Free, open-source 3D sculpting for Linux, macOS and Windows: dense meshes, remeshing, painting, built-in AI (image/text to 3D), Send to Unreal, game-ready LODs, Gaussian splats and OpenUSD.
C++ 3d3d-sculptingcppdear-imguidigital-sculptingfbxgame-developmentgaussian-splattinggltfimage-to-3dmcpopenglopenusdsculptingtext-to-3dunreal-engine
Docs and comments: no personal names (maintainer instead)
islandmed575be5967ef8
diff --git a/docs/design/core-200m.md b/docs/design/core-200m.mdindex 7a6b8af9..563d649e 100644--- a/docs/design/core-200m.md+++ b/docs/design/core-200m.md@@ -160,7 +160,7 @@ Every number twice: this laptop and the floor profile (`SF_HW_SIMULATE=floor`),Nothing here merges into a release without the floor-tier numbers.-## 7b. The novel track (Tony, 11-03: beyond the known playbook)+## 7b. The novel track (11-03: beyond the known playbook)Sections 1-6 are the known playbook done well (multiresolution offsets, cluster LOD to a pixel budget, a cachedscreen, scale-adaptive dabs). The novel track changes *what a stroke is*, not only how fast it runs. Each item is adiff --git a/docs/design/perf-moonshots.md b/docs/design/perf-moonshots.mdindex a5974274..ef98f17b 100644--- a/docs/design/perf-moonshots.md+++ b/docs/design/perf-moonshots.md@@ -4,13 +4,13 @@ Research branch `research/perf-moonshots` (2026-11-01). Goal: the cost of a stroand the brush footprint, never with the total polygon count. This note surveys recent work, rates eight ideas,reports three prototypes measured with `--benchmark standard-33m`, and recommends what to ship.-Starting point (Tony's clean 33M run, M1 Pro, Asahi, Mesa GL 4.6, unified memory): orbit 89 fps; Clay small p99+Starting point (a clean 33M run, M1 Pro, Asahi, Mesa GL 4.6, unified memory): orbit 89 fps; Clay small p9938 ms; Clay large p99 119 ms; 240 px brush p99 338 ms; Move 18 fps. The branch is built on `perf/bigbrush-p99`(dabs of meshes from 2M vertices go to the solver thread without the preview), which already takes the strokep99s to about 30 ms; what it leaves is the surface trailing the pen (each 240 px dab at 33M still costs100-200 ms of CPU) and Move, which never used the solver path.-## 0. Hard requirement: the floor tier (Tony, 11-03)+## 0. Hard requirement: the floor tier (11-03)Weak GPUs must get great performance too, not only big cards. Every idea, prototype and default below is held tothis; an experiment that misses it stays off for that tier or doesn't ship.@@ -33,7 +33,7 @@ this; an experiment that misses it stays off for that tier or doesn't ship.octahedral normals, implicit adjacency), and micro levels (1.25 bytes per triangle) above ~8M.5. **Benchmarks.** Every number is reported twice: this laptop and the **iGPU-simulated floor profile** (below),with p99 frame time and stroke latency at 5, 20 and 50M triangles. A beta must not regress the floor tier.-6. **The floor decides; a big GPU never does (Tony, 11-03).** Performance must never be gated by an expensive GPU:+6. **The floor decides; a big GPU never does (11-03).** Performance must never be gated by an expensive GPU:that is the value proposition. Every perf claim and every beta gate is judged on the floor tier first (thesimulated profile and this laptop), and only then on the RTX 3060, as a sanity check. The 3060 PC is for testspeed and Windows driver checks, not the target. A result that meets its target only on the 3060 has not met it.@@ -396,7 +396,7 @@ here), which is the canvas track's job (§7) and CPU dab throughput (core-200m Ndata (3) belongs with the Sculpt Timeline; coarse-first (2) and learned prediction (7) don't pay once thepreview is on.