diff options
Diffstat (limited to 'doc')
| -rw-r--r-- | doc/forum/managing_todo_lists.mdwn | 29 |
1 files changed, 29 insertions, 0 deletions
diff --git a/doc/forum/managing_todo_lists.mdwn b/doc/forum/managing_todo_lists.mdwn index 238b000f1..b55420aca 100644 --- a/doc/forum/managing_todo_lists.mdwn +++ b/doc/forum/managing_todo_lists.mdwn @@ -2,3 +2,32 @@ I keep some TODO lists on ikiwiki pages. I'm half-tempted to write a plugin to make ticking items off and adding items easier via the web interface. I'm aware though that this is not really what ikiwiki is designed for. Would anyone else find this useful? -- [[users/jon]] + +---- + +My subsequent thoughts about how to approach this are two-fold. + +Firstly, a filetype for todo lists, probably OPML, but I haven't looked to see +if there is something more suitable. A plugin that converts this source into a +traditional page output, i.e. a DOM tree of ul or ol and li elements. + +Secondly, some magic javascript to make editing the list via the web page +more interactive: add items, strike items out, reorder items etc., without +round-tripping to the cgi for each operation. + +Finally, a mechanism whereby the changes made to the page live can be +committed back to the repository: + + * ...perhaps the input → output conversion is reversible, and the HTML DOM + representing the list can be transformed back into the source and submitted + to the cgi like a regular edit: issues include the result of other + postprocessing: templates, wikilinks, etc. + * perhaps an embedded copy of the source is included in the output and the + javascript operates on that in tandem with the static copy + * perhaps the "output" is generated live by the JS at view time (with maybe + a plugin-generated rendered output for non JS environments) + +I envisage a button called "commit changes" appearing once some changes are +made that submits the changes to the CGI, perhaps via a back channel. I'm not +sure how to handle embeds or challenges from the CGI such as a login challenge +(maybe the back channel would not be necessary in the first cut). |