More than 500,000 comments have been posted with the Reply tool.
Reply Tool: The Reply Tool is available by default to all editors at all WMF-hosted wikis except for Meta-Wiki, Wikispecies, and the English, Finnish, and Russian Wikipedias.
New Topic Tool: A large A/B test will begin soon at 20 Wikipedias. To disable the New Discussion tool, go to Special:Preferences, click on the Editing tab, and scroll to the end of the page. Turn off "Quick topic adding" and save this change to your preferences.
Topic Subscriptions (Notifications): Manual topic subscriptions are available in the Beta Feature at all wikis. Automatic topic subscriptions have entered early testing.
Soon, the team will begin working on a set of enhancements to make it easier for volunteers, across experience levels, to identify and understand the conversations happening on talk pages. As we near starting the design process for this work, we have a question for you all:
🤔 When looking at a talk page, what do you think would make it easier for you to know what conversations are worth reading?
Please share what you think on the project's talk page here.
Summary
GIF showing an in-progress design for the new talk page empty state experience.
The team has been focused on the following projects. More details can be found below.
New Topic Tool: People who have the New Topic Tool enabled will soon have access to a new experience on empty, not-yet-created talk pages.
Topic Subscriptions (Notifications): At the end of June, manual topic subscriptions were made available as a beta feature at an initial set of seven wikis. Development of automatic topic subscriptions is now underway.
Reply Tool: Work continues on optimizing the database the Reply Tool depends on. Once this work is complete, we will resume offering the Reply Tool as an on-by-default feature at more wikis.
Visual enhancements: the design strategy, and set of interventions we are planning to work on as part of this project, is now published on the project page here.
Extension:DiscussionTools/How it works: This is a new resource available to developers interested in writing gadgets or SQL queries that depend on DiscussionTools
Usage: to date, people have used the New Topic Tool to start 30,000+ conversations.
Deployments: in the coming weeks, the New Topic Tool will become available as a default feature on the first set of wikis. More details about the specific wikis and deployment timing will be posted to the project page in the coming weeks.
Empty talk pages: on many wikis, the majority of talk pages have not yet been created. To make the experience of creating these "empty" talk pages more intuitive for Junior Contributors, we are working on some improvements. Soon, when people who have the New Topic Tool enabled arrive on a talk page that has not been created, they will see a new placeholder that explains the purpose of talk pages as well as a clear call to action to start a new conversation. Note: people will still have access to the existing, full-page editing experience, by way of the page's standard Create Source button.
A screenshot showing a talk page where Topic Subscriptions are enabled. Notice the [ subscribe ] link that appears within the section heading.
Usability: In June, the team completed usability testing with Junior and Senior Contributors. In short, Junior and Senior Contributor test participants found the Topic Subscription prototype to work as they intuitively thought it would/should. You can review the test findings in detail here.
Deployments: On 28 June, manual topic subscriptions became available as an opt-in Beta Feature at the first set of seven wikis.
Scaling: In the coming weeks, we will begin offering manual topic subscriptions as an opt-in Beta Feature at even more projects. Follow T287276 for updates.
Trying topic subscriptions: You can experiment with Topic Subscriptions for yourself on any project by adding ?dtenable=1 to the URL of the talk page you are viewing.
To date, people have used the Reply Tool to make ~250,000 edits. Source
Usage: to date, people have used the Reply Tool to post ~250,000 talk page comments.
Deployments: the team continues to receive requests from wikis to make the Reply Tool available to all people by default. Despite this demand, and the data that demonstrates the positive impact the tool has on participation, deployments are paused while we continue with to make performance enhancements to ensure the tool continues to function as expected at this larger scale. More information about this can be found in T280599.
Strategy: In June, we published the design strategy and set of interventions we are planning to work on as part of the larger effort to help people, across experience levels, identify and understand the conversations happening on talk pages. Read more about this strategy on the project page here.
Summary
Percent of Junior Contributors that completed at least one comment attempt on a talk page during the AB test. | Source
The team has been focused on the following projects. More details can be found below.
Reply Tool: on 23-April-2021 we completed the analysis of the Reply Tool A/B test that ran from 11 February through 10 March on 22 Wikipedias. The results show the Reply Tool decreases reverts and causes Junior Contributors to have more success publishing comments. More below.
New Topic Tool: in the coming weeks, you will have the option of enabling tools for inserting pings and links in the tool's source mode. You can try the prototype on a test wiki here.
Notifications: A prototype of the new topic subscription feature is ready for you to try. You can try the prototype by following these instructions.
New requirements for user signatures: as of 29-April-2021, signatures will be required to contain a single line of wikitext. This affects the literal carriage return and line feed characters, not <br> and <p>. More information here.
Visual enhancements: you can read about the objectives of this upcoming part of the Talk pages project on the recently-expanded project page here and follow the work as it is being defined in Phabricator here.
The results are in from the A/B test of the Reply Tool that ran from 11 February through 10 March on 22 Wikipedias. What follows are the conclusions we are drawing from these results and the steps we are taking next.
The Reply Tool causes Junior Contributors to have more success publishing comments :
72.9% of Junior Contributors who opened the Reply Tool were able to publish at least one comment during the A/B test compared to 27.6% of Junior Contributors who attempted a comment using page editing. This amounts to a 160% increase in the comment completion rate of Junior Contributors who used the Reply Tool during the test than those who used page editing.
Junior contributors using the Reply Tool are 7.2 times more likely to publish a comment they start than Junior Contributors using page editing.
The Reply Tool decreases disruption:
There was a 79.5% decrease in the revert rate of comments Junior Contributors made with the Reply Tool compared to comments Junior Contributors made with page editing.
There was no significant increase in the number of Junior Contributors who were blocked after publishing a comment with the Reply Tool
Next steps: considering the A/B test has shown the Reply Tool causes a greater percentage of Junior Contributors to publish a comment without a significant increase in disruption, we will proceed with the plan we shared in T252057 to begin making the Reply Tool available as an opt-out preference at all Wikimedia projects. The sequence and timing of these deployments will be finalized in T280388.
A screenshot showing the New Topic Tool's new Souce mode.
In response to consistent feedback people have shared since the Reply and New Topic Tools were first made available, we are preparing to introduce tools for adding links and pings within the two tools' source modes.
To start, these new tools will be made available as an opt-in setting. Once we have verified these new tools work in the ways people expect and value, they will be offered by default. More information can be found in Phabricator here.
A screenshot showing the topic subscription prototype.
A prototype of the new topic subscription feature is ready for you to try. This prototype enables you to elect to receive a notification via Echo when someone posts a new comment in any conversation you have decided to "subscribe" to.
To start, you will need to manually elect to receive notifications when new comments are posted in sections you are interested in. In a future iteration, we will offer people the ability to automatically subscribe to conversations they post comments in and/or start. More information about manual and automatic topic subscriptions can be found here and here.
Summary
The team has been focused on the following projects. More details about each can be found below.
Reply Tool: On 20 February, the 100,000th comment was posted with the Reply Tool.
New Topic Tool: On 18 February, the New Topic Tool became available as a Beta Feature at the Arabic, Czech and Hungarian Wikipedias.
Notifications: Design and engineering work continues on developing a new feature that will enable people to subscribe to receive updates about new comments in specific talk page discussions.
A/B Test: On 11 February, an A/B test of the Reply Tool started at 22 Wikipedias (full list here). This test will help determine the impact the Reply Tool is having on how people post comments to talk pages. This test will also help determine whether the tool is ready to be offered as an opt-out setting at more Wikipedias.
Scaling: a conversation is ongoing at the English Wikipedia where volunteers are sharing the experiences they have had using the Reply Tool. This conversation is part of a larger effort to learn whether volunteers at en.wiki think the tool is valuable enough to be offered as a beta feature. Read the conversation and/or add to it here:Wikipedia talk:Talk pages project#Experiences
Initial deployment: the New Topic Tool is currently available as a Beta feature at the Arabic, Czech and Hungarian Wikipedias. The people who have used the tool have had positive things to say so far. Of the people who have tried the tool, most are using it more than once. More info here: https://quarry.wmflabs.org/query/52630.
Scaling: over the next couple of weeks the team will be continue monitoring how people are using the tool and explore offering the New Topic Tool as a beta feature at more projects.
The team is actively designing and building a way for people to elect to be notified when another person adds a new comment in a specific talk page discussion (read: section). In the next 2 months, we expect there to be a prototype for you all to try.
There is a conversation happening on the Notification project's talk page where volunteers are sharing feedback about the approach we have in mind for implementing topic subscriptions. Read and/or join in here: Topic:W3c084ihivgkm9vt.
Summary
The team has been focused on the following projects. More details about each can be found below.
Completion rates for comments made with the Reply tool and full-page wikitext editing. Details and limitations are in this report.
Reply Tool : On 5 January 2021, the team completed an analysis of the first ~2.5 months of Reply Tool usage at the three wikis where it is available by default (Arabic, Czech and Hungarian). A larger A/B test to evaluate the tool's impact on participation will start soon.
New Topic Tool: Junior and Senior Contributors recently completed testing the New Topic Tool prototypes. In the coming weeks, the tool will be offered as a Beta Feature at the Arabic, Czech and Hungarian Wikipedias.
Notifications: A new tool for subscribing to updates to specific talk page sections will be the first intervention we work on to increase the likelihood contributors receive relevant responses to the things they say.
Visual enhancements: one of the final components of this first phase of the Talk pages project will be visual enhancements to help people instinctively recognize and use talk pages as spaces to communicate with people on-wiki. The stub for the project page is now up.
On 5 January 2021, the team completed an analysis of the first ~2.5 months of Reply Tool usage at the three wikis where it is available by default (Arabic, Czech and Hungarian). The findings are encouraging, tho not conclusive considering the limited sample size.
A notable finding from this analysis: the edit completion rate for Junior Contributors using the Reply Tool is 4x higher than the edit completion rate for people using full-page wikitext editing.
You can review these findings in more detail by reading the Metrics section below and/or reviewing the full report: Reply Tool workflow engagement metrics.
Screenshot of version 1.0 of the New Topic Tool prototype.Over the past couple of months, Junior and Senior Contributors have been trying the New Topic Tool prototype to ensure it is both intuitive to use and complimentary to how people currently add new topics to talk pages.
Of the Junior Contributors who participated in the usability test, 9 (or 90%) positive adjectives to describe the experience, such as "easy", "intuitive," "straightforward" and "obvious."
Of the Senior Contributors who experimented with the prototype, most appreciated the consistency between how the New Topic Tool and Reply Tool look and function. Some people also noted how it could be valuable to make it so the affordance (think: link, button, etc.) for starting a new topic is easier for people who are new to notice.
During Talk pages consultation, editors expressed that it ought to be easier to know about new activity in conversations they are interested in. To address this concern, the team will soon begin work on talk page Notifications. As part of our research and design process, we are curious to know: What would help you become aware of this new activity? What's working with the current system? Which pages at your wiki should the team look at?
The team has been focused on the following projects. More details about each can be found below.
Reply Tool : we have continued making the Reply Tool available to more people at more Wikipedias. On 4-November-2020, the Reply Tool became available to another ~250 Wikipedias as a Beta Feature. More details in this 4-November update.
Notifications: we started planning and researching the work involved with improving the likelihood people know when others are saying something relevant to them on talk pages. More details on the new project page.
This image shows a screenshot of what it could look like to add a new discussion topic to a Wikipedia talk page to someone using the New Topic Tool'svisual mode.Dirty Diff Detector : we have continued working to improve the reliability of the Reply Tool. Over the past 5 days, an average of ~99.50% of edits made with the Reply Tool have been posted to talk pages without affecting any other parts of page.
Scaling: on 4-November-2020, the Reply Tool became available as a Beta Feature at an additional ~250 Wikipedias. The next step in the scaling process is to conduct an A/B test to evaluate the impact of the Reply Tool. This will help us determine whether the process for offering the tool as an opt-out feature at all projects is ready to begin. More information about this A/B test can be found on the project page: Metrics.
Measuring engagement: The team is analyzing the Reply Tool usage data from the Arabic, Czech and Hungarian Wikipedias, where the tool has been available as an opt-out feature since 24-September-2020. This analysis will help us determine whether people, across experience levels, are having success using the Reply Tool and whether people are using the tool in ways that disrupts others. Once complete, the analysis results will be published on the project page here: Metrics.
Initial designs: in late October, we shared and invited feedback about version 1.0 of the New Topic Tool designs on mediawiki.org and 15+ other Wikipedias.[1][2] Everyone who shared feedback expressed support for the general design direction. A more detailed summary of the feedback and plans for addressing it can be found here: Feedback summary: version 1.0 designs.
Prototype: within the next two weeks we will have a functional prototype ready that we would value people trying and sharing what they think about it. To learn when the prototype is ready, we recommend adding the New Topic Tool's project talk page to your watchlist; we will post a link to the prototype there when it is ready.
Notifications
The next area of the talk page experience we will set out to improve as part of the Talk pages project will be notifications.
This work is intended to increase the likelihood Junior and Senior Contributors receive timely and relevant responses to the things they say.
We are currently in the planning and technical exploration phase of this project. We plan to focus on the user-facing components of the implementation during the first three months of 2021.
Summary
The team has been focused on the following projects. More details about each can be found below.
Dirty Diff Detector: to ensure the new tools introduced as part of this project do not affect existing content in undesirable ways, we have created a way to detect when edits made with DiscussionTools affect other parts of the talk page.
Reply Tool: as peoples' use of, and confidence in the tool grows, we have been making the Reply Tool more widely available. The tool is now available at 41 Wikipedias. This is ~3x the number of Wikipedias the tool was available at the time the September update was written.
New Topic Tool: we are working on mockups of the New Topic Tool. In the coming weeks we will be sharing them and will value hearing what you think of them.
Suspicious diffs: this chart shows the percentage of diffs posted with the Reply Tool that affect other parts of the talk page in potentially disruptive ways.
To ensure the new tools introduced as part of this project do not affect existing content in undesirable ways, we have created a way to detect when edits made with DiscussionTools affect other parts of the talk page.
Over the past three weeks, an average of 99% of Reply Tool comments are posted to talk pages without affecting any other part of the page. You can review the results here: https://dtcheck.toolforge.org/dtstats.html.
Reply Tool comment volume: a chart showing the total number of comments posted with the Reply Tool between 31-March and 15-October-2020.
Metrics: usage of the Reply Tool continues to grow.
Comment volume: the number of comments people make with the Reply Tool each month has been growing by 31% (on average) for the past 5 months.
Unique people using tool: The number of unique people who have posted at least 1 comment with the Reply Tool has been growing by 36% on average each month, for the past 5 months.
Scale: to give you a sense of scale, people have used the Reply Tool to publish 30,000+ comments to date.
Deployments
14-October: the Reply Too became available as a Beta Feature at another 29 Wikipedias, bringing the total number of Wikipedias where the Reply Tool is available to 41.
24-September: the Reply Tool became available to all users (logged in and out) at the Arabic, Czech and Hungarian Wikipedias. This deployment marks the first time the tool has been made available to everyone at a project by default.
Enhancements
As of 17-September, anyone who uses the Reply Tool will have the ability to customize the edit/comment summary that is posted along with it.
In response to feedback volunteers have been sharing, the tool will soon "remember" whether you would like the Comment summary field shown or hidden when you use the Reply Tool.
The team is working on a few different design directions the New Topic Tool could take. These design directions are intended to help us make the following implementation decisions:
Initial touchpoint: Where and what do people click to start the process of adding a new discussion topic to a talk page?
Composition:Where do people draft the title and body of the topic they are planning to publish?
Posting/Feedback:What is the best way to communicate to people that the topic they intended others to be able to see and interact with has been posted successfully?
To see how we are thinking about the New Topic Tool more broadly, you might review the design approach on the project page: New topic#Approach.
4 Sep 2020
Summary
A journey map visualizing the kinds of interactions the Editing Team is intending the new talk page tools and enhancements to case more of.
The team has been focused on the following projects. More details about each can be found below.
Talk pages project: the team has decided on the new talk page tools and enhancements it plans to work on during this next fiscal year as part of this project.
RfC for mutli-line comment syntax: in the coming weeks, the Editing Team, in collaboration with the Parsing Team, will start an RfC to define new wikitext syntax for demarcating multi-line comments.
On 28-July-2020, the team completed an analysis of how people who enabled the Reply Tool Beta Feature at the Arabic, Dutch, French and Hungarian Wikipedias used the tool between 31-March and 30-June-2020. This analysis found:
328 people posted at least one comment using the Reply Tool
69.5% of people posted ≥ 2 comments with the Reply Tool
61.3% of people posted a comment using the Reply Tool on ≥ 2 distinct days
These results are encouraging and lead the team, and volunteers, to become more confident the tool is valuable.
DeploymentsCustom edit summary: a screenshot showing a *provisional* implementation of custom edit summaries within the Reply Tool.
On 6-August-2020, the Reply Tool became available as a Beta Feature at the following Wikipedias: Catalan, Chinese, Czech, Georgian, Korean, Serbian, Sorani Kurdish and Swedish. The Reply Tool is now available at 13 total projects. A full list can be found here.
In the coming weeks, the team intends to make the Reply Tool available as a Beta Feature at even more Wikipedias and as an opt-out preference at the Arabic, Czech and Hungarian Wikipedias.
Inspired by volunteer feedback [i][ii][iii][iv], the Reply Tool will soon offer people the ability to customize the edit summaries that accompany comments posted with it. You can try experiment with a prototype that incorporates this functionality by visiting this link on a desktop computer: Reply Tool with custom edit summary.
Junior and Senior Contributors are aware when someone is talking to them and when someone says something new in a conversation they are interested in. See: T233447; project page not yet published.
Junior Contributors recognize article and user talk pages as places to communicate with other people and understand the kinds of topics these spaces are used to discuss. See: T249579; project page not yet published.
The list you see above is built on the Talk page consultation and other research the team has done. To help visualize the kinds of talk page interactions and experiences we aspire Junior and Contributors to have as a result of this work, the team created the journey map pictured above.
RfC: new wikitext for multi-line comment
List items are fragile in wikitext and there are many things (e.g. tables and images) that cannot be effectively embedded in lists. The Editing Team and Parsing Team will soon be seeking your input on a proposal to introduce new wikitext syntax that will eliminate the aforementioned constraint and unlock new possibilities.
You will soon be able to read more information about what approaches are being proposed and why. In the meantime, you can see more information on Phabricator here: T246960.
Summary
Reply Tool: a screenshot showing the Reply Tool's new visual mode, equipped with a feature for searching for and pinging other people.
The team has been focused on the following projects. More details about each can be found below.
Reply tool: On 17-June, Version 2.0 of the Reply tool became available as a Beta Feature at our partner wikis and via a custom URL on all projects. Information about how you can turn it on at your Wikipedia below.
New user signature requirements: On 6-July, the new signature requirements were implemented on all wikis. Details about these requirements and how they are being rolled out is below.
On 17-June, Version 2.0 of the Reply Tool was deployed as a Beta Feature at our four partner Wikipedias: Arabic, Dutch, French and Hungarian. This version's notable features include: a new visual mode for writing comments, complete with a new way to search for and ping other people.
Prior to the 17-June deployment, we completed a series of usability tests with Junior and Senior Contributors to evaluate whether they found the new features being introduced in Version 2.0 intuitive. You can review the results of these findings on the project page: Version 2.0 usability testing.
The usability test findings, usage data, and comments from people who have been using the Reply Tool, are giving us confidence the tool is ready to be deployed more widely. Conversations are now active at our four partner wikis about making the Reply Tool available by default there.
Anyone can try Version 2.0 of the Reply Tool at their Wikipedia by adding the following parameter to any talk page URL: ?dtenable=1.
On 6-July, the new requirements for user signatures were turned on. This means attempts to save new signatures that meet any of the following conditions will be rejected:
Signature does not contain a link to a user's user page, talk page or contributions page at that wiki
Signature contains "nested" substitutions
Existing signatures that do not meet the new requirements will be unaffected for now. Eventually, all signatures will need to meet the new requirements. Before this happens, every person who has a signature that does not meet the new requirements will be notified.
All of the above was announced in Tech News; announcements were also made at various projects to increase the likelihood people who follow wiki signature policy were aware.
You can see an overview of the process in this Phabricator task: T248632.
Other talk page initiatives
Editors at Basque, French, Hebrew, Persian Wikipedias, French Wiktionary, and Portuguese Wikiversity might find a different appearance on their talk pages.
You can learn more about the research that informed the Web Team's decision to implement this change and ask questions/provide feedback about it by going here:
In the next couple of weeks, this new edit conflict resolution interface will become available as an opt-out preference on the Wikipedias listed below.
Reply tool designs: Version 2.0 of the Reply tool is ready for testing and we would value hearing what you think about it. To test this new version of the tool, please review the testing instructions here:Version 2.0 test instructions.
Version 2.0 usability testing: in an effort to make sure the tool functions as people expect, we are running three usability tests:
A test with more experienced contributors at mediawiki.org.
A test with people new to using Wikipedia talk pages on usertesting.com.
A test with people who have already used the Reply tool in production on our partner Wikipedias.
Version 1.0 of the tool is now available for testing as a Beta Feature on mediawiki.org. To turn it on, visit Beta Features in your preferences and make sure the box next to "Discussion tools" is checked (☑️).
Next: Once we have evidence Version 2.0 of the tool is working well and meeting peoples' expectations, it will be deployed as a Beta Feature on our partner Wikipedias (Arabic, Dutch, French and Hungarian).
Initially, we will focus on improving the following aspects of the new discussion workflow:
Adjusting the language throughout the workflow to make it more discussion-specific.
Enhancing the editing tool so people can write what they want to talk about without needing to learn or know about wikicode.
Introducing subtle automations (e.g. automatically signing posts) that guide people towards writing and posting topics that make it easy for others to understand and engage with.
You can read more details about this approach and examples of interventions we are exploring here: Design approach.
The scope of changes to new user signature requirements will be limited to those that directly affect the reliability and availability of DiscussionTool features. These changes are:
Unclosed HTML tags will be not be allowed in signatures
Some uses of subst: markup and tildes will not be allowed in signatures
A link to a user page, talk page or contributions page will be required and at least one of these links will need to be a local page.
An important part of making these changes is that everyone who is affected be made aware of them. As such, we have put together an implementation and communication plan which you can review in Phabricator here: phab:T248632.
Next: the finalized changes and the plan for communicating these changes will be posted to to the proposal page: New requirements for user signatures.
Trying DiscussionTools
If you would like to see your community try DiscussionTool features next, please contact User:Whatamidoing (WMF).
Summary
The team has been focused on the following projects. More details about each can be found below.
Replying tool: on 31-March, Version 1.0 of the new Replying tool became available as a Beta Feature on the following partner wikis: Arabic, Dutch, French and Hungarian.
Starting a new discussion: the project page has been published for the next feature we will work on, an easier and more intuitive way for starting new discussions on talk pages.
Reply tool designs: we published the proposed designs for Version 2.0 of the new Replying tool and we would value hearing what you think about them. You can review the designs and share what you think of them here: Design feedback: version 2.0 mockups.
Starting a new discussion: Are there aspects about how you go start new conversation threads you find take a lot of effort and would like to see improved? Is there functionality you want to make sure this new tool does not affect? If so, please share them on the talk page here: Current ways of starting new discussion threads.
For ~4 weeks, Version 1.0 of the new Replying tool has been available as a Beta Feature on our partner wikis. While there is more work to be done, usage and explicit feedback, like the quote below, has been positive:
"I am now very enthusiastic, good development. Delicate, intelligent, does not overwrite the usual things…" | source
Next up: we will be sharing and inviting feedback on a prototype for Version 2.0 of the new Reply tool which will introduce a new visual mode as well as a simplified way for pinging/@-mentioning other people.
To better understand how people who are new to editing Wikipedia experience starting new discussions on talk pages (article and user) we conducted four rounds of usability tests.
Here is some of what these tests uncovered:
None of the test participants did not sign the discussions they started
Test participants did not understood how talk pages worked and many found it not clear where the new discussion they started would be shown and to whom it would be visible.
Next up: we are starting work on an initial set of designs that we will share and invite feedback on on the project page: Talk pages project/New topic.
After 4 weeks and 150+ comments, we refined the scope of new signature requirements.
Next up: we will be sharing a plan for what changes will be made and when they will be made. This information will be shared ahead of time, so people directly affected by this change can make the necessary adjustments to their signatures.
Other talk page improvements
We thought people following this project would be interested in knowing about another talk page-related project Wikimedia Deutschland'sTechnical Wishes Team is working on: an improved interface for resolving edit conflicts on talk pages.
This feature is scheduled to be deployed to an initial set of wikis in May and on all wikis sometime in June or July. Once the exact deployment date has been set, we will share it here.
Trying DiscussionTools
If you would like to see your community try DiscussionTool features next, please contact User:Whatamidoing (WMF).
Version 1.0 deployment: The deployment of the the new replying workflow as a Beta Feature on Arabic, Dutch, French and Hungarian Wikipedias will be delayed for an unspecified amount of time. The team will share an update on this page as soon as a new date is set.
Over the past month, the team has been focused on the following:
Mockup: this mockup shows a proposed design for version 2.0 of the new Replying feature.
Version 1.0 deployment: Version 1.0 is scheduled for deployment on Monday, 16-March-2020 as a Beta Feature on four partner Wikipedias: Arabic, Dutch, French and Hungarian.
The next initiative the team will be working on as part of the talk pages project will be improvements to the workflow for starting a new discussion. This initiative is a part of the team's broader effort to make it easier and more intuitive to participate on talk pages.
Usability testing: completed an initial round of usability testing to inform what improvements we prioritize to make it easier and more intuitive for Junior Contributors to start conversations on talk pages.
To make it easier to reply to specific comments on talk pages, the team published, and started inviting feedback about, a proposal to codify some community-defined signature requirements into Wikipedia's software.
Over the past few weeks, the team has been focused on the following:
Partner wikis: finalizing the wikis we will be partnering with to trial early iterations of new talk page features.
Replying to comments: preparing for the release of version 1.0 to our partner wikis in the next 2-3 weeks and finalizing the requirements for version 2.0.
Infrastructure improvements: preparing an RfC to determine what wikitext syntax should be used to demarcate multi-line comments.
Partner wikis
At the end of 2019, we invited four wikis to work closely with our team to trial early iterations of new talk pages features. In January 2020, those wikis concluded their community conversations. We are now excited to share Arabic, Dutch, French and Hungarian will be partnering with the Editing Team for this project. This means, new talk page features will be deployed to these wikis first, beginning with version 1.0 of the new replying workflow in the next 2-3 weeks.
On-wiki testing: over the past 6 weeks, 15+ contributors have tested and shared their feedback about version 1.0 of the replying prototype. Their feedback has been helpful in identifying issues that needed to be resolved and refining the design of the feature itself. For more details about what came out of those conversations, see this thread: Topic:Vcwvt3bq03o5gv8h.
Version 1.0 deployment timing: In the next 2-3 weeks, Version 1.0 of the new replying workflow will be deployed as an opt-in Beta Feature on Arabic, Dutch, French and Hungarian Wikipedias.
Version 2.0: the enhancement people are most likely to notice in Version 2.0 is a way for drafting replies using a rich text editor. To learn what other enhancements to expect in Version 2.0, please visit: Talk_pages_project/replying#Versions.
Infrastructure improvements
Currently, contributors are limited in the types of components (e.g. tables) they are able to embed in lists (read: indented talk page comments). This prevents them from being able to communicate about and collaborate on things like navboxes, series templates, and infoboxes within indented comments on talk pages.
Introducing new wikitext syntax to demarcate multi-line comments would remove this limitation and also create additional benefits like being able to more reliably edit specific comments.
The team is preparing an RfC to determine what this new wikitext for multi-line syntax ought to be. We expect for the commenting period to begin in the next two weeks on MediaWiki.
Summary
Over the past few weeks, the team has been focused on the following:
Replying to comments: refining version 1.0 of the prototype, completing a usertesting.com test of the version 1.0 of prototype with Junior Contributors and started on-wiki testing with contributors who have more experience using talk pages. More information: replying user testing.Screenshot of version 1.0 of the new workflow for replying to talk page comments.
Partner wikis: starting conversations with select wikis about partnering to try early iterations of new features.
Technical decisions: sharing an overview of the underlying technical decisions the team has made so far. More information: discussion.
We need your help
In the last update, we mentioned there would soon be a prototype of the new workflow of replying to specific comments ready for public testing. That prototype and the structured user test is now ready. If you are interested in trying an early version of the new software, the team would value your participation in the user test.
Replying in long conversations:here, contributors are talking about how the new replying workflow might be improved for conversations where there are many replies.
In February 2020, the team is planning to deploy version 1.0 of the new replying workflow to a group of partner wikis. This limited rollout is an effort to make sure the tool is behaving in ways contributors expect and find useful before it is deployed more broadly.
We have invited an initial set of wikis to partner with the team on this work and they are now discussing the opportunity in their communities. We expect to have more details to share about which wikis we will be working with in late January/early February, 2020.
Since the last update, we have completed two usertesting.com tests with Junior Contributors. One test sought to identify the challenges Junior Contributors face when trying to participate in conversations on talk pages using the existing tools (read: full page editing). The second test sought to understand how version 1.0 of the prototype changed Junior Contributors' experience replying to comments. Results for both tests can be found here: User testing.
The team has also started on-wiki testing for the version 1.0 prototype that we would value your participation in. More details can be found here: Replying version 1.0 prototype test.
Summary
Over the past few weeks, the team has been working on the following:
Partner wikis: defined a set of criteria we will use to identify the wikis we work closely with to design and test early versions of new features.
Replying to comments: prototyped an initial version of the improved workflow for replying to comments on talk pages.
Understanding the impact of talk page usage: completed the initial stages of a collaboration with the Research Team to better understand the impacts of talk page usage.
We need your help
We would value your input on the questions below. If you have answers to any of the questions, please say so on the talk page: 27 November 2019 update.
Talk page research: What questions do you have about the initial talk page usage research? What refinements do you think we should consider making to the next iteration of this research?
Talk page usage metrics: How might we measure/monitor whether the changes to talk pages made as a part of this project promote "unproductive behavior" on talk pages?
In an effort to increase the likelihood the tools we build to improve on-wiki communication meet the needs of the range of people these tools are intended to serve, we are striving to partner with a limited number of Wikipedias to design and trial early versions of new features.
The team is currently putting together a list of candidate wikis we think would make for good "partners" with the following criteria in mind:
Do contributors to this Wikipedia depend on talk pages to collaborate?
Do contributors to this Wikipedia see improving talk pages as a necessary and important endeavor?
Among the Wikipedias the Editing Team partners with, is there at least one wiki whose contributors use a [Right-to-left language] on their wiki?
Among the Wikipedias the Editing Team partners with, is there at least one whose contributors use a non-[Latin script] on their wiki?
We now have an initial version of a prototype that supports replying to specific comments running on an internal test wiki. Over the next couple of weeks, we will be making iterative improvements to the prototype before posting it on the project page for feedback as well as conducting an initial round of usability tests using usertesting.com.
There are still open design questions to be answered. You can read more about these and share your input in the soon-to-be created discussion: Talk:Talk pages project/replying.
Understanding impacts of talk page usage
The team has been working with the Research Team to better understand what – if any – impact talk page participation (across namespaces) has on contributor productivity. In this initial research project, we have translated this broader curiosity about talk pages' importance into the following research question: Do interactions on talk-pages increase or decrease contributions to edits on articles?
Our initial analysis suggest:
Edits to talk-pages seem to create additional edits to subject-pages
Edits to user-talk pages have a negative effect on the number of future edits to the main namespace (en.wiki)
For other wikis, no such negative effect can be found and article-talk pages play a more important role
To learn more about these initial results, as well as the methodology used to arrive at them, we recommend reading this report: Research: Usage of talk pages.
Also, if you have any comments, questions and/or thoughts about this research, we would value you sharing your thoughts here: 27 November 2019 update.
---
That's it for this update. If you have any questions, comments or thoughts about anything included, please add them on the talk page:27 November 2019 update.
Started collaborating with user:Enterprisey (author of en.wiki's reply-link gadget) who has been, among other things, helping the team identify edge cases we will need to design for
In the 15-October update, we shared the goal of this project is to evolve talk pages in ways that help contributors work together more effectively.
We now have a set of metrics to quantify this objective. Those metrics are:
Increase in the number of JuniorContributors participating on talk pages
Increase in the retention of JuniorContributors participating on talk pages
Increase in the number of productivecontributions to articlepages made by Junior Contributors
More details about how terms in the above are being defined (e.g. "productive contributions) can be found in this task: Create OWC metric definitions. These goals and definitions will also soon be posted to the main project page and project glossary.
The first enhancement to talk pages will be an affordance intended to make replying to specific comments on talk pages easier and more intuitive for contributors across experience levels.
We now have a set of specifications for v1.0 of the feature as well as a rough idea for how it might evolve. For now, links to the specifications for each version can be found in this task: T233443
In addition to finalizing the specifications for v1.0, we also now have designs for this first version. For now, the designs can also be seen in Phabricator, in this task: T235592
The software can now "read" time and dates. This is a big step that will make it possible for talk pages to do things like:
Showing you the time and date comments were posted in your local timezone and
Showing you new comments posted to the talk page since you visited it last See the 27-September-2019 update for an introduction to this work.
