- patches/flare/000{1..5}-*.patch: typing indicators, formatted
messages, edited messages, multi-select with delete-for-me, and
in-channel message search. Mirror the matching commits in
~/projects/forks/flare and apply cleanly on top of upstream 0.20.4
(which is what nixpkgs ships).
- home/profiles/gui.nix: include a flare-signal override that appends
the patches via overrideAttrs. None of them touch Cargo.lock so the
cargoDeps hash stays valid; signal-desktop stays alongside it.
agenix activation runs from initrd-nixos-activation-start, which fires
right after /sysroot/persistent is mounted but before impermanence's
stage-2 bind mounts. The TPM identity at /var/lib/agenix/tpm-identity
was therefore unreadable at activation time, and every secret silently
failed to decrypt: 'no readable identities found'. Visible downstream
fallout was pull-update-apply hitting HTTP 401 against the binary cache
because nix-cache-netrc was never written to /run/agenix.
Mark /var/lib/agenix as neededForBoot via a bare fileSystems entry,
mirroring the existing /home/${username} bind. Drop the now-redundant
environment.persistence directory entry to avoid two competing units.
`bun2nix.hook` (used by upstream omp's package.nix) sets
patchPhase = bunPatchPhase
at the end of its setup-hook unless `dontUseBunPatch` is already set.
`bunPatchPhase` only runs `patchShebangs` plus a HOME mktemp; it never
iterates over `$patches`. The standard nixpkgs `patches` attribute
therefore went into the derivation env but was silently ignored at
build time, leaving the deployed omp binary unpatched.
Switch to applying the two patches via `prePatch` (which `bunPatchPhase`
does call). Verified with strings(1) over the rebuilt binary that both
patch hunks land:
/wrong_api_format|...|invalid tool parameters/ (patch 0001)
stubsReasoningContent ... thinkingFormat == "openrouter" (patch 0002)
DeepSeek V4 Pro (and similar reasoning models reached via OpenRouter) reject
multi-turn requests in thinking mode with:
400 The `reasoning_content` in the thinking mode must be passed back
to the API.
omp's existing kimi placeholder injection (`requiresReasoningContentForToolCalls`)
covered this requirement only for `thinkingFormat == "openai"`. OpenRouter
sets `thinkingFormat == "openrouter"`, so the gate never fired even though
the underlying providers behind OpenRouter (DeepSeek, Kimi, etc.) all enforce
the same invariant.
This patch:
1. Extends `requiresReasoningContentForToolCalls` detection: any
reasoning-capable model fronted by OpenRouter now sets the flag.
2. Extends the placeholder gate in `convertMessages` to accept
`thinkingFormat == "openrouter"` alongside `"openai"`.
Cross-provider continuations are the dominant trigger: a conversation
warmed up by Anthropic Claude (whose reasoning is redacted/encrypted on
the wire) followed by a switch to DeepSeek V4 Pro via OpenRouter. omp
cannot synthesize plaintext `reasoning_content` from Anthropic's
encrypted blocks, so the placeholder satisfies DeepSeek's validator
without fabricating a reasoning trace. Real captured reasoning, when
present, short-circuits the placeholder via `hasReasoningField` and
survives intact.
Side benefit: also closes a latent gap where Kimi-via-OpenRouter
(`thinkingFormat == "openrouter"`) had the compat flag set but the
placeholder gate silently rejected it.
Applies cleanly on top of patch 0001.
Steam interprets exit 0 from 'steamos-update check' as 'update applied
successfully' and shows a persistent 'update available' notification.
The SteamOS convention is exit 7 = no update available.
