A sitesources run that GitLab rate-limited dropped about twenty trees it
never read from website/src/data/source-paths.json. A tree key names a
pinned sha, so the last record's entry for an unread tree still holds and
is carried over, in sitesources and in profile_sync's refresh.
The Vulkan renderer merge (75 commits): no ROM loading change. The CMakeLists.txt organization and app name definitions moved down six lines (647-648).
183 upstream commits. Ultima's data_archive.cpp was folded into Common::load_engine_data, which opens ultima.dat and checks its version.txt the same way (common/engine_data.cpp:273-279). No dists/ or gui/ file changed, so the bundled hashes hold. Adds japanesemacfonts.dat, which the Mac GUI reads for Japanese fonts and ScummVM does not ship.
The fork rebased on trzy 1481545 (Src/Version.h:32 now 0.3a-1481545-mm); the two commits touch tilegen timings, the PowerPC core and the GUI only. upstream_commit follows to 1481545. The 11726c4 release archives carry the same Games.xml, Music.xml and crosshair bitmaps. RetroPie builds the main branch tip on x86, so the pin follows HEAD; the ROCKNIX arm build stays a note, with the version it prints.
101 upstream commits. Japanese glyphs now come from M PLUS 1p embedded with the Inter font (imgui_drawer.cc:481-498), so msgothic.ttc and the fontconfig CJK match are no longer read; XConfig gained a per-title random MAC address. Still no system file read. The libretro core is still at f23c14c.
Three upstream commits on tilegen write timings and the VF3 start-up: no cited file changed and every ref anchors at 1481545. Version.h still reads 0.3a-WIP. Libretro-Supermodel is still at 84bc106.
47 upstream commits. The ECam patch gained a Symbian 8.1a build (ecam_v81a.dll, [epoc81a] in ecam.dll.map) and per-version map sections; the mediaclientaudio and mediaclientvideo patch DLLs were rebuilt. Sizes and SHA-1 read from src/patch at d55da28. Firmware and ROM loading unchanged; the libretro branch is still at e39ccac.
A long systems list opens on None. A there unticks every system and
moves to the first one, and A on a system with nothing ticked installs
that one alone, so Enter still continues. Inside muxterm, which keeps
PgUp and PgDn for its scrollback, Left and Right on the D-pad page.
One system of 66 at 40x15 took 17 presses with a pad, 57 in muxterm;
it now takes 12 on both. The pyyaml check of the device lists skips
where pyyaml is not installed.
One run of profile_sync --all --rebase-refs --fresh. Every guard held:
no ref to review, no prose left under an older pin, no file name or hash
literal gained in a cited file since the pin, no built_by pin and no pin
keyed by build mode. The site's record of cited source paths is rebuilt
with them.
A system that runs from its SD card (Onion, spruceOS, muOS, KNULLI,
MinUI...) named with --platform on a computer no longer takes that
computer's RetroArch folder: one mounted card of that system is used,
at a terminal the card root is asked, without questions the run stops
on the --dest example. The interface opens the card screen with the
mounted cards first.
The result pager scrolls and counts lines, so its last line is
reachable at every size. The setup filter knows the devices each
setup runs on, from the guides and the firmware pages. A new event
kind, data, gives the bytes of the data trees, and the plan shows
them.
Refs that still anchor say nothing of what the code added around them:
mame2003_plus's zn.c gained two Taito G-NET BIOS sets while its cited
lines only shifted. --rebase-refs and --bump-commit now refuse a profile
whose cited files gained a file name or a hash literal since the pin, or
could not be read, until --accept-changed says the additions were read.
With this, --all --rebase-refs --fresh is safe to run as is.
citron's taken domain now resolves to an unrelated host that fails TLS,
so the pass reported an upstream error where the profile already states
the forge is gone. Any forge error now agrees with upstream_gone; a rate
limit still does not.
The console switch tests put tty2 in front and switched to the pty the test was given. When that pty was /dev/pts/2 the launcher found itself in front already, and the refused-switch case returned 0.
Each pin advance left website/src/data/source-paths.json behind and the
site's freshness test red until sitesources.py --fetch ran by hand.
profile_sync now rebuilds it after any run that wrote a profile, online
and not in a dry run.
After the second bring-to-head pass: 40 bumps, ten profiles by hand,
five off-forge apps, the G-NET BIOS sets and the new hiscore.dat.
Manifests generated twice, identical.
The Apple II page named its start files with no visible way to get one: the download sat after the collection path, fourth of a card's facts. Each card now carries a Download button in its title, each start-line name links to its card, and a card held on another part page is reached by its permanent link.
--add-app, the result screen and a [y/N] question put the launcher of
muOS, KNULLI or ROCKNIX on the card the BIOS folder is on. The files come
from install/launchers/index.json at the same ref, each checked by size
and SHA-256 before the first is written, and the app event reports it.
Branch tips, tags and releases are cached for six hours, and the
EKA2L1, trzy/Supermodel and VICE heads read as behind their real HEAD.
--fresh asks the forges again, and is refused with --offline.
A recale on a profile whose pins are keyed by build mode said the prose
had to move first. It now says the pins are set by hand.
Abdessamad Derraz