Hey folks Apologies for not getting the upstream bugfix release out sooner. I’ll take a look at that tomorrow.
Adam (Chris?), I think your diagnosis is correct: When I test i3 from git, I do not see any delay. I agree that the issue seems specific to Debian package i3-wm 4.25-2. In fact, when extracting the upstream tarball, we see no sleep: root@7c65d5ad4056:/usr/src# tar xf i3-wm_4.25.orig.tar.xz root@7c65d5ad4056:/usr/src# cd i3-4.25/ root@7c65d5ad4056:/usr/src/i3-4.25# grep sleep i3bar/src/workspaces.c root@7c65d5ad4056:/usr/src/i3-4.25# …but when using dpkg-source to extract and apply patches, we do see the sleep call being inserted into the source: root@7c65d5ad4056:/usr/src# dpkg-source -x *.dsc Signing key on 7BF5F6AC36431F5D40DC137A4CF2B218F54DAE3D is bad: The primary key is not live because: Expired on 2026-01-21T08:05:54Z dpkg-source: warning: cannot verify inline signature for ./i3-wm_4.25-2.dsc: no acceptable signature found dpkg-source: info: extracting i3-wm in i3-wm-4.25 dpkg-source: info: unpacking i3-wm_4.25.orig.tar.xz dpkg-source: info: unpacking i3-wm_4.25-2.debian.tar.xz dpkg-source: info: using patch list from debian/patches/series dpkg-source: info: applying fix-perl-interpreter-path dpkg-source: info: applying bts1126791 root@7c65d5ad4056:/usr/src# cd i3-wm-4.25/ root@7c65d5ad4056:/usr/src/i3-wm-4.25# grep sleep i3bar/src/workspaces.c sleep(1); root@7c65d5ad4056:/usr/src/i3-wm-4.25# cd .. That’s a pretty big footgun IMHO! Worth a bug report. Best regards Michael On Thu, Feb 5, 2026 at 5:30 AM, Adam Marchetti <[email protected]> wrote: I can confirm this behavior after updating i3-wm on sid as well. Though I can't reproduce the behavior after building from upstream: $ git clone https://github.com/i3/i3 $ mkdir -p i3/build ; cd i3/build $ meson setup $ ninja $ ./i3bar Curiously, the version of that patch from debian/patches (https://salsa.debian.org/i3-team/i3-wm/-/blob/debian/unstable/debian/patches/bts1126791) includes both the actual patch as well as an indented patch used for testing. The testing patch includes "sleep(1);" when parsing the workspaces IPC response. I wonder if this part of the patch file is *also* applied to the source, in addition to the rest of the patch file? ~ Chris -- Best regards, Michael
