This is the last weekly sharing at my company. I mainly talked about how I handle Jira tickets.
Principles
- Focus on one thing at a time
- Avoid start from scratch
- Don't repeat yourself
Focus on one thing at a time
Every time I'm going to process a Jira ticket, I save it to Notion with the Notion Web Clipper extension, then apply a pre-defined template to the Notion page. For example, there are four basic templates for Jira tickets: Bug, Task, Story and Epic, I apply the template according to the type of ticket, then snippets or scripts are auto filled, it really saves me large amounts of time.
The most interesting part is for Story and Epic. I create a Workona workspace for them when I begin to develop, it helps me focus on one thing at a time. When things are done, I archive the workspace for later reference, I can recover all tabs, notes and tasks at any time.
For Epic, I also create a Board View in Notion, then link all related pages to the Epic, it helps me to quickly get a glance on how things are going.
For each In Progress ticket, Zapier helps to create a GitHub issue automatically. Later I either convert it to a pull request or copy-paste the contents from Notion to help others know how to handle such tickets in the future.
Avoid start from scratch
There are always similar tasks that have been done before, so take advantage of existing solutions is better than start from scratch.
Search before going
I search Jira or GitHub before I start to process, it can reduce much time on investigating, debugging and developing.
I created some custom queries to ease the search with Alfred's Web Search functionality:
I search Jira or GitHub before I start to process, it can reduce much time on investigating, debugging and developing.
I created some custom queries to ease the search with Alfred's Web Search functionality:
- Search Jira tickets
https://JIRA_CLOUD_DOMAIN/secure/QuickSearch.jspa?searchString={query}
- Search my own issue
https://github.com/issues?q=is%3Aissue+is%3Aopen+assignee%3Agoofansu+archived%3Afalse+{query}
- Search the whole GitHub
https://github.com/search?q={query}&ref=opensearch
Utilize the command line tools
- Install command line tools:
- Checkout any pull request:
git branch --all | grep -v HEAD | string trim | fzf | read -l result; and git checkout "$result"
- Checkout any commit:
git log --pretty=oneline --abbrev-commit --reverse | fzf --tac +s -e | awk '{print $1;}' | read -l result; and git checkout "$result"
Don't repeat yourself
- If some scripts are always used, I create a rake task and submit to the repo.
- If some scripts need to run among several servers, I create ansible playbooks, execute a command and waiting for results coming to my machine.