xr-reference
Use when checking an XR anti-pattern, looking up a latency, frame rate, field-of-view or resolution number, checking the platform landscape (the three-platform picture and the OpenXR and WebXR position, verified August 2026), finding resources, or needing a picker and a ship checklist. Companion to the other vr-ar-development skills.
- License 许可
- MIT
- 最近更新
- 2026-11-15
- Stars
- ★ 5
- Forks
- 3
01 前言 Front Matter
- name
- xr-reference
- description
- Use when checking an XR anti-pattern, looking up a latency, frame rate, field-of-view or resolution number, checking the platform landscape (the three-platform picture and the OpenXR and WebXR position, verified August 2026), finding resources, or needing a picker and a ship checklist. Companion to the other vr-ar-development skills.
03 SKILL.md 正文
VR and AR Development: Anti-Patterns, Numbers, the Platform Landscape, and Resources
Part 5 of 5 of the VR and AR Development reference (plugin
vr-ar-development), covering §12–§17. Sibling skills:xr-perceptual-constraints-displays-and-tracking(§0–§3),xr-rendering-input-and-spatial-understanding(§4–§6),xr-platforms-audio-and-comfort(§7–§9),xr-ux-design-and-performance-budgeting(§10–§11). Section numbers are shared across the set; a reference written as §N →skillpoints into that sibling skill.Currency: The perceptual constraints are human physiology and permanent; the platform and hardware landscape moves fast. See §14 below for the current picture.
Scope. Assumes a graphics reference for the rendering fundamentals (rasterization, PBR, the rendering equation) and a computer-vision reference for SLAM and multi-view geometry. This is the XR-specific layer built on top of both.
⚠️ GOTCHA boxes mark what makes users physically ill or breaks presence.
The three facts that make XR its own discipline:
- ⚠️ You are rendering to a nervous system, not a screen. The frame budget is set by human perception, and missing it doesn't drop quality — it makes people nauseated (§1 →
xr-perceptual-constraints-displays-and-tracking). No other software has this property.- ⚠️ Latency is the master constraint and it is not a performance metric. Motion-to- photon under ~20 ms is the consensus threshold, and it governs your entire architecture (§1.1 →
xr-perceptual-constraints-displays-and-tracking).- ⚠️ Presence is fragile and asymmetric. It takes sustained correctness to build and one bad frame, one wrong-scale object, one mistracked hand to destroy. Everything in §9 →
xr-platforms-audio-and-comfortand §10 →xr-ux-design-and-performance-budgetingis about protecting it.
§12. Anti-Patterns
| Anti-pattern | Why |
|---|---|
| Testing comfort only on yourself | ⚠️ You acclimated. Fresh users are the test (§1.2 → xr-perceptual-constraints-displays-and-tracking) |
| Taking camera control from the user | ⚠️ Reliable nausea (§1.2 → xr-perceptual-constraints-displays-and-tracking, §10 → xr-ux-design-and-performance-budgeting) |
| Smooth locomotion with no comfort options | First-session sickness (§5.4 → xr-rendering-input-and-spatial-understanding) |
| Smooth turning as the default | ⚠️ Yaw rotation is the worst axis. Snap turn (§5.4 → xr-rendering-input-and-spatial-understanding) |
| Relying on ASW/reprojection to hit frame rate | ⚠️ It's a safety net, not a budget (§4.2 → xr-rendering-input-and-spatial-understanding) |
| Frame rate as the thing you sacrifice | ⚠️ Drop render scale instead (§11 → xr-ux-design-and-performance-budgeting) |
| Profiling for 20 seconds on a mobile SoC | Thermal throttling changes everything (§11 → xr-ux-design-and-performance-budgeting) |
| Optimizing to average frame time | ⚠️ A 1% spike is a visible hitch (§11 → xr-ux-design-and-performance-budgeting) |
| Rendering the scene twice, naively | Single-pass/multiview exists (§4.1 → xr-rendering-input-and-spatial-understanding) |
| TAA in XR | ⚠️ Ghosting is far worse in stereo. Prefer MSAA (§4.4 → xr-rendering-input-and-spatial-understanding) |
| Rendering at exactly display resolution | ⚠️ Distortion resampling eats centre detail (§2 → xr-perceptual-constraints-displays-and-tracking, §4.4 → xr-rendering-input-and-spatial-understanding) |