-## 7. Canvas track (the cached screen; Tony, 11-03)+## 7. Canvas track (the cached screen; 11-03)Generic techniques, designed here from first principles (clean room: no decompiling, no copied code or docs). Goal:a dab costs about the brush's pixels, whatever the mesh size, so the floor tier sculpts 20M like a small mesh.diff --git a/docs/design/release-train.md b/docs/design/release-train.mdindex c474408f..dc30d0f7 100644--- a/docs/design/release-train.md+++ b/docs/design/release-train.md@@ -6,7 +6,7 @@ one-command releases (`scripts/release.sh`). v2 is about shipping more often witthey want it, risky features ship dark, and hardware we can't run in CI is checked by people before a release.Every item below adds no port, no listener and no code from the network. Where an item touches the network or-accounts it says so, and needs Tony's OK before it is built.+accounts it says so, and needs the maintainer's OK before it is built.## 1. Channels: Stable, Preview, Nightly@@ -40,8 +40,8 @@ opens the download page (no download or install from the app: that would be new**Insiders.** Insider builds (private sculptforge-insiders repo) stay separate from Preview: the public wording is"Insiders will get each release about two weeks early, starting with SculptForge 1.0." Before 1.0, Preview is the-public release candidate. Open question for Tony: from 1.0 on, is Preview for Insiders only (then the channel shows-only to signed-in members, which needs an account check: not before Tony decides), or public with Insiders getting+public release candidate. Open question: from 1.0 on, is Preview for Insiders only (then the channel shows+only to signed-in members, which needs an account check: not before the maintainer decides), or public with Insiders gettingfeatures two weeks earlier through feature flags (section 2)?## 2. Feature flags@@ -56,7 +56,7 @@ Risky or unfinished features merge early and ship dark instead of waiting on lon- A flag is deleted within two releases of going on by default; `ctest flags_have_owners` fails on a flag olderthan that (a date in the table).- No remote flags: switching features from the feed would be remote configuration of every install (attack surface,- and surprising). If ever wanted, Tony decides.+ and surprising). If ever wanted, the maintainer decides.## 3. Hardware matrix gate@@ -90,7 +90,7 @@ For "the pen does something weird" reports that we can't reproduce.## 5. Screenshot-diff approval-Tony's polish bar: every visible change is looked at before merge.+The polish bar: every visible change is looked at before merge.- The UI tests already save screenshots (`shot` in tests/ui/*.sfui, uploaded as `ui-shots-*` artifacts). Main'sLinux x86_64 shots (Mesa llvmpipe, deterministic enough) become the baseline artifact for each main commit.diff --git a/src/link/PhoneLink.h b/src/link/PhoneLink.hindex 7ed4369c..bb1d5dab 100644--- a/src/link/PhoneLink.h+++ b/src/link/PhoneLink.h@@ -57,7 +57,7 @@ struct PhoneLinkOptions {std::string inboxRoot; // the Captures folder (created on first use)uint16_t port = kPhoneLinkPort;bool localNetworkOnly = true;- std::string desktopName; // shown on the phone ("Tony's PC")+ std::string desktopName; // shown on the phone ("Studio PC")std::string version; // SculptForge version, shown on the phonestd::string siteUrl; // face page; empty = kPhoneLinkSiteUrl (or SCULPTFORGE_LINK_URL)std::string relayUrl; // empty = kPhoneLinkRelayUrl (or SCULPTFORGE_LINK_RELAY); "off" = none;diff --git a/tests/test_memory_guard.cpp b/tests/test_memory_guard.cppindex 177eeb25..9fe03a9a 100644--- a/tests/test_memory_guard.cpp+++ b/tests/test_memory_guard.cpp@@ -29,7 +29,7 @@ TEST(memory_check_margin) {}TEST(memory_benchmark_scene_skipped) {- setAvailableMemoryOverride(size_t(1.3 * double(kGB))); // Tony's machine: 1.3 GB free+ setAvailableMemoryOverride(size_t(1.3 * double(kGB))); // a real report: 1.3 GB freeCHECK(availableMemoryBytes() == size_t(1.3 * double(kGB)));const MemoryCheck c33 = checkMemory(sceneNeed(33));CHECK(!c33.enough);