Commit

Deciding to work on the /dev/null graph was, how should I say, not the best time saving decision I made. Wanted to fix a tiny rendering bug on the dahsed lines, ended up dealing with SVG, canvas, performance issues and so on. Now I'm looking into how the SSR is done on the app, trying to find ways to improve the rendering performance. I had to create a full testing harness (https://github.com/jubianchi/dev-null.sh/tree/test/graph-labjubianchi/dev-null.sh:graph-lab/) along the way because, man, there are so many cases to test each time we touch something on this graph (Claude was here to help 🤖) This was a long streak of PRs (https://github.com/dev-null-sh/dev-null.sh/pull/724dev-null-sh/dev-null.sh#724, https://github.com/dev-null-sh/dev-null.sh/pull/783dev-null-sh/dev-null.sh#783, https://github.com/dev-null-sh/dev-null.sh/pull/956dev-null-sh/dev-null.sh#956, https://github.com/dev-null-sh/dev-null.sh/pull/980dev-null-sh/dev-null.sh#980, https://github.com/dev-null-sh/dev-null.sh/pull/1019dev-null-sh/dev-null.sh#1019) and one last to come to hopefully provide the best possible experience. And now, @kachelle is also doing his part (https://github.com/dev-null-sh/dev-null.sh/pull/1135dev-null-sh/dev-null.sh#1135 — which I have to review) 👍