Q&A highlight
Posted20 days ago
Key Takeaways
https://github.com/orhun/git-cliff
https://www.conventionalcommits.org/en/v1.0.0/
This changelog is copied into the release on github, or wherever the release is announced.
Auto-genertaed changelogs lack business-aware context about what is important. You get a big list of new features, but which ones are the most important to stakeholders? You have a few breaking changes, which are likely to have the most widespread impact? Without being judicious about what information is included, you risk overwhelming readers with line noise and burying important notes.
One of OP's motivations for starting this thread was to see how people tailor changelogs to different types of stakeholders; techincal vs non-technical, for example. This approach doesn't solve that problem. In fact, I think it's worse due to an additional side effect: the commits are now forced to do double duty; they must be useful commits for developers looking at code history, but now they also must be useful messages to be included in a changelog. While there is some overlap, it's hard to do both simultaneously. One must pick between writing good commit messages for the codebase & developers, versus writing a coherent changelog.
As a matter of personal taste, I think it looks lazy. Changelogs are a unique opportunity to communicate something important, they're written once and read by many. with a list of commits, myself and all other readers must now put in the work to find out what's relevant - it's disrespectful of others' time.
I worked for one startup with one major customer who was really skeptic of investing further because of stability problems, feature delay problems, and lack of transparency. Along with a complete list of changes that gave them insight into how we prioritised between stability and feature development, I wrote a human summary of what this meant — experiments, summaries of statistics, summary of most important changes to business logic.
Writing personally to your stakeholders does not exclude being systematic, and vice versa.
> As a matter of personal taste, I think it looks lazy.
That’s funny, because I find the lack of automation to be the lazy choice. Forgetting to add to the changelog because the requirement is checked by humans, or because single commits fix things below some bar of noteworthiness that is entirely subjective and driven by lack of structure. Not writing commit messages worth putting in release notes (fix sht, asdasdasd, etc.)
>
Changelogs are a unique opportunity to communicate something important, they're written once and read by many. with a list of commits, myself and all other readers must now put in the work to find out what's relevant - it's disrespectful of others' time.*When I migrate software, I’m very interested in the complete picture. I’ll ask my AI agent to go over the links in the changelog and summaries for me what are the breaking changes and what manual steps do I need to take. Having them in human-readable form ahead of time would be nice.
Since git-cliff has different sections, I can skip changes to documentation. Because of SemVer, I know if there’s something breaking.
> That’s funny, because I find the lack of automation to be the lazy choice.
Automation is cheap these days. Many automations make things that exceed human ability, but this isn't one of those cases. You'll get something good-enough, but not great. Perhaps that's what your organisation has time and budget for, in which case your use of automation makes sense, but if we're trying to make the best audience-tailored summaries of software releases with specific purpose, the strategy falls short.
> Forgetting to add to the changelog because the requirement is checked by humans,
Individual developers definitely can, which is why you must also have organizational process. If a valid changelog checked by a release manager is a requirement for a software release, it can't be forgotten.
> I’ll ask my AI agent to go over the links in the changelog and summaries for me what are the breaking changes and what manual steps do I need to take.
I really think this is backwards and exactly the thing I was advocating against. The changelog you have is so large and poorly-structured that you need to use an AI to summarise it for you, and gather the information that should have been in the changelog in the first place. If that needs to be done to make the changelog useful, clearly its original state is deficient?
If affirmation is cheap, why hire humans to sit and manually validate if a text file was updated? Maybe you have a lot of cheap labor or a lack of automation experts.
> The changelog you have is so large and poorly-structured that you need to use an AI to summarise it for you
https://terranix.org/news-2025-11-09_website_renewal.html
Not really that long and poorly structured, no.
I see our difference in perspective. My organisation has over 100k people
I've found this to be pretty useful for my projects, and users can quickly get a list of what they care about for changes (usually the features part). Since it's pretty automated, the amount of effort also means that even if barely anyone reads them, it's not a huge waste of time. It's actually kind of made me write better, more useful commit messages as I know that the first line of the commit will actually be presented to the user.
v1.4.18 - "Bug fixes and performance improvements"
v1.4.17 - "Bug fixes and performance improvements"
v1.4.16 - "Bug fixes and performance improvements"
While I'm morally tempted to do the same, many of the apps guilty of this are the major ones one uses, and as time goes by, I somehow find myself with less and less time on my hands, so I have to be selective with the things I want to do right and proper. Thus, by means of inaction, I indirectly contribute to the circle of enshittification, and there is no stopping it.
