On 31/07/2007, at 16:49, Joerg Sonnenberger wrote:

> On Tue, Jul 31, 2007 at 03:51:43PM +0200, Dieter Baron wrote:
>>   I think sooner or later pkg_add will have to consult this file as
>> well, and I don't want pkg_add to parse this huge file on each run.
>
> The only reason I can think of that pkg_add should consult pkg_summary
> file is to avoid the penalties of http fetch. But if you want to do
> that, you have a *local* copy already and just use it to construct the
> URLs directly. That can be done with a frontend to pkg_add already  
> and I
> think it is the better idea to do that.

I just hope we don't end with a pkg_add similar to Fedora's yum.  It  
takes *ages* to download lots of tiny files (the "headers" for each  
package) from a remote server.

-- 
Julio M. Merino Vidal <jmmv84@gmail.com>