Subject: Re: Updating packages
To: Matthias Scheler <email@example.com>
From: Aleksey Cheusov <firstname.lastname@example.org>
Date: 04/30/2007 17:12:11
> On Sat, Apr 28, 2007 at 01:36:34AM +0300, Aleksey Cheusov wrote:
>> I expect that this
>> - builds package B
>> - creates B-x2.y2.z2.tgz in PACKAGES
>> - installs A from .tgz package from PACKAGES
> "make update" doesn't use binary packages. You should look at one of
> package management tools like "pkg_chk".
16.17. Other helpful targets
Install a binary package from local disk and via FTP from a list
of sites (see the BINPKG_SITES variable), and do a make package if
no binary package is available anywhere. The arguments given to
pkg_add can be set via BIN_INSTALL_FLAGS e.g., to do verbose
The following variables can be used either on the command line or
in /etc/ mk.conf to alter the behaviour of make update:
Install target to recursively use for the updated package and
the dependent packages. Defaults to DEPENDS_TARGET if set,
"install" otherwise for make update. Other good targets are
"package" or "bin-install". Do not set this to "update" or you
will get stuck in an endless loop!
I asked developers and those who know pkgsrc internals,
i.e. know how it is implemented, why
doesn' work together.
I think 'make update' mode that
1) uses binary package if they are available
2) build these binary packages
would be very nice feature.
>> but A is not installed at all. Why?
> What do you mean bu "A is not installed"? It should be built from source.
"is not installed" means that
process of building from sources were not started at all.
I know about pkg_chk. It is extreeeeeeeeeeamely slow.
Before starting real job (building/installing)
it eats MINUTES on my old home machine and on more modern machines too.
With my patch it works ~10 times faster in some modes
but the patch is not complete yet.
It is so slow not because it is written in shell but because
it has O(N^2) complexity of some operations where O(N) is possible.
Best regards, Aleksey Cheusov.