Not only nobody reads them, but Apple forces you to translate them into languages even less than nobody read. It'd be an improvement if they only required English text.
https://apps.apple.com/us/app/chime-mobile-banking/id8362152...
"Choose your own update notes adventure! Pick only one: A, B, or C.
A. The holidays are coming up and we've been busy planning a celebration of bug bashes and performance enhancements, so get merry and smash that update button.
B. Shorter days, longer nights, colder temps. You know what helps the Winter blues?
Instant gratification. Tap update, watch that progress bar fill up, and feel the dopamine flow.
C. Whoever made up the mistletoe thing was crazy. You know what's not? Updating your app."
This... also serves no useful purpose.
Rev number go up!
Nowhere else in society do we allow such self-serving laziness and unethical negligence (looking at you, purposely destroying backwards compatibility of APIs) at a professional level. Most other professions have steep legal consequences if they hide their actions or inactions.
And sometimes they do actually add a feature... but they'll mention it within the app itself despite the app updates not mentioning it. Or even more funny is how often I'll see a news article talking about the new feature, but then it never even gets mentioned in the release notes anywhere.
- One location, people who may fall into multiple categories (or none) don't need to check multiple places, users also know that all my communication will be via that page/they don't have to wonder if they're missing something
- As much as some detail doesn't matter to certain audiences, I find being able to give all the detail you want a user to know while maintaining readability to less technical audiences is a skill worth developing because the result is regardless of where your notes end up, the person will understand what's changed and why it matters
- Maintaining multiple versions leads to mistakes, at some point you'll leave out a detail to one audience that matters so letting the user mentally filter what they don't care about takes the onus to get it right 100% of the time off of you. I'll often categorize my changes by the section that had the change to help users with this.
- This is a personal preference and you touched on this one but it's just far less work, I've found it common in tech that people don't want to do things more than once or they'll automate it/look for shortcuts and this is no different. This isn't always a bad thing but getting release notes right means your users stay informed/use new features which is why we build them so I think it's worth putting my energy into doing it properly every time
The change log for end users comes from the JIRA board for the release, looking at what tickets got closed that release cycle, and it usually requires some amount of human effort to rewrite.
I've used feat/feat(pub):, fix/fix(pub):, etc before to automatically separate changelogs into internal/public.
If features need more explanation, we create a wiki page and link it in the release.
Sometimes we feel like there should be a changelog for devs but in the end git blame is used anyways.
I use Github generated release notes based on Pull Request bodies and this is more than sufficient for my dev team.
I think have a small app in my different internal web application called `releases` (a Django app) that has a simple model called Release. I hand write the release notes, copy and paste the version number, and select date for each release (sometimes I'll batch add them if I get behind).
But I had write them with my audience in mind (business system users - who care about specific bugs or features; i leave out any internal things that wouldn't be of interest).
At the footer, or a small indicator in side bar, I display the version number that was created through our automated release process (Github Release triggers a release and the tag that is the version number gets set as an env var in my Docker image). Clicking that takes the user to a /release-notes/ page that has a descending order list of the notes that I hand author in the Django admin.
Pretty low tech but I find I like to be able to hand author the notes with the user's in mind.
You've pretty much got the gist of it, the way we're going after the problem is producing internal notes rolled up from the code changes, and the main thing that changes as we move that communication from internal to external is frequency and delivery method.
We're still fairly early, but I think it's a mistake to think the contents of that communication should change. As soon as you get past the developers who wrote the code the primary thing people care about is customer benefit and how this work contextualizes into our goals, so we start there.
Internal comms comes out in internal channels (eg, Slack), and gets updated frequently (> 1x a week). As we bring that message out to customers we offer more self serve options (eg, hosted URL, embedded widget), and then only recommend pushing a notification once a month in the form of a recap.
(But you should still have a place that potential customers can see all the work your team is doing)
Would love to talk more on it, and thanks for brining this question up, very cool to see all the responses.
I would love every professional software to have a full release note for each update, as well as options to not have your web app update until you are ready for the changes to your flow.
The first sentence should inform of any breaking changes or major (e.g. security) fixes. For instance:
We are pleased to announce the release of version 2.2.1, which includes several fixes for major issues, includes one change which will require action on your part, 36 bug fixes, 12 enhancements, and the new combine harvester feature.
And then you drill into the top level. Also, categorise, it helps people find stuff.
