summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorjoey <joey@0fa5a96a-9a0e-0410-b3b2-a0fd24251071>2007-04-22 20:33:10 +0000
committerjoey <joey@0fa5a96a-9a0e-0410-b3b2-a0fd24251071>2007-04-22 20:33:10 +0000
commit496d0f819dc4b391f264484198c3f34713f0a9a8 (patch)
tree4e8e86f46df484e63169818b5eac88227878d0fd
parent94294260cc6c349e83f79941692ee89ab4443abf (diff)
web commit by http://id.inelegant.org/
-rw-r--r--doc/todo/fileupload/soc-proposal/discussion.mdwn5
1 files changed, 5 insertions, 0 deletions
diff --git a/doc/todo/fileupload/soc-proposal/discussion.mdwn b/doc/todo/fileupload/soc-proposal/discussion.mdwn
index 8461c829c..ff92264a9 100644
--- a/doc/todo/fileupload/soc-proposal/discussion.mdwn
+++ b/doc/todo/fileupload/soc-proposal/discussion.mdwn
@@ -1,7 +1,12 @@
There's nothing in [[fileupload]] that suggests putting the file upload limit in the body of a page. That would indeed be a strange choice. Ikiwiki already uses [[PageSpecs|PageSpec]] in the Preferences page (for specifying locked pages, banned users, and subscriptions), and I had envisioned putting the file upload controls there, and possibly subsuming some of those other controls into them.
+> Thanks for clarifying; I clearly misunderstood the original text.
+
It's not clear to me that the concept of attaching files to a page fits ikiwiki very well; unlike most wikis, ikiwiki supports subdirectories and [[SubPages|SubPage]], which allows for hierarchical placement of uploaded files, which is a much more flexible concept than simple attachment. Futhermore, the idea of listing all attached files at the bottom of a page seems somewhat inflexible. What if I want to make a podcast, using inline's existing support for that -- I won't want a list of every "attached" file at the bottom of my podcast's page then.
+> If a file was attached to _some-dir/some-page_, it would be stored in _some-dir/_ and linked from _some-page_. That would seem reasonably hierarchical to me. What do you suggest as an alternative?
+> As for the attachment list, I envisaged that being optional.
+
I don't understand why the file size would need to be stored in the index file; this information is available by statting the file, surely? Similarly, the mime type can be determined through inspection, unless there turns out to be a reason to need to cache it for speed.
--[[Joey]] \ No newline at end of file