Text input for drafting a reply | Follow along: T234403
In order to reply to a comment, there needs to be a way for contributors to write out their comments. Over the past couple weeks, the team has started prototyping a text-input to make replying possible.
The first version of the text input will support wikitext only. In time, a rich text editing mode will likely be added.
There are open questions about how multi-line comments should be saved and what should happen when contributors attempt to insert a table inside a comment. These are questions we will explore on the project page: Talk pages project/Replying
That's it for this update. If you have any questions, comments or thoughts about anything included, please add them on the talk page: Discuss 5-Nov Update.
Summary
Over the past two weeks, the team has been working to:
Define the goals the team will strive to achieve with this project
Programmatically identifying signatures on talk pages
Supporting autocomplete which could be applied to things like seamlessly pinging/@-mentioning a specific user while writing a comment
We need your help
A few questions came up in the course of our work that we would value your input on. If you have answers to any of the questions below, please say so on the talk page: Discuss 15 October 2019 update.
1. What templates and/or gadgets do you use most on talk pages?
Context: the team wonders whether these gadgets will give deeper insight into how contributors use and/or would like to use talk pages.
2. What non-Wikimedia projects/services do you think handle commenting/discussions well? Reddit? Discourse? Something else? What about these product(s) do you value?
Context: as the team begins to design the first set of improvements, they are curious to know what functionality you have found to be helpful in contexts similar to talk pages.
3. What do you type on a talk page when you want to notify someone in your post?
Context: the team is looking for conventions that might be specific to certain wikis. More information in this ticket: T234472
It is one thing to make certain workflows "easier" (i.e., replying to comments, indenting, and signing your name), but it is another to make sure those changes amount to contributors working together more effectively. | Source
It is with this in mind, that the team is working to define goals that honor that intention and a corresponding set of quantitative metrics they will use to track their progress towards achieving them.
For Talk pages to be valuable, contributors need to intuitively know how to participate in conversations.
Trouble is (as the Talk Page consultation and the team's research over the past couple of weeks uncovered) contributors, across experience levels, find replying to specific comments on Talk pages to be challenging.
Specifically, the team's research has found:
More experienced contributors find it difficult to locate the specific comment they are wanting to reply to when participating in long, multi-person conversations, within big blocks of wikitext.
Newer contributors report not being sure how to reply to a conversation, regardless of its length or complexity.
The team is now drafting specifications and beginning to develop designs for the first iteration of an enhancement that would make replying to specific comments on talk pages easier and more intuitive.
Technical research
The team is continuing the work on parts of the underlying architecture future features will rely on.
The next phase of this work will involve enhancing the tool so it can "read" the dates and times included in signatures. Doing this, will enable functionality like showing contributors when posts/comments were made on talk pages, in their local timezone and preferred date format
Text input for drafting a reply | Follow along: T234403
One of the first improvements the team plans to work on is a way for contributors to reply to specific comments on talk pages.
There are several components that will need to be designed and built to make this possible.
One of those components is a place (read: a "text input") to write a reply. Right now, the team is researching "text inputs" that have been built in the past to help inform their approach for doing something similar on talk pages in the near future.
Summary
The team has been, and is continuing to work on:
Design and product research: forming a better understanding of the most common talk page use cases and the opportunities for improvement within them.
Technical research: investigating existing gadgets and developing early proof of concepts for parts of the underlying technical architecture future features will depend on.
Getting involved
Do you have a question? Is there something you think the team could benefit from knowing? If so, please say something on the talk page: Discuss 27 September 2019 update
Design and product research
The team is working to form a more complete understanding of where and how contributors communicate on-wiki.
Right now, they are cataloging the most common use cases of article and user talk pages and the pain points and moments of delight contributors experience using them.
This work will help us to develop a roadmap. That is, an overview of the specific set of improvements we will propose working on and the order in which we are thinking about working on them.
You can see this work in more detail by visiting the following links:
Journey maps: this is where the team is creating visual representations of specific use cases, the steps that make up these use cases and the opportunities for potential improvements we see within them.
Research: this is where the team is cataloguing the use cases, design patterns and social conventions that help to make up "on-wiki communication."
Technical research
The team is working on two parts of underlying architecture future features will depend on:
One of the biggest technical challenges of making improvements to talk pages is their unstructured nature.
Without a clear structure, it is difficult to build software that can do things like enabling you to reply to specific comments, receive notifications when someone replies to a comment you've made, receive notifications about a conversation you are interested in following, etc.
Right now, the team is developing a proof of concept to identify signatures on talk pages. Signatures contain important pieces of metadata like where a comment ends, who is speaking and the order in which people are speaking...information that is needed to build features like those mentioned in the bullet point above.
Autocomplete can speed up how quickly you can express something by predicting what you are trying to type before you have finished typing it.
Right now, the team is developing the infrastructure to enable this kind of functionality on talk pages.
Functionality that could make it possible to seamlessly ping/@-mention a specific user while typing a comment on a talk page.
Summary
Members of the Editing Team were in Stockholm, Sweden to attend this year’s Wikimania. We arrived with a few questions in mind:
What are contributors' current experiences with Talk Pages? What could be improved?
How have the Foundation's past efforts to improve on-wiki communication impacted contributors?
What new questions will the team need to consider answering?
To help answer these questions, we engaged with Wikimania in the following ways:
We led a session in the Community Growth Space to share the Talk Page Consultation's findings and outcomes
We had casual conversations with contributors from around the movement
We met with the Foundation’s Parsing team to explore how wikitext syntax might be evolved to enable some of the improvements that surfaced in the Talk Page Consultation
We attended sessions relevant to Talk Pages throughout the three conference days
Below are our findings and notes from this year's conference...
After flow: a new direction for improving talk pages
The team led a session in the Community Growth Space to bring attendees up to date on what we've learned from the Talk Page Consultation and to discuss some of the decisions and tradeoffs we will collectively need to make to improve how contributors collaborate using talk pages.
Around 35 people attended the session who, generally, agreed talk pages need to be improved and were glad the WMF is prioritizing this work. With this said, some people expressed skepticism about whether the Foundation would follow through on completing the improvements we are proposing. More specifically, attendees were clear about how – in the past – the Foundation has promised changes to improve on-wiki communication and collaboration that have not been fully delivered (e.g., Flow).
The Q&A that followed the presentation was lively. People asked about the future of Flow and asked for more details about the improvements to existing talk pages. Notes from the Q&A portion of the session will be posted to the project FAQ. In the meantime, they are available in this Etherpad: AfterFlow session notes
Session takeaways:
Communicate and follow through. We need to acknowledge we do not have people's full trust. We need to appreciate how the Foundation promised big changes with Flow and has not been able to deliver on all of them. In response, we need to make improvements more iteratively and communicate any changes about these plans early, often and thoroughly.
Create spaces for conversations. During the Q&A, there were many questions we – the team – do not yet have answers to. Questions like: What improvements will be implemented? How will they be implemented? When will they be implemented? These are valid questions that need to be answered. They are also valid questions we do not yet have definitive answers to. In being open about not having all the answers, we noticed people in the audience started trying to help. This is exactly how we think this work needs to feel and flow...for us all to be asking and answering questions together, as partners. This interaction was a good reminder. It reminded us that we need to be intentional about creating environments where diverse groups of people feel invited to speak up. And we think this will partly depend on us continuing to be open about what we do and do not know.
More communication doesn't mean more collaboration. Attendees highlighted the relationship between communication and collaboration and how more of the former doesn't mean there will be more of the latter. To quote one attendee who put this well by saying there is a, "...tension between improving communication and collaboration on Talk pages vs impact on growth and contributions." The full context of this exchange is posted here.
Causal conversations
Following the team’s session in the Community Growth Space, an impromptu and productive feedback session happened with experienced contributors from Hungarian, German, Swedish and English Wikipedias.
Highlights:
Often times contributors are wanting to find the specific conversation where a decision was made about the content of an article, but that conversation is difficult to find. Being able to more easily search archived conversations could help this. This was something that came up in the Talk Page Consultation as well: #Searching.
Contributors who focus on helping newcomers learn Wikipedia observe newcomers write “everywhere” looking for help. One person suggested communication on the wikis be centralized to make a system that can be more easily understood by newcomers.
In relation to some of the criticism about unfulfilled promises, contributors generally agreed that an approach of incrementally developing features seems like a good idea.
Editing Team + Parsing Team
Background
Considering the direction that came out of the Talk Page Consultation is to build new features atop wikitext talk pages, it is important we start conversations with the Parsing Team early to understand what technical considerations will need to guide how the team designs and builds these new features.
Meeting details
At Wikimania, the team met the Parsing team to discuss what changes to wikitext ought to be considered in order to deliver on some of the changes that came up in the Talk Page Consultation. Changes like watching and being notified about replies to specific discussions, replying to specific comments and automatically appending signatures to content contributors post on talk pages.
To build features like this, the software will need to be able to effectively know where comments begin and end. One idea for doing this is introducing new wikitext syntax to make content on talk pages more "machine readable".
Takeaways
A number of syntax changes were proposed and are some are documented on Phabricator:
List item attributes (e.g. for thread IDs): T230658
The conversation surfaced some new questions we will need to answer:
To what extent do new tools need to be compatible with existing conversations?
More details: by "existing conversations" we mean: active conversations on talk pages and conversations that have been archived
Context: if we introduce new features that depend on new wikitext, these tools might not work with existing content that does not contain this new wikitext markup.
It is generally felt that the tools should work reasonably well with "existing conversations" but that it would not be essential to apply any syntax changes to existing conversations.
How does the software determine whether a page contains discussions or not?
Ideas: we could use a magic word to explicitly communicate "talk" on a page
All Talk? The effects of easier communication interfaces (than user talk pages) on collaborative production
In this session, researchers shared their findings about how Fandom's/Wikia's Message Wall feature impacted new editor behavior. The researchers were specifically interested in understanding whether the feature's release led to an increase in the number of article contributions from new editors.
Takeaways
Presenter's conclusion: while the release of Message Wall* increased communication among all editors and newcomers, the feature decreased article contributions from new editors.
The Editing Team's takeaway: this study's findings helps validate feedback that was shared in the Talk Page Consultation and something the team has been mindful of ever since: making it easier for contributors to communicate does not necessarily mean people will make more contributions to articles.
It was interesting to see how this conversation about the relationship between communication and collaboration resurfaced among attendees of the team's After Flow... session. Here is a snippet of an interaction that surfaced on the session's Etherpad: After Flow Etherpad:
Person A:"It would be good to consider the tension between improving communication and collaboration on Talk pages vs impact on growth and contributions. There was a research session this morning where increased communication bt users on Message Walls in Fandom/Wikia saw *decreased* article contributions?"
Person B:"I think something like this really depends on what your ultimate aim with improving communication actually is ... I didn't see this presentation but there are more things to be done and communicated about on the projects than raw article growth. -foks"
Person A:"Agree. I think that's is part of the question the team posed "How do we measure collaboration?" - is it about making it easier for newcomers and experienced editors to communicate and make connections 'healthily' instead of getting into edit/revert wars (in which case less article contributions may be a positive sign of reduced conflict)"
*It should be noted that Fandom's/Wiki's "Message Wall" feature is the equivalent of user talk pages on Wikipedia.
.png)
.png)
