On 8 September 2014 05:54, Marc A. Pelletier marc@uberbox.org wrote:
And yet, after over a decade of open-ended design through social convention, the end result is... our current talk pages. Perhaps another decade or two will be needed before that document-centric architecture gives us a half-decent discussion system?
Marc, I'm not arguing against having a discussion system. In fact I think having threaded comments happen by default is a great idea that will make the conversation interface far more usable, both on desktop and mobile (I agree with Gerard that the mobile editing experience is dreadful).
The problem I see is with having that discussion system as the 'only' option, making refactoring of conversations limited and difficult, and removing the open-ended and flexible platform we currently rely upon, when several important workflows and goals such as accountability and building new workflows for projects are based on the well understood capabilities of a wiki system.
The system I envision as a suitable, modern replacement would be based on proven enterprise collaboration platforms like Microsoft OneNote or Atlassian Confluence, which include discussion systems as modules integrated within the platform. I simply can't see the benefit of losing the collaboration capabilities of wiki software in favor of enforced structured discussion, when we can have both.
Now if Erik vision for the deeper than I give him credit for, and he is able to build a OneNote-like application on top of the suggested architecture for Flow, I will eat my words with an apology :-) However, that capability of the system should be better explained so that we can understand it and discuss its ramifications.
On 7 September 2014 23:53, Diego Moya dialmove@gmail.com wrote:
On 09/06/2014 17:06 PM, Marc A. Pelletier wrote:
On 09/06/2014 12:34 PM, Isarra Yos wrote:
if the designers do not even understand the basic principles behind a wiki, how can what is developed possibly suit our needs?
You're starting from the presumption that, for some unexplained reason, collaborative discussion benefits from being a wiki (as opposed to, you know, the actual content).
Wikipedia has been built using that platform. I'd say that's a very good reason to trust that the model is at least capable. :-)
Very many people, myself included, believe that a wiki page is an *atrocious* medium for discussion.
Sure, and I agree there are many way to improve how users are engaged into discussion and to keep it manageable. But what is missing from this conversation is the point that Wikipedia talk space is not *merely* a medium for discussion: there are other vital roles that may be hindered by a radical focus on conversation:
tl;dr version: there are times and places that Wikipedia discussion system needs to be a Microsoft OneNote, and Flow is building us a Twitter (minus the 140 characters limit).
- The talk space has a strong expectation that it serves as an
archive of all decisions taken in building the articles, i.e. to show how the sausages are made. The disembodied nature of Flow topics, which may be shown out of order and distributed to many boards, makes it hard to recover a sequential view of the conversations in order as they happened.
- Same thing for keeping user's behavior in check - policy
enforcement often requires that the reviewers can see exactly what the users saw when they performed some particular disruptive action, to assess whether it was made in good faith from incomplete information or a misunderstanding.
- Comment-based discussion is not the only way editors collaborate;
nor discussions are limited to users expressing their particular views at ordered, pre-defined processes. Some fellow users have already pointed out how the wiki page works as a shared whiteboard where semi-structured or free-form content can be worked upon by several editors, and improved iteratively in an opportunistic way. Sometimes, that re-shaping of text is made onto the form of the conversation itself, by re-factoring, splitting, merging and re-classifying comments from many editors. This would be hard or impossible to do if the layout of the discussion is fixed in hardware and comments belong to the poster.
- Wikiprojects develop over time new procedures that better suit
the workflow of their members to achieve their goals. Their project pages are free-form collages of all the relevant information they require to do their work, plus discussion processes that may involve just its members or any other external participant. As projects cover all the aspects of human knowledge, it would be difficult to provide a one-size-fits-all interface that may cover all their needs - the flexibility to compose new layouts and compilations of content is core to achieve their goals.
- There's a sense now that the community owns the content of all
pages including talk, and can manage it to their liking. That will disappear if the base model is changed to one based on user-owned comments. There has not been enough discussion of how that will affect all the existing projects and guidelines, or whether such change is acceptable and beneficial to the goal of writing the encyclopedia.
A wiki-like system is very good at achieving those. Some of these needs have surfaced at previous discussion at Talk:Flow, and some have been already addressed or influenced the design, but some others are squarely opposed to the nature of a threaded discussion where the structure is enforced by the platform. It would be sensible that the process to gather feedback from the community includes solid answers to these facets of the tool.