Skip to the content ForoHackers
Log inJoin

git_blame_myself

Joined 5 February 2024

Posts in the forum

  1. Replied in 100vh on mobile browsers is lying to you (y cómo arreglarlo)

    Shipped a login screen once where the submit button was below the fold on every phone. Tested only in the desktop browser's responsive mode. The commit message said "mobile support". The commit messa…

  2. Replied in Lab writeup: the box where the 'vulnerability' was a forgotten .bak file

    Confession time: I once committed a `.env` file to a public repository, noticed twenty minutes later, force-pushed it away, and felt clever. It was already in the clone someone had made, and in the f…

  3. Replied in Poner la licencia en cada fichero: SPDX en dos minutos

    Add it to a pre-commit check so missing headers fail the build. I once found an entire directory with no header three years after it was merged. Nobody could say whose it was, which is exactly the pr…

  4. Replied in I rewrote shared history last week. A gentle guide to --force-with-lease

    Doing the drill tonight. Also adding the lease to my shell alias so that plain `--force` is no longer one keystroke away.

  5. Started I rewrote shared history last week. A gentle guide to --force-with-lease

    Confession time. Last week I rebased a branch that two colleagues had already pulled, pushed with `--force`, and deleted a commit one of them had added an hour earlier. Nobody died. The commit was re…

  6. Replied in Refactorizar el bucle de juego de 900 líneas con tests de caracterización

    Commit after every green step. When I did something similar I once extracted three functions in one go, the golden test failed, and I had no idea which of the three was the culprit. Tiny commits make…

  7. Replied in The skeleton of a good postmortem, and the section everybody skips

    "What was lucky" is the section I steal for everything. In my last postmortem the luck was that the broken config was pushed on a Friday morning, not at 17:55. I put that in writing and the team fina…

  8. Replied in How I lost three days fixing a first layer

    The "config in git, one commit per reason" part is gold. I once changed three settings at once on a machine, it got worse, and I had no idea which one to blame. Spoiler: it was me, as always.

  9. Replied in RAID0 for speed: a short post-mortem, with arithmetic

    For the record, I lost a week of work once by running `git clean -fdx` in the wrong directory. We have all been the cautionary tale at least once. The difference is writing it up.

  10. Replied in A benchmark score is a measurement. Where are the error bars?

    And the other source of optimism: if you tune hyperparameters on the test set, the "test" number stops being a test. I did that. See my leakage post-mortem. Keep a held-out set that nothing touches u…

  11. Started I reported 97% accuracy and it was data leakage. A post-mortem.

    Sharing a mistake because writing it down is the only thing that stops me from making it again. I trained a classifier on a table of events and got 97% on the held-out set. I told the team. A week l…

  12. Replied in Config files, build pipelines, and my ongoing quarrel with indentation

    I once shipped a version string `1.10` that a YAML parser read as the float `1.1`. The release notes said 1.1 for a week. I only noticed because a user asked where 1.10 was.

  13. Replied in Cardinality: the metric label that ate the monitoring server

    We did it with a `pod_name` label on a short-lived job. Every run created new pods, so the series count grew forever, and old series lingered in memory. The fix was dropping the label at ingest. I al…

  14. Started My CI cache key was the branch name. Confession and fix.

    Sharing a mistake so it is one fewer for you. Our pipeline cached dependencies to speed up builds. I set the cache key to the **branch name**, thinking "each branch gets its own cache". Looked cleve…

  15. Replied in I wanted to renew one certificate. Now I run my own CA.

    Same shape as my worst weekend: I tried to bump one dependency and ended up rewriting the build. My rule now is a timer. If the "quick fix" has no result in 30 minutes, I stash everything and apply o…

  16. Replied in Alert on burn rate, not on every blip

    We had the opposite: 40 alerts for one dead database, because every service that talked to it screamed separately. Alerting on symptoms users can see (error ratio at the edge) instead of causes (CPU,…

  17. Replied in AddressSanitizer: what it catches that your tests do not

    I added an ASan job to our CI after reading something like this. First run: three bugs in code that had been "stable" for years. Humbling and very worth it.

  18. Replied in Threat-modelling your dependencies: what pinning does and does not buy you

    Mine, after getting burned: automated update PRs are welcome, but they never auto-merge. Somebody reads the changelog and skims the diff of anything that touches install scripts or network code. It c…

  19. Replied in How I force-pushed over a coworker's work (and the command that would have saved me)

    That is a very good point and I did not know it. Adding it to my notes, along with the apology I owe my colleague.

  20. Started How I force-pushed over a coworker's work (and the command that would have saved me)

    Confession time, because my mistakes are the only lessons I reliably learn from. Last month I rebased my feature branch and needed to push it. Git refused, as it should. So I did what everyone does:…