000
> @@ -14,7 +14,7 @@
>  Prototype: vs*
>  Help: error({str}*): abort script with error message str
>  Description: 
> - (?gen,...):void  pari_err(talker, "${2 format_string}"${2 format_args})
> + (?gen,...):void  pari_err(user, "${2 format_string}"${2 format_args})
>  Doc: outputs its argument list (each of
>   them interpreted as a string), then interrupts the running \kbd{gp} program,
>   returning to the input prompt. For instance

I am not sure whether you intended to send this hunk, but I have
considered doing that already and I am of two mind about it:

Using 'user' is more 'correct' than 'talker' but pari_err(talker,...)
make the generated code looks more like libpari code.

Very personnaly, I find the way user error are displayed under GP terribly

I think a goal for PARI/GP could be to allow to write GP functions that
looks exactly like PARI functions from the outside. error() breaks that.