Submissions/Wikitext is broken, long live wikitext (2.0)

After careful consideration, the Programme Committee has decided not to accept the below submission at this time. Thank you to the author(s) for participating in the Wikimania 2015 programme submission, we hope to still see you at Wikimania this July.

Submission no.
Title of the submission
Wikitext is broken, long live wikitext (2.0)
Type of submission (discussion, hot seat, panel, presentation, tutorial, workshop)
Short presentation (~15 min) followed by discussion / panel
Author of the submission
Subramanya (Subbu) Sastry, Brion Vibber
E-mail address,
User:SSastry (WMF), User:Brion Vibber (WMF)
Country of origin
Affiliation, if any (organisation, company etc.)
Wikimedia Foundation
Personal homepage or blog
Abstract (at least 300 words to describe your proposal)

Tongue-in-cheek attention-grabbing title apart, this session is to start discussing about how to evolve wikitext.

Wikitext, as it currently exists, is broken for a few reasons:

  1. Its design makes high-performance parsing difficult.
    • You cannot parse transclusions in parallel, for example.
    • You cannot do incremental updates of parsed HTML easily. As it exists, minor edits still lead to full-page reparsing.
    • Template edits for popular templates are extremely expensive because of this reason.
  2. It doesn't play well with HTML-based editing.
    • Visual editing is here to stay and should have been here long back.
    • Heroics of Parsoid and VisualEditor apart in getting where it is now, it still makes visual editing difficult because not all wikitext is representable as editable HTML.
  3. Lots of edge cases makes life for a new parser (ex: Parsoid) difficult.
    • Wikitext errors are especially problematic since the output is now dependent on implementation details of the parser rather than the wikitext itself.
  4. The current templates are ghastly.
    • This is not because template editors are doing a bad job, but because they are doing their very best to extract as much juice they can out of a broken setup. Lua-based templating fixes some of this.

All that said, no matter how good visual editing gets, some editors and some kinds of editing tasks are going to be easier in a "text editor". So, let us call that text-based format wikitext 2.0. Rather than think of this as competing edit modes, it is perhaps useful to think of this as how can we evolve wikitext to support multiple editing modes without making life difficult for either.

We have a few ideas / fixes in mind for issues 1. and 2. above which we will present. However, this is also a session to discuss more widely how we should go about fixing this brokenness and perhaps start thinking about how to get there from here. Part of this should also be about setting clear expectations that an updated markup format should meet regarding usability, performance, editability, consistency, backward compatibility, and ability to transition smoothly.

Technology, Interface & Infrastructure
Length of session (if other than 30 minutes, specify how long)
60 minutes (Given the nature of the session, this will benefit from an initial introduction setting up the discussion a longer discussion. A 30-min slot makes it difficult to accomplish much)
Will you attend Wikimania if your submission is not accepted?
Slides or further information (optional)
Special requests
Timing should not conflict with the Templates are dead! Long live templates! talk.

Interested attendees

If you are interested in attending this session, please sign with your username below. This will help reviewers to decide which sessions are of high interest. Sign with a hash and four tildes. (# ~~~~).

  1. Daniel Mietchen (talk) 12:59, 28 February 2015 (UTC)[reply]
  2. Amir E. Aharoni (talk) 15:36, 6 March 2015 (UTC)[reply]
  3. Cscott (talk) 20:03, 10 March 2015 (UTC)[reply]
  4. CT Cooper · talk 00:01, 11 March 2015 (UTC)[reply]
  5. EpochFail (talk) 21:10, 20 March 2015 (UTC)[reply]
  6. Quiddity (WMF) (talk) 06:43, 23 March 2015 (UTC)[reply]