Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 20 May 2015 11:59:01 +0100
From:      Arthur Chance <freebsd@qeng-ho.org>
To:        Luca Ferrari <fluca1978@infinito.it>, freebsd-questions <freebsd-questions@freebsd.org>
Subject:   Re: ports, packages, jails
Message-ID:  <555C68F5.4080205@qeng-ho.org>
In-Reply-To: <CAKoxK%2B4fAsPATNfHEvkneUFxN_hJyCQ4fGeRq8rC8G-n_JTVPQ@mail.gmail.com>
References:  <CAKoxK%2B4fAsPATNfHEvkneUFxN_hJyCQ4fGeRq8rC8G-n_JTVPQ@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On 20/05/2015 11:14, Luca Ferrari wrote:
> Hi all,
> reading some recent discussions I start wondering what the real
> problem with mixing ports and packages is and why one should use jails
> to build ports (e.g., as poudriere does). What is the real advantage
> of using a building system with regard to a "normal" usage of ports
> and packages?


To answer your questions in reverse order, building in a jail has the 
advantage that it doesn't affect the main host until everything has been 
compiled successfully. In the past, when I used portmaster for updating, 
on a few occasions it would would update some ports and then fail on one 
with a buggy update, leaving my machine in mixed state that meant I 
couldn't use a graphical desktop and/or some services didn't work. 
Making the update of the running machine separate from rebuilding the 
ports in poudriere means I don't get those problems.

As for mixing ports and packages, there are a couple of possible 
problems. One is that using ports probably means you've changed some 
options to non-default values. Packages are built with default options, 
so clashes can occur. The other is that (re)building all the packages 
takes time, so for a period a package may be at an earlier version than 
the same software built from /usr/ports. This can also cause problems.


-- 
Those who do not learn from computing history are doomed to
GOTO 1



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?555C68F5.4080205>