[petsc-dev] Stokes flow code PETSC

[Matthew Knepley]

On Mon, Sep 12, 2011 at 10:37 AM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> On Mon, Sep 12, 2011 at 17:20, Jie Chen <jiechen at mcs.anl.gov> wrote:
>
>> One case though, that I think it would be more convenient if the options
>> specified in the file are overridden by command line options is, when people
>> want to tune parameters. The varying parameters are better changed in
>> command line than in the options file. In the case that for some reason the
>> file already specifies the parameters, I would like to see them overridden
>> by command line options.
>
>
> I agree about usually wanting the command line to have priority. The
> question is
>
> 1. Is it any hardship to always put -options_file first in the command line
> so that later args always override it?
>
> 2. Is it more confusing to have -options_file parsed first, or to have all
> options parsed in the sequence where -options_file is effectively an
> "include"?
>

I definitely do not want this.

   Matt

-- 
What most experimenters take for granted before they begin their experiments
is infinitely more interesting than any results to which their experiments
lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20110912/0d1feb95/attachment.html>