# How an HTTP/2 bug led me to an emergency Ferron 3 release

**URL:** <https://discuss.ferron.sh/t/how-an-http-2-bug-led-me-to-an-emergency-ferron-3-release/14>\
**Category:** Core Development & Architecture\
**Created:** [September 4, 2026, 1:38pm UTC](https://discuss.ferron.sh/t/how-an-http-2-bug-led-me-to-an-emergency-ferron-3-release/14 "2026-09-04T13:38:02Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![dorianniemiec](https://discuss.ferron.sh/user_avatar/discuss.ferron.sh/dorianniemiec/32/14_2.png) [@dorianniemiec](https://discuss.ferron.sh/u/dorianniemiec)\
**Post date:** [September 4, 2026, 1:38pm UTC](https://discuss.ferron.sh/t/how-an-http-2-bug-led-me-to-an-emergency-ferron-3-release/14/1 "2026-09-04T13:38:02Z")

</div>

Remember when I was excited to [show Ferron 3.0.0-rc.1](https://ferron.sh/blog/ferron-3-rc)? Well, that faded away when I visited the Ferron website, and found that it (or rather `zincio-http`) had a critical bug inside HTTP/2 implementation… 😔

I was then debugging myself using Wireshark and found that `HEADERS` frame for empty responses was missing (!). So I quickly patched `zincio-http`, Ferron 3, and did a quick emergency release of Ferron 3.0.0-rc.2 with just that bugfix.

If you used Ferron 3.0.0-rc.1, I recommend **immediately updating to Ferron 3.0.0-rc.2**.

_Reposted from [Emergency Ferron 3.0.0-rc.2 release is here (post-mortem?) - Ferron Blog](https://ferron.sh/blog/ferron-3-0-0-rc-2-emergency)_