Re: [PATCH] remote show: fix the -n option

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Jun 9, 2008, 16:31 UTC
Message-ID: <alpine.DEB.1.00.0806091729020.1783@racer>
In-Reply-To: <484D4FB3.2090309@freesurf.fr>

Hi,

On Mon, 9 Jun 2008, Olivier Marin wrote:

Show 14 quoted lines

> Olivier Marin a écrit :
> > Johannes Schindelin a écrit :
> > 
> >> I would not like to remove the got_states.  I think this is the wrong 
> >> direction.  Rather change the die() into a return error().
> > 
> > OK, I will try that.
> > 
> 
> After reading some more code, I can say that changing die() in return 
> error() won't change anything here because, in get_ref_states() we only 
> die() if get_fetch_map() return an error. But guess what, 
> get_fetch_map() never return an error. It just die() or return 0. And I 
> can't change it without breaking "clone" and "fetch".

So you think it is okay, because the result is the same? I think not. I think this is exactly the way of thinking that makes reusing unlibified parts of Git's source code hard. I think that this is exactly the style of programming I try to avoid, because it messes up clean concepts.

And I am utterly embarassed that we are talking about my code here.

Ciao, Dscho