Re: [pacman-dev] [PATCH] add support for .so dependencies
On Sun, Aug 16, 2009 at 12:45:50AM +1000, Allan McRae wrote:
"arch" is a reserved variable name in PKGBUILDs. This will break split package support for packages that have both binary and arch=any packages (which will be supported in the future).
Okay, this one should be easy to fix.
As far as I can tell, IFS does absolutely nothing here.
IFS is set to newline.
Well, the patch being broken and so I assume untested was my main issue. I will not review a patch the submitter has not even tested.
Bluewind tested it yesterday, and it worked for him.
Well, then... what is the point? Versioned dependencies already cover soname values if the package is listed as a dep in addition to is library.
The point is that the dependencies are magically generated and this greatly reduces the work for the maintainer to specify the explicit versioned dependencies (which doesn't happen right now). Also the .so dependencies are not about the library version, but about the content of the package.
greetings, Christoph Schied
Christoph Schied wrote:
On Sun, Aug 16, 2009 at 12:45:50AM +1000, Allan McRae wrote:
As far as I can tell, IFS does absolutely nothing here.
IFS is set to newline.
Ah. I forgot about that bash internal... Piping through something like
tr " " "/n"
would be better and far less ugly.
Well, then... what is the point? Versioned dependencies already cover soname values if the package is listed as a dep in addition to is library.
The point is that the dependencies are magically generated and this greatly reduces the work for the maintainer to specify the explicit versioned dependencies (which doesn't happen right now). Also the .so dependencies are not about the library version, but about the content of the package.
The "magically" adding depends is my concern. If the libraries provided by a package are to be in a depends array, then there is no need for that package to be in the depends array (but should be in makedepends). Otherwise this makes pacman have to do redundant checks and slows down dependency resolution.
What I would like is a description of the problem this is trying to solve that could not be already handled by versioned deps?
Allan
participants (2)
-
Allan McRae
-
Christoph Schied