We should further have a think about what worked more or less well with our old structure and what structure we want for the current projects with its two (fairly distinct) subparts.

This is a blocker for creating new tasks

Reflections on past setup:

Terminology: From the pilot document Manusskapare would be a Component and Manusinput (under it) would be a Function

Initial set up of tasks
Before:

• We created story tasks for each component mentioned in the pilot study and broke each one down into further tasks. This was all done in the intiial setup.

For the new setup we suggest:

• We should not mass create tasks all the way down, this makes it inflexible when we need to change how something is done and means tasks are easily lost in a sea of unstarted tasks. Creating initial umbrella tasks for each Function should work. When work on a Function starts the first task becomes to break it down into smaller tasks (some of which can be broken down further once work on them starts). Function tasks should include info from the pilot document including responsible partner and during which phases it is planned to be worked on.

Projects/Sub-projects/Milestones
Before:

• Due to technical limitations Wikispeech was originally set up as a set of different projects (see description) per component. * Sprints were milestones under the central project.
• Tasks were additionally tagged with the responsible group (e.g. Wikispeech-WMSE).
• Story-tasks were additionally grouped by Phase sub-projects.

For the new setup we suggest:

• A top level project Wikispeech - no board and should in theory not contain any tasks
• Three sub-projects under Wikispeech (Wikispeech-Jobrunner, Wikispeech-Speech-Data-Collector, Wikispeech-Text-to-Speech(?))
  • Wikispeech-Speech-data-collector and Wikispeech-Text-to-speech(the old project) contains one column per Component plus an "unsorted" default column.
  • Wikispeech-Jobrunner has the columns: Incoming (default), backlog (and possibly a proposed for sprint column)
    • Sprint milestones are created under this project.
      • Sprints have a backlog, in-progress and done column (and possibly a proposed for sprint column)
    • Any task in one of Wikispeech-Speech-data-collector or Wikispeech-Text-to-speech projects should also be on the Wikispeech-Jobrunner board.
      • A Herald rule can be set up to add tasks not satisfying this requirement to the Incoming column)
• Tasks should be tagged with one or more of the group tags (Wikispeech-WMSE, Wikispeech-STTS, Wikispeech-KTH)

Left to decide:

• Should we have the proposed for next sprint column in the Jobrunner or in the sprint itself
• What to actually call Wikispeech-Text-to-Speech
• What (if anything) should be in the WMSE-Wikispeech-Speech-Data-Collector-2019 and WMSE-Wikispeech-2019 projects
• How do deal with tasks which don't belong to any Component e.g. "Organise a fika for all of the project partners"

In T230765#5478802, @Lokal_Profil wrote:

• A Herald rule can be set up to add tasks not satisfying this requirement to the Incoming column)

This will have to be created by someone with the appropriate permissions, see https://www.mediawiki.org/wiki/Phabricator/Help/Herald_Rules#Global_Herald_rules.

In T230765#5478802, @Lokal_Profil wrote:

• Wikispeech-Speech-data-collector and Wikispeech-Text-to-speech(the old project) contains one column per Component plus an "unsorted" default column.

Suggestion is that we also add an out-of-scope/for-future-consideration column to keep generally good ideas/new features which are not in scope for the current project but which should not be lost.

In T230765#5478802, @Lokal_Profil wrote:

• What (if anything) should be in the WMSE-Wikispeech-Speech-Data-Collector-2019 and WMSE-Wikispeech-2019 projects

For WMSE-Wikispeech-2019 I'd say we move any remaining tasks to WMSE-Wikispeech-Speech-Data-Collector-2019 then archive it. The decision on WMSE-Wikispeech-Speech-Data-Collector-2019 depends a bit on how we chose to handle purely organisational tasks.

• Proposed for next sprint will be kept in the jobrunner and removed from the sprint board, making backlog the default
• As preparation for each new sprint. Go through and empty Wikispeech:incoming and Wikispeech-Jobrunner:incoming
• Every task in Wikispeech-Text-to-Speech/Wikispeech-Speech-Data-Collector should also be on the Jobbrunner board. With the exception of "Out of scope ones"
• We keep the name Wikispeech-Text-to-Speech
• We keep the name Out of scope
• Administrative tasks should live in WMSE-Wikispeech-Speech-Data-Collector-2019, they can however also be part of the Jobrunner board and sprints and be tagged with Wikispeech-WMSE etc.
  • Empty and archive WMSE-Wikispeech-2019

Updates to the mock:

• Document * and ^.
• Make connection between sprint column and board more obvious
• Add Wikispeech-WMSE etc. to the mock
• Add WMSE-Wikispeech-Speech-Data-Collector-2019 to the mock

New questions:

• What to do with the info encoded in the columns of Wikispeech-WMSE, could we use subtypes to capture this?

In T230765#5585919, @Lokal_Profil wrote:

Updates to the mock:

• Document * and ^.
• Make connection between sprint column and board more obvious
• Add Wikispeech-WMSE etc. to the mock
• Add WMSE-Wikispeech-Speech-Data-Collector-2019 to the mock

Updated the mock. @Sebastian_Berlin-WMSE please take a look to see if it makes sense

Several of the tasks marked as story only have a single task with basically the same information. I merge these together.

I have gone through the old tasks in Wikispeech and subprojects and moved tasks to Wikispeech-Text-to-Speech and Wikispeech-Speech-Data-Collector. Most were put into the "Future development" column and some where closed as no longer relevant.

The old subprojects have been archived.

In T230765#5667058, @Sebastian_Berlin-WMSE wrote:

I have gone through the old tasks in Wikispeech and subprojects and moved tasks to Wikispeech-Text-to-Speech and Wikispeech-Speech-Data-Collector. Most were put into the "Future development" column and some where closed as no longer relevant.

The old subprojects have been archived.

Thanks.

I think it would also be good to have a short description on each of the subprojects. Might also be good to mention the Wikispeech-Jobrunner in the Wikispeech project description (and maybe also WMSE-Wikispeech-Speech-Data-Collector-2019 since some of the info is there).

I also archived WMSE-Wikispeech-2019

Other than that is there anything left to do on this?

Yes, adding/updating descriptions is something that should be done for the current projects.

I've added minimal descriptions to the projects. They will be extended somewhat when T227068 is done by adding links and/or copy pasting some information.