summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorintrigeri <intrigeri@boum.org>2010-08-02 16:00:37 +0200
committerintrigeri <intrigeri@boum.org>2010-08-02 16:00:37 +0200
commitc5b052cbbda14c5c33127890d4a9842b8dee826a (patch)
tree25788be549bd49727a7bafc2172af569e2101fcf
parent97a706f73ce2cedc8fd8a9bbeea6d25783260b2e (diff)
explored possible ways to do it, got trapped.
-rw-r--r--doc/todo/po:_remove_po_files_when_disabling_plugin.mdwn7
1 files changed, 7 insertions, 0 deletions
diff --git a/doc/todo/po:_remove_po_files_when_disabling_plugin.mdwn b/doc/todo/po:_remove_po_files_when_disabling_plugin.mdwn
index 26b1964ba..0801f7fcd 100644
--- a/doc/todo/po:_remove_po_files_when_disabling_plugin.mdwn
+++ b/doc/todo/po:_remove_po_files_when_disabling_plugin.mdwn
@@ -2,3 +2,10 @@ ikiwiki now has a `disable` hook. Should the po plugin remove the po
files from the source repository when it has been disabled?
> pot files, possibly, but the po files contain work, so no. --[[Joey]]
+
+>> I tried to implement this in my `po-disable` branch, but AFAIK, the
+>> current rcs plugins interface provides no way to tell whether a
+>> given file (e.g. a POT file in my case) is under version control;
+>> in most cases, it is not, thanks to .gitignore or similar, but we
+>> can't be sure. So I just can't decide it is needed to call
+>> `rcs_remove` rather than a good old `unlink`. --[[intrigeri]]