From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Thu, 10 Apr 2014 23:19:19 +0200 Subject: [Buildroot] [PATCH 0/3] [RFC] pkg-autotools: introduce infra to gettextize packages (branch yem/gettextize) In-Reply-To: References: Message-ID: <20140410231919.2538fac1@skate> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Dear Yann E. MORIN, On Thu, 10 Apr 2014 00:20:00 +0200, Yann E. MORIN wrote: > This small series introduces the gettextize of a package as an optional > sequence in the autotools package infra. > > WARNING! Completely untested, only an RFC! > > Currently, wget is the sole user, and the way it is done for wget is > just working by chance (see commit log of path 3 for the complete > explanations.) > > So, move the gettextization to the pkg-autotools infra, so we have a > chance to do it correctly and consistently across packages. I haven't looked at the details. But to me, when only one package needs something, it is *way* too early to make changes to the common infrastructure. Especially when there is a solution that allows to solve the problem by making changes to the package .mk file itself. I believe changes to the package infrastructure should only be made when: * There is really no solution to solve the particular problem at the package .mk file level; * There are a sufficient (say 5 or 10 maybe) packages that have a similar problem, and therefore factorization would make sense. So far, I don't really see a compelling reason to complexity the package infrastructure just for the need of one particular package. Need that may very well be transient, because a future bump of wget will most likely remove the need for the gnulib patch, therefore for the autoreconf. Thomas -- Thomas Petazzoni, CTO, Free Electrons Embedded Linux, Kernel and Android engineering http://free-electrons.com