Filter Cheatsheet
term
Matches a case-sensitive substring.
"two terms"
Allows querying with spaces by using a quoted string.
/directory
Filters items that belong to a specific directory or iteration.
#hashtag
Identifies items that include a specific hashtag.
@mention
Finds items that mention a specific person.
or
Introduces a logical OR operator to stop the previous query and define another one for finding more items.
-token
You can negate any token to exclude it from your results. e.g. Handy to hide closed items.

🔖 open/add_a_cli_tool_for_combining_issue_cli_with_git.md

Add a cli tool for combining issue cli with git

A single command for working with issues to create branches and commits based on issues.

Use cases

Create a new branch and an empty commit with the issue as the commit message.

Examples

git issue ./issue/open/some-issue.md

# git checkout "issue/some-issue"
# git commit --template <(uses a commit message friendly version of the issue)

@bas080 is trying a quick and dirty bash version on his local machine.

author:@bas080 priority:low

🔖 open/improve_state_module_performance.md

Improve state module performance

All listeners are called twice for each state push. This is not ideal as multiple pushes can occur on one event. Let's say you have 5 listeners and 2 pushes. The amount of calls end up being:

listeners = [1,2,3,4,5]

calls = [
  firstPushState,
  ...listeners,
  ...listeners,
  secondsPushState,
  ...listeners,
  ...listeners,
]

This double walking over the listeners has to happen because state of one listener can be the dependency of another listener.

A more ideal situation would be to compose all the push call within that tick and then walk the listeners twice.

calls = [
  firstPushState,
  secondsPushState,
  ...listeners,
  ...listeners
]

We can cut down the amount of calls by quite a bit. Especially when the amount of pushes is on the high side.

Conceptually this should be backwards compatible as these amount of passes should allow for the state to stabilize.

#browser #performance

assigned:@bas080

🔖 open/create_a_release_on_tag_creation.md

Create a release on tag creation

It #must create a new release by pushing a new git tag.

It #could also populate the release body with release notes that are generated using some git changelog tool. An option for such a tool could be npm:auto-changelog.

@rage has done this before and shared his previous work and experience.

Requirements

  • The commits since last release do not contain fixup or wip commits messages.
  • Update the changelog with a git log to changelog tool.
  • Creates a github release with the cli build for that specific tag version.

author:@bas080 priority:medium

🔖 open/add_a_negate_operator.md

Add a negate operator

If you would for example exclude all /closed items you could write -/closed

Candidates

- and !

I have implemented the not for negation and decided to instead opt for the -.

Currently this behavior is implemented but not document in the usages.

Tasks

  • [x] Document negation.
  • [ ] Give visual feedback in filter buttons that it is a negated token.
  • [ ] Have the UI report correctly on count matches for negated tokens.
🔖 open/make_it_easy_to_link_issues_to_one_and_other.md

Make it easy to link issues to one and other

With github issue tracker you can use the # to reference other issues.

The hashtag is used for labels. Either rethink this or use something else to reference issues.

Both @rage and @bas080 agree that ! is a good candidate.

For this to work we also need a way to assign the id to an issue file. Possible implementations:

  • Edit the file meta data to include an id. This requires the use of the cli and is less transparent. Suggested by @rage.
  • Have it be part of the filename. This requires a convention to be followed.
  • Have each issue be placed in a directory that has an issue id. /open/abc12/my-issue.md. This also allows bundling possible dependencies like images within the issue.

priority:high

🔖 open/support_passing_a--templatetoissue_html_.md

Support passing a --template to issue html

issue html --template=./issue/template.html

In this example we stored a template in the issue directory to be used to generate a richer and interactive issues webpage.

The template contains a comment <!--issues--> that is replaced with the output of issue html.

🔖 open/do_not_match_on_content_when_searching_with_slash.md

Do not match on content when searching with slash

When searching for /open it should not match on arbitrary /open strings. It should only match on the file path.

Other search operations might also need reconsideration but this issue is only concerned about the path search.

🔖 open/use_the_ocaml_build_to_generate_pages_html.md

Use the oCaml build to generate pages html

Currently the perl script is used in the .github/workflows/pages.yml workflow. We want to use the new and shiny oCaml implementation in the workflow.

We need a way to download a specific version of the oCaml build and use it to generate the issue html.

author:@bas080

🔖 closed/do-not-use-hash-to-store-query.md

Do not use hash to store query to leave it for issue bookmarks

The # in the window.location is used for anchors with hrefs that start with #.

Instead use the well known convention of using the q search param.