Skip to content

docs(hermes): mark the RADE/DAX null-deref rows done, per §18.3 - #5383

Open
on8st wants to merge 1 commit into
aethersdr:mainfrom
on8st:docs/roadmap-reconcile
Open

docs(hermes): mark the RADE/DAX null-deref rows done, per §18.3#5383
on8st wants to merge 1 commit into
aethersdr:mainfrom
on8st:docs/roadmap-reconcile

Conversation

@on8st

@on8st on8st commented Sep 2, 2026

Copy link
Copy Markdown

Three roadmap rows and the suggested-order list still carry the RADE/DAX-bridge panStream() null-deref as open — item 10, item 24, gap 18, and §18.7's "do this first". §18.3 of the same document already records both halves as closed, and even carries its own correction of an earlier claim that both RADE and the DAX bridge crashed.

This only makes the tables agree with the section. No code change.

I checked the tree rather than trusting §18.3

It would have been circular to cite §18.3 as evidence for editing rows that point at §18.3, so:

  • activateRADE() guards panStream() at its top and returns before anything mutates station state — it declines with a log line and a message box, and un-sticks the RADE toggle. The bare connect(..., &PanadapterStream::daxAudioReady, ...) further down is inside that guarded region on every path, and the inner lambda carries its own null check.
  • startDax() bails at entry with a debug line, before the bridge is constructed.
  • Every other panStream() call site in the tree is guarded too — AutomationServer, TciServer, MainWindow_Wiring.

So both halves of the item really are closed, independently of what §18.3 says about them.

On the convention

Struck rows use the ~~12a~~ … **DONE** form already used by items 8, 12a, 16 and 25, and set effort to to match.

Scope

Deliberately narrow. There are other roadmap entries I noticed that disagree with each other, but I have only read those — I have not verified them against the code, and a contradiction you have only read is not a finding worth sending you. They are left alone.

🤖 Generated with Claude Code

https://claude.ai/code/session_01CkisdWQiVxcUG6r6X9W3Fg

Three roadmap rows and the suggested-order list still carry the
RADE/DAX-bridge panStream() null-deref as open and "do this first".
Section 18.3 of the same document already records both halves as closed
-- RADE fixed on feat/rx-audio-bus, the DAX bridge never affected
because startDax() has always guarded panStream() at entry, and 18.3
carries its own correction of the earlier claim that both crashed.

Verified against the tree before editing, not taken from 18.3: the
guard is at the top of activateRADE(), which returns before anything
mutates station state, and the bare connect() to daxAudioReady sits
inside the guarded region. startDax() bails at entry with a debug line.
Every other panStream() call site in the tree is guarded too.

Documentation only, no code change.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CkisdWQiVxcUG6r6X9W3Fg
@on8st
on8st requested a review from a team as a code owner September 2, 2026 09:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant