timothy/ersatztv
1
0
Fork 0
Code Issues 54 Pull Requests 6 Actions Packages Projects Releases 9 Wiki Activity

The QSV frame-pool 422 can name a control the editor is not rendering #848

New Issue
Open
opened 2026-08-26 23:36:54 +02:00 by timothy · 0 comments
timothy commented 2026-08-26 23:36:54 +02:00
Owner
Copy Link
Copy Source

Split out of #735 (PR #847) as an accepted low-severity residual, with the review's own repro.

Repro

  1. Take a profile written before #529's floor existed that stores qsvExtraHardwareFrames = 0 (or any value below 64).
  2. Set its Hardware acceleration to something other than QSV — or copy it, since a copy is POSTed to the create handler where the legacy exemption deliberately does not apply.
  3. Save.

Observed: Save is disabled with QSV extra hardware frames must be at least 64, and FFmpegProfilesScreen.tsx renders that field only when hardwareAcceleration === 'Qsv' — so the message names a control that is not on screen. The operator has no hint that the workaround is "switch acceleration to QSV, fix the value, switch back".

Why the check is unconditional (do NOT re-gate it)

This is not a regression to undo by restoring the hardwareAcceleration === 'Qsv' gate in validate(). The server checks the field unconditionally, so with the gate the SPA let a draft through that the POST then 422'd anyway — the same dead end, one round trip later and further from the control. api.ffmpeg-profile-numeric-bounds records the rule: a conditionally-rendered field still needs its unconditional check. Re-gating would reopen that.

The reachability is unchanged by #735 — the create handler rejected this copy before too. What changed is that the rejection now happens in the form, which is why the missing control became visible.

Two candidate fixes (either, not both)

  • Render the field whenever the stored/draft value is out of range, regardless of acceleration — the control appears exactly when it is actionable, which is the smaller conceptual change.
  • Append the workaround to the message — "… — switch Hardware acceleration to QSV to change it". Cheaper, but leaves a form that names an absent control.

Prefer the first if it does not complicate the row's conditional (it currently shares a branch with the Tonemap algorithm row).

Done-when

  • Copying or re-saving a legacy sub-64 profile on a non-QSV acceleration either shows the control or names the workaround
  • The unconditional check in validate() is preserved (a test pins that it still fires for a non-QSV draft)
  • Adversarial review passed
Split out of #735 (PR #847) as an accepted low-severity residual, with the review's own repro. ## Repro 1. Take a profile written before #529's floor existed that stores `qsvExtraHardwareFrames = 0` (or any value below 64). 2. Set its **Hardware acceleration** to something other than QSV — or copy it, since a copy is POSTed to the create handler where the legacy exemption deliberately does not apply. 3. Save. **Observed:** Save is disabled with `QSV extra hardware frames must be at least 64`, and `FFmpegProfilesScreen.tsx` renders that field **only** when `hardwareAcceleration === 'Qsv'` — so the message names a control that is not on screen. The operator has no hint that the workaround is "switch acceleration to QSV, fix the value, switch back". ## Why the check is unconditional (do NOT re-gate it) This is not a regression to undo by restoring the `hardwareAcceleration === 'Qsv'` gate in `validate()`. The server checks the field unconditionally, so with the gate the SPA let a draft through that the `POST` then 422'd anyway — the same dead end, one round trip later and further from the control. `api.ffmpeg-profile-numeric-bounds` records the rule: *a conditionally-rendered field still needs its unconditional check*. Re-gating would reopen that. The reachability is unchanged by #735 — the create handler rejected this copy before too. What changed is that the rejection now happens in the form, which is why the missing control became visible. ## Two candidate fixes (either, not both) - **Render the field whenever the stored/draft value is out of range**, regardless of acceleration — the control appears exactly when it is actionable, which is the smaller conceptual change. - **Append the workaround to the message** — "… — switch Hardware acceleration to QSV to change it". Cheaper, but leaves a form that names an absent control. Prefer the first if it does not complicate the row's conditional (it currently shares a branch with the Tonemap algorithm row). ## Done-when - [ ] Copying or re-saving a legacy sub-64 profile on a non-QSV acceleration either shows the control or names the workaround - [ ] The unconditional check in `validate()` is preserved (a test pins that it still fires for a non-QSV draft) - [ ] Adversarial review passed
timothy added the bugfrontendpriority: low labels 2026-08-26 23:37:06 +02:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
ad-hoc
One-off / ad-hoc work not tracked by a dedicated issue
 api
REST API / HTTP endpoints
 bug
Something isn't working
 ci-cd
Build, test, deploy pipeline
 content
Channel content / schedules / playlists
 dependencies
Dependency updates (Renovate)
 enhancement
New feature or improvement
 frontend
ChicoryTV React SPA frontend
 in-progress
Claimed by an active session — do not pick up
 jellyfin
Jellyfin tuner / IPTV integration
 parked
Excluded from automatic queue pickup; work only when explicitly selected
 priority: high
priority: low
priority: medium
review
Adversarial review finding
 security
Security / vulnerability fix
No labels bug frontend priority: low
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
renovate timothy (Timothy)
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: timothy/ersatztv#848