| Head-locked UI panels | Uncomfortable, breaks presence (§10 → xr-ux-design-and-performance-budgeting) |
| Desktop-sized text | ⚠️ PPD is not monitor PPI. Test on device (§10 → xr-ux-design-and-performance-budgeting) |
| Content anchored to world origin in AR | ⚠️ Origin drifts as SLAM refines. Use anchors (§6 → xr-rendering-input-and-spatial-understanding) |
| Shipping AR without occlusion or contact shadows | ⚠️ Objects float; the illusion never forms (§6 → xr-rendering-input-and-spatial-understanding) |
| Expecting real-world occlusion on optical see-through | ⚠️ Physically impossible — additive light only (§2 → xr-perceptual-constraints-displays-and-tracking) |
| Untracked audio | Destroys presence instantly (§8 → xr-platforms-audio-and-comfort) |
| Complex hand gestures as core input | ⚠️ Pinch is reliable; little else is (§5 → xr-rendering-input-and-spatial-understanding) |
| Interaction requiring extended arms | ⚠️ Gorilla arm within minutes (§5.3 → xr-rendering-input-and-spatial-understanding) |
| Depth as the only channel for critical info | Not everyone fuses stereo (§9 → xr-platforms-audio-and-comfort) |
| No boundary/guardian consideration | People hit walls and each other (§9 → xr-platforms-audio-and-comfort) |
| Assuming inside-out tracking always works | ⚠️ Mirrors, low light, moving vehicles (§3 → xr-perceptual-constraints-displays-and-tracking) |
| Single-platform native SDK for a multi-platform product | ⚠️ OpenXR exists for this (§7 → xr-platforms-audio-and-comfort) |
| Treating eye-tracking data casually | It's biometric (§5 → xr-rendering-input-and-spatial-understanding) |
§13. Numbers
LATENCY ⚠️ THE CONSTRAINT
Motion-to-photon target <20 ms ⚠️ the consensus rule
Gaze-to-photon (foveation) 42–91 ms tolerable; artifacts unchanged under ~60 ms
⚠️ a different, looser budget than head MTP
FRAME BUDGET
72 Hz → 13.9 ms · 90 Hz → 11.1 ms · 120 Hz → 8.3 ms ⚠️ for BOTH eyes
Low persistence illumination ~2 ms
OPTICS
⚠️ Human acuity ~60 PPD (the "retinal" target)
Render target typically 1.2–1.4× display resolution (distortion headroom)
IPD range ~54–72 mm typical adult
SENSORS
IMU ~1000 Hz, low latency, ⚠️ unbounded drift
Camera/SLAM 30–60 Hz, drift-free, ⚠️ high latency
→ fuse for both (§3)
UX
Comfortable content depth ~0.5–20 m · reading sweet spot ~1–3 m
Primary content within ~±30° of forward gaze
⚠️ Looking up is worse than looking down
§14. The Platform Landscape — verified August 2026
⚠️ GOTCHA — this section decays fastest in the document, and the sourcing is weaker than the rest. Much of what follows comes from XR trade press, vendor blogs, and development agencies marketing their services. ⚠️ The structural picture is corroborated across sources; specific prices, percentages and roadmap dates are not measurements. Verify before making a platform commitment. §1–§11 →
xr-perceptual-constraints-displays-and-tracking,xr-rendering-input-and-spatial-understanding,xr-platforms-audio-and-comfort,xr-ux-design-and-performance-budgetingdo not depend on any of it.
14.1 The three-platform picture
⚠️ The consistent framing across sources is that platform choice follows use case, not quality:
- Meta Quest — ⚠️ the installed base and mature catalogue; the default for scale and lowest per-seat cost. One trade study found Meta or Quest referenced in 38% of XR stories over a four-month 2026 window, more than the next two companies combined. ⚠️ Horizon OS still favours Meta's own hardware and store, though it does support OpenXR.
- Apple Vision Pro / visionOS — ⚠️ vertically integrated, premium fidelity; the pick for visualization, design review and executive demos. ⚠️ Coverage in 2026 has been substantially retreat-flavoured — a reduced headset roadmap and cancelled products are reported, alongside a 2025 refresh with an upgraded processor and improved headstrap.
- Android XR (Google + Samsung + Qualcomm) — ⚠️ the fastest-rising, and reported as the #2 platform by coverage share. Runs standard Android apps, supports Unity 6, OpenXR 1.1 and WebXR, with Samsung Galaxy XR shipping and XREAL glasses in the ecosystem. The explicit strategