Repository navigation
flymsg 0.7.1 — replay stays visible
The widest bloom mip wrote NaN on this GPU. The composite stretched it over the empty part of the frame, and adding the glow back turned those pixels black, so a paused pIP10 replay went black while the page still counted its surfaces.
The blur inlines its weights and unrolls the taps, and a non-finite sample adds nothing. One bad mip cannot erase the picture. Clip planes track the volume. Geometry is unchanged.
- The README states the N1 result released in 0.7.0: the traced graph passes 29/29, the shuffle passes 5/29, and every specific check fails.
w_synstays 0.192. - Figures from this viewer: LC4_R at 120 ms, the pIP10 pathway at 150 ms, and pIP10, dPR1 and dMS9 in the neuropil shells.
LICENSEis the plain MIT text. What it does not cover (connectome data, vendored three.js and lil-gui) is inNOTICE.
Full notes: CHANGELOG.