Welcome to the roundup of GitClear Q3 upgrades and improvements! With 100k of Diff Delta in the past quarter, our recent progress is 43% greater than the progress made by the Facebook React team over the same time range. We've got a lot to cover, starting with our biggest feature debut since pull requests:
link🎯 Headline launch: Goals
After 5 years of building visualizations, we have only too many ways to visualize the data that dev teams care about. Whether it's "repo progress," "pull request collaboration," or "DORA-related" metrics, new users often report that we offer more data than they have time to explore. That should be a good thing?
But it's not good if the team's most actionable or applicable data is buried on a sub-tab of a page you don't have time to visit.
Enter GitClear Goals. This marks the beginning of a new era of GitClear, where data presented to you is sorted by how relevant it is to your role and your stated goals. If you can enumerate what your "ideal dev team" looks like, we will help you become that team. Our new Goals Setup Instruction page illustrates the breadth of goals that can now be pursued:
Already more than 40 team goals available to choose from
Once you've set your sights on your preferred targets, we will notify you with specific details when your performance falls short of the intended target. You can also see any graphs available that correspond to the goals you've set by visiting the new "Goals" tab under "Highlights":
Charts that correspond to the goals picked by our team
Soon, we will begin showing a team's weekly goal history on the "Highlights" => "Goals" tab, as well as on the forthcoming "Performance Review" tab.
link🗺️ CAB Dependency and Migration Detection
Not all commits are created equal when it comes to their reliance to a Lead Developer.
Now, when a commit introduces a database migration, or changes the dependencies for a repo, these commits will receive special branding:
These should be automatically detected. If your project uses a migration/dependency that’s unique, you can define it in your code categories. If you use a conventional approach and your commits aren’t automatically recognized, drop our support a line and we will update our detection algorithms.
link📈 Google DORA section
Instead of spreading DORA charts across 3 tabs, why not present them together? We have spent a lot of time refining our DORA infrastructure this quarter, and consolidating the measurements was one key part of the effort.
Equally useful: DORA setup can now be completed in a single tab. If you haven’t taken time to decide on what constitutes a “Critical Defect” yet, drop us a line at support@gitclear.com, or just click here and consider what measurements should qualify as “key importance.”
Currently, the DORA Stats page includes the following graphs:
Release Frequency - What it sounds like
Critical Defects Detected - Raw count of critical defects detected during the time range chosen
Mean (Business) Hours to Resolve (MTTR) - How many business hours elapsed between when critical defects were detected and marked "resolved" during the chosen interval?
Release Cycle Time - How long did it take to usher a pull request from its first change to being available to users?
PR Cycle Time - How much time passed between the first change for a commit and its PR being merged?
PR Lead Time - Similar to "Release Cycle Time," but averaged by pull request vs. averaged by "issue resolved."
link🏝️ Planned absences/PTO tracking
With the variety of goals offered to detect anomalies in code delivery, it has become important to algorithmically distinguish between “gone on PTO” and “MIA.” To that end, we now offer an API call, a Contributor Setting option, a CAB button, and email links to easily designate when a delivery anomaly is just PTO.
Read all about setting planned absences on its help page.
link🔍 Pull request time breakdown
Now show a breakdown of how much time was used in each phase
From the newly combined "Pull Requests" => "List" tab, which by default includes pull requests that are currently under review alongside those that were recently completed.
This breakdown should make it much more straightforward to see how long each pull request was stuck during each phase of its evolution.
link📣 Goal ideas?
Drop a line in the comments if there is anything you have wished you could instrument? Chances are, we are already recording what you want to know, or a close proxy for it.
link🔮 What’s next?
Things don't get any slower from here. Now that we goal tracking is available, we need to ensure that they're easy to locate and use.
We will soon begin showing a weekly goal history for each goal when visiting "Highlights" -> "Goals."
We are crafting an LLM built to better summarize code and review PRs compared to the suggestions offered by OpenAI.
Chart Glimpses (public publishable, optionally updating) will be implemented for almost every goal. Chart Glimpses have already been implemented for all of the DORA goals.
Speaking of Chart Glimpses, they are on the way to being upgraded such that developers can easily drop a screenshot on each of their achievements. This will let Chart Glimpses function as a semi-automated changelist, where either a developer or a product manager can handle translating "commit message" to "public announcement." The developer screenshots will not only inspire customer confidence (for customers using the glimpse as a Changelog), it will also let the developer themselves savor how much progress they've made over the past year.
New research on how AI is impacting code quality will be released before the end of Q4.
And of course, continued incremental evolution of pull request review tools. If you are paying for GitClear, you use pull requests, and you’re not using GitClear yet, why not? You have nothing to lose and 30% more time to gain.
Thanks for keeping up with our progress!