[ccoventry]

Release note: none

Part of #704.

A debug cvar, ds_debug_log_gun_lowered, logs every time the first-person gun goes down or comes back up, with the demo player's clock:

• POV: from client.dll's g_ihidexhair, found by a pattern that is unique and the same on both installs;
• HLTV in-eye: from ds_spec_match_pov's copy (spectator_gun), with anim_fix's reason (sprinting, prone transition, crawling, in the air, on a ladder).

Played over the same stretch of a POV demo and the HLTV demo of that half, the two lists line up. The findings are on #704: the two agree 94-96% of the time, with four kinds of difference. The jump one turned out to be the log's, not the game's (see the limit below), so no decision is needed there.

It's a measuring tool; the fixes it points at come in their own PRs (#767 for the mid-sprint bounce).

Tests: cargo test -p goldsrc-hooks passes, and clippy 1.98.1 (host and i686) is clean. Commands.txt and console_commands_data.js are updated.

After merging dev (#719), the name is ds_debug_log_gun_lowered, as the #555 table lists it. It has a row in docs/dodstudio_commands.md now and is listed in console_prefix's NEW_SINCE_555. All checks pass: hook tests (687), workspace tests, clippy (both targets), fmt and npm run test:unit. I rebuilt the test DLL into local\test-704b\. game_probe on PRE and POST played wsod25_grp2_h1_hltv from 1499 s, in-eye on player 13. Both logged gun_log: down at 1540.949 (spec 13: sprinting)-style lines and never logged could not find g_ihidexhair. Reports: local\game-probe\707-ds-pre\, 707-ds-post\.

Merged dev again (ready for review). The only conflicts were console_prefix.rs's NEW_SINCE_555 (kept both sides' names) and console_commands_data.js (regenerated with console_names.py, not hand-merged); lib.rs, commands.rs, Commands.txt and docs/dodstudio_commands.md merged cleanly. Checks: hook tests (736), native tests (769), clippy 1.98.1 on host and i686, fmt and npm run test:unit (779) all pass. The new test DLL is in local\test-707\.

The log's limit: in a POV demo it reads client.dll's own lowering sum (g_ihidexhair), not the gun that is drawn, so it misses POV's lowering on a jump or a crouch-jump. That is what made the first finding ("POV never lowers for a jump") wrong; see the correction on #704. Treat a POV-vs-HLTV difference it reports as a lead to check against the drawn frames. The cvar's docs row, its Commands.txt line and gun_log.rs's module doc now say so.

🤖 Generated with Claude Code

[ccoventry]

Merged dev in (after #719) at 16085259.

• The cvar is now ds_debug_log_gun_lowered, the name the Rename the in-game commands: area as the second word (spec_, hud_, world_...), and a shorter prefix #555 table gives it. The code's suffix already matched; I updated the comments and the PR title.
• Conflicts: lib.rs keeps dev's module list, with gun_log added in order. Commands.txt keeps dev's side, with the entry re-added under DEBUG. console_commands_data.js is regenerated.
• docs/dodstudio_commands.md had no row for this cvar, so I added one.
• It never shipped, so it gets no old_names.rs entry. It is listed in console_prefix's NEW_SINCE_555.
• Local checks: fmt, clippy 1.98.1 (workspace and i686), hook tests (687), workspace tests, npm run test:unit: all pass.
• Test DLL rebuilt: local\test-704b\dodstudio_goldsrc_hooks.dll.
• game_probe re-run on PRE and POST: wsod25_grp2_h1_hltv from 1499 s, in-eye on player 13, logging on for 45 s. Both logged lines like gun_log: down at 1540.949 (spec 13: sprinting) / up at .... Neither logged could not find g_ihidexhair. Both PASS (local\game-probe\707-ds-pre\, 707-ds-post\).
• If While a demo loads, the demo player bar shows how far along it is (#465) #758 or R&D: a restyled health readout drawn by the hook (part of #849) #882 merges first, this PR's NEW_SINCE_555 line will conflict with theirs. Keep both names in the list.

🤖 Generated with Claude Code