You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Most unused parameters are in the DataLayer subclass headers. Fixing them with [[maybe_unused]] or unnamed parameters is good first-contribution work.
Order:move stb_image_write.h to third_party/ (#192), FFPoint (#191), then decide where the generated colormap.h goes, then the layer headers, then Command.cpp / FireDomain.cpp. Enable FOREFIRE_WARNINGS_AS_ERRORS for each file as it is cleared. Finish this before #164.
Drafted with Claude Opus 5, reviewed by a maintainer.
I compile with clang, I can test with GCC. Except for the -Wreorder, the other flags should be idempotent related to compilers.
From what we saw recently with the Meso-NH team and pyfortool (Fortran, but same aproach), some false flags were raised but the rest was pretty much old code that needed an update.
Opened #238 to remove the unused listenMode and mark the MPI headers as SYSTEM (no logic change). After it, 21 unique warnings left (GCC 13.3, -Wall -Wextra, NetCDF + MPI), all in physics code:
flux/ models (13) and IsotropicIsospeed (1): unused valueOf/et/at
FarsitePropagationModel (5): variables computed and never used
CLibForeFire.cpp (2): atmoSurf computed but never used in the coupling, maybe a missing division
Leaving these for Filippi, since silencing them blindly could hide a real bug.
#255 marks 18 of the last 21 as [[maybe_unused]]. Left 3 on purpose because they look like real bugs: Farsite.adjustementWind is read but never applied, and atmoSurf is computed but unused in injectedAtGround/spotAtGround.
Activity
filippi commentedon Oct 21, 2026
Why not @HugoFara hope it'll pass on exotic compilers, bu can be custom changed anyway
HugoFara commentedon Oct 22, 2026
I compile with clang, I can test with GCC. Except for the
-Wreorder, the other flags should be idempotent related to compilers.From what we saw recently with the Meso-NH team and pyfortool (Fortran, but same aproach), some false flags were raised but the rest was pretty much old code that needed an update.
perf: inline hot-path accessors, make FFPoint and FFVector trivially …
Dev to master -> v2.6.0 (#224)
revert: keep atmoSurf and the Farsite rm for Filippi to decide (#161)
antonio-leblanc commentedon Oct 28, 2026
Opened #227 for the reorder and dead variables (286 → 276 unique warnings, GCC 13.3
-Wall -Wextra). Suggestion for the rest:unused-parameterin layer headers (comment out the name)unused-parameterin core.cppflux/,propagation/,atmoSurf, Farsite leftoversThe last one touches physics, so it waits for Filippi. The first two are mechanical, Hugo or I can take them.
refactor: clear unused-variable and reorder warnings (#161) (#227)
refactor: clear unused-parameter warnings outside the models (#161) (#…
antonio-leblanc commentedon Oct 29, 2026
Opened #238 to remove the unused
listenModeand mark the MPI headers asSYSTEM(no logic change). After it, 21 unique warnings left (GCC 13.3,-Wall -Wextra, NetCDF + MPI), all in physics code:flux/models (13) andIsotropicIsospeed(1): unusedvalueOf/et/atFarsitePropagationModel(5): variables computed and never usedCLibForeFire.cpp(2):atmoSurfcomputed but never used in the coupling, maybe a missing divisionLeaving these for Filippi, since silencing them blindly could hide a real bug.
rm unused listenMode and include MPI headers as system (#161) (#238)
plot: use the WARNING: prefix for the non-square ASCII cells warning (f…
plot: warn when ASCII cells are not square (#161) (#249)
antonio-leblanc commentedon Nov 3, 2026
#255 marks 18 of the last 21 as
[[maybe_unused]]. Left 3 on purpose because they look like real bugs:Farsite.adjustementWindis read but never applied, andatmoSurfis computed but unused ininjectedAtGround/spotAtGround.refactor: mark unused physics variables [[maybe_unused]] (#161)
Merge pull request #255 from antonio-leblanc/refactor/silence-physics…
Clear the last warnings from #161 (#267)
antonio-leblanc commentedon Nov 10, 2026
Done: #255 and #267 cleared the backlog. 0 warnings with GCC 13.3
-Wall -Wextra, with and without MPI.