On Mon, Mar 10, 2008 at 07:48:34PM -0500, Dan McGee wrote:

On Tue, Jan 29, 2008 at 7:44 AM, Nagy Gabor wrote:

...

...

From 9274e45ad97cce0749e23340cf51d8a3de0935b9 Mon Sep 17 00:00:00 2001 From: Nagy Gabor Date: Tue, 29 Jan 2008 13:43:08 +0100 Subject: [PATCH] New sync1007.py pactest

This pactest shows that checkdeps can be fooled when a package exists in multiple repos.

Signed-off-by: Nagy Gabor --- pactest/tests/sync1007.py | 18 ++++++++++++++++++ 1 files changed, 18 insertions(+), 0 deletions(-) create mode 100644 pactest/tests/sync1007.py

diff --git a/pactest/tests/sync1007.py b/pactest/tests/sync1007.py new file mode 100644 index 0000000..4245142 --- /dev/null +++ b/pactest/tests/sync1007.py @@ -0,0 +1,18 @@ +self.description = "Conflicting package names in sync repos (checkdeps is fooled)" + +sp1 = pmpkg("pkg1") +sp1.depends = [ "pkg2=1.0" ] +self.addpkg2db("community", sp1) + +sp2 = pmpkg("pkg2", "1.1-1") +self.addpkg2db("testing", sp2) + +sp3 = pmpkg("pkg2", "1.0-1") +self.addpkg2db("extra", sp3) + +self.args = "-S pkg1 pkg2" + +self.addrule("PACMAN_RETCODE=0") +self.addrule("PKG_EXIST=pkg1") +self.addrule("PKG_EXIST=pkg2") +self.addrule("PKG_VERSION=pkg2|1.0-1") -- 1.5.3.8

Did we ever resolve the disparity with this pactest and sync134.py?

I think I agree that we should handle 1 replacer as our normal case. However, I think sync134 was meant to test something different, although it may be invalid. Was it intended to test a package being split into two? If so, more than likely one of them would depend on the other and they would both get pulled anyway, so I am not sure we need to keep it around.

Let me know, and then we can work on getting this merged.

lol, it's rather funny what happened. Nagy submitted two totally different sync1007 pactests the same month. (he already noticed it a while ago, after it happened). The one you are referring to (which isn't the one you quoted) is the multiple replacer one, first submitted here : http://www.archlinux.org/pipermail/pacman-dev/2008-January/010819.html then a patch for fixing it (which took care of sync134 too) : http://www.archlinux.org/pipermail/pacman-dev/2008-January/010877.html and the last comment about it : http://www.archlinux.org/pipermail/pacman-dev/2008-February/011289.html So as I said, I find the current behavior alright. But I don't think the patch is wrong either, so I won't oppose to merging it :) Now, about the second sync1007 pactest above, I don't think it has been discussed anywhere. It might be worth merging too.