aboutsummaryrefslogtreecommitdiff
path: root/0.8.0/doc/doc_known_issues.doxygen
diff options
context:
space:
mode:
authorAvatar Ciaran McCreesh <ciaran.mccreesh@googlemail.com> 2006-10-11 20:56:59 +0000
committerAvatar Ciaran McCreesh <ciaran.mccreesh@googlemail.com> 2006-10-11 20:56:59 +0000
commite3b86134db7baada7b3a1e779e7d6d046ab53a1a (patch)
tree6acb3c0de68e1b95dbba8f0e00f770b8f31e4bac /0.8.0/doc/doc_known_issues.doxygen
parentab9cdb1150d97449e857cd55fe706377c9ab2422 (diff)
downloadpaludis-ecde7a6645289acadcd0998d220203363d7d3e77.tar.gz
paludis-ecde7a6645289acadcd0998d220203363d7d3e77.tar.xz
Tag release 0.8.00.8.0
Diffstat (limited to '0.8.0/doc/doc_known_issues.doxygen')
-rw-r--r--0.8.0/doc/doc_known_issues.doxygen80
1 files changed, 80 insertions, 0 deletions
diff --git a/0.8.0/doc/doc_known_issues.doxygen b/0.8.0/doc/doc_known_issues.doxygen
new file mode 100644
index 0000000..db7b562
--- /dev/null
+++ b/0.8.0/doc/doc_known_issues.doxygen
@@ -0,0 +1,80 @@
+/* vim: set ft=cpp tw=80 sw=4 et : */
+
+/**
+\page KnownIssues Known Issues and Non-Issues
+
+\section KnownIssuesMain About the issues
+
+This document describes some of the more commonly encountered problems, issues
+and things that aren't bugs but might look like they are. Pestering anyone about
+these is liable to get you hurt.
+
+\section KnownIssuesMain About the non-issues
+
+This document also describes some things that are not bugs or missing functionality. Pestering anyone about these is liable to get you hurt a lot.
+
+\section KnownIssuesPackages Packages
+
+\subsection KnownIssuesUdev sys-fs/udev
+
+Problem: <code>sys-fs/udev</code> will fail at the merge stage.
+
+Workaround: Use <code>~arch</code> for this package. We don't support merging
+fancy things like device files.
+
+\subsection KnownIssuesKDE Some KDE libraries are broken on some amd64 profiles
+
+Problem: Various KDE applications install libraries into the wrong libdir.
+
+Workaround: Make sure you're running 0.6.1 or later.
+
+Workaround: Check your profile. You can't use <code>default-linux/amd64/</code>
+directly -- you have to use one of the release subdirectories. If this still
+fails, someone who knows what they're doing needs to sit down and figure out
+exactly what's going on for the profile in question.
+
+\subsection KnownIssuesCheck Various packages fail src_test
+
+Problem: Various packages fail <code>src_test</code>
+
+Workaround: You can <code>export SKIP_FUNCTIONS="test"</code> to skip tests. This
+is best done on a per-package basis via <code>bashrc</code>:
+
+\verbatim
+case "${PN}" in
+
+ automake|glibc|e2fsprogs|neon|gettext|tar)
+ export SKIP_FUNCTIONS=test
+ ;;
+
+esac
+\endverbatim
+
+Unfortunately not all package maintainers care about making their package's test
+suite work. This is a nuisance, a) because it makes things much harder for arch
+teams and b) because it makes it harder for users to catch bugs.
+
+\subsection KnownIssuesRootSandbox Sandbox violations when ROOT!=/
+
+Problem: Sandbox violations when ROOT!=/
+
+Workaround: Paludis enforces ROOT. However, some packages don't honour ROOT. To
+temporarily disable sandbox for these packages, <code>export SANDBOX_PREDICT=/</code>.
+
+\section KnownNonIssuesPortageThings Things Portage Does
+
+\subsection KnownNonIssuesWget wget Resume Support
+
+Non-Problem: With Portage, <code>wget -c</code> is used to attempt to resume downloads of partial files. With Paludis, resumes are not allowed.
+
+Rationale: This leads to corruption and wasted bandwidth far too frequently. In particular, if an error page that isn't recognised as a 404 is fetched from one server (this is common for <code>mirror://sourceforge/</code>), resume support means <code>wget</code> would then download all but the first few hundred bytes of the file from somewhere else, leading to a corrupt distfile notice only after lots of bandwidth has been wasted.
+
+\subsection KnownNonIssesResume Resume / Skip First Support
+
+Non-Problem: Paludis doesn't have an equivalent to --resume --skipfirst in Portage.
+
+Rationale: Too unreliable, too flaky and far too widely abused. There's talk of echoing a command (<code>paludis -i10 =sys-apps/foo-1.23-r1 =app-misc/fnord-2 ...</code>) that can be used to resume if an ebuild exits with an error, but that's not set in stone.
+
+*/
+
+