Code review comment for lp:~maddevelopers/mg5amcnlo/2.2.0

Hi Olivier,

Wow, a really tough one to investigate.
No I don't want to fix this because in my opinion the problem lies in the
fact that the syntax for setting options of the check command is really bad
right now and not uniform. Ideally we should have all options set with the
syntax '--<option_name>=<option_value>'. That would make the parsing
simpler and unambiguous. But that will be for some next release.

Cheers,

On Tue, Sep 23, 2014 at 8:10 AM, Olivier Mattelaer <
<email address hidden>> wrote:

> Review: Approve
>
> Hi Valentin,
>
> I have found some time to investigate this bug, and I can now confirm that
> this is a bug of my computer rather than the code itself.
> The reason of this problem is that for some reason a file named "e+" was
> present on my local directory. (Do not ask me why probably a wrong
> manipulation). So the command check_check was interpreted the check e+ e- >
> t t~ as if the first entry was an input file rather than a particle...
>
> We can off-course add a check that this entry is not a particle name (in
> top of checking if the file exists) but I think that such situation should
> be rare enough. So It's up to you if you want to have a fix for this or not.
>
> Cheers,
>
> Olivier
>
> --
> https://code.launchpad.net/~maddevelopers/mg5amcnlo/2.2.0/+merge/234925
> Your team MadDevelopers is subscribed to branch
> lp:~maddevelopers/mg5amcnlo/2.2.0.
>

--
Valentin