If you just need a simple thing, query what you are releasing (from jira or whatever tracking system you are using) and package them up into categories of features/bug fixes and keep the release notes general.
if it's important to have an accurate curated set of release notes, create a field in your bug db for external release notes---leave those for tehncical product managers, support etc to edit as they want. THen you can have internal and external ones.
you always need to review for language and sensitive data, so human review (but again LLMs are helpful now fo this stuff too)
Each user story has separate fields with summary information, testing notes, and technical information for developers. The release process pulls the information from the linked user stories into an Excel spreadsheet, and the non-technical users just ignore that column.
One of the things that stood out was the need for the docs teams to have visibility and early notice about what is going into a release. Some teams mentioned using slack emojis for markers to help review what is proposed for promotion for eg. another was the buy-in to treat these various docs as strict release requirements (will you be willing to block a release because the docs aren’t ready?)
Lots of LLM-driven tooling attempts, but the Ghostty one is the only one I remember reading publicly.
What I've seen work well for a UI-centric service: Let a UX person craft the release notes and a product person edit/tweak them.
* consisting of PRs merged since last release. (This is better than manually updating CHANGELOG. Do not allow direct commits to main w/o a PR.)
* internal audience is engineering, PMs, customer success & support, these do not leave the company.
* PRs can be examined for more context if needed. This is a good enough balance between noise & automation.
External release notes:
* manually updated log of important changes, such as new features or other larger changes. These do not include all bug fixes.
* visible to customers.
* do not mention version numbers, only dates. You do not want to leak how often you release, or customers will start demanding release notes per version or dictating your release schedule.
When you fix bugs for customers, tell them what day/time the bug fix went live.
Obtaining set of PRs merged since last release is non-trivial, but doable.
Unless you’re developing open source or developer APIs/SDKs, end users don’t care about release notes. The KB needs to get updated, and meaningful feature improvements get announced in newsletters or blog posts every N weeks. A good customer experience team will take care of this based on raw release notes, and also notify customers who reported bugs when those bugs are fixed.
Engineers working on the product get raw release notes.
Engineers integrating the product ideally get something edited a bit to be maximally useful when working out how to upgrade — Django’s release notes are something to aspire to.
No. The release notes are for the end user.
There should be a separate changelog for technical users. This documents changes to the software including things that are invisible to users. For example, adding some unit tests wouldn't be in the release notes but it would be in the changelog.
Stakeholder comms is an entirely separate, but equally as important, thing. That should include information about impact the release is expected to have, what dependencies it impacts, and who gets the credit for work in the release.
As the author of an R package, my release notes are much drier and businesslike. The package is quite static, so releases are mainly bug fixes. I start each item with either 'Add' or 'Change', then I name the function, and then I supply a short descriptive phrase and end with a link to the github issue where where users can see why the change was made, and what the code differences were.
I realize that this is not an answer to the question, really, because all users of the R package are basically on an even footing, in terms of knowing the R language and the science that the package is intended to support. If there is something transferrable to the OP's use-case, I guess it is to be systematic and terse, and to use a fairly fixed way of writing (being aware that not all users have English as the first language).
I think crafting good release notes (that go beyond automated release notes generated based on conventional commits, which are mostly good for bug hunting), is still a mainly manual and tedious process, though after the first few decent hand-crafted ones, the first draft can be handed off to an LLM these days.
- The git commit log is the raw material. We try to have clean commits, but it's as messy as it is.
- This gets compiled into CHANGELOG.md at release time; we include all functionality and bugfixes, basically anything that any user or non-team dev might be interested or care about. But if some feature required multiple commits, we only include one line item for it. And if a feature gets reverted, we don't include both the feature and the reversion (that would be very confusing). This is for posterity.
- From the CHANGELOG we gather the "important subset" for the github release notes; this includes all features and major bugfixes, but only major API additions or changes. This has "see the CHANGELOG[link] for the full list of changes" at the bottom. This is for developers and users who follow us on github and are therefore more dev-savvy.
- From these release notes we produce the website release notes. This includes a complete list of new features, options, and commands, and important bugfixes (ones that a user might have experienced and would compel them to upgrade). But not any API changes unless it was a topline item for this release. This is for users and links back to the CHANGELOG.
- From the release webpage we pull highlights for social media, which link back to the release webpage.
We can always target different groups on social media with different subsets of functionality, but linking back up through the funnel leaves a trail of breadcrumbs for people who are more interested in the details.
Not affiliated with Hacker News or Y Combinator. We simply enrich the public API with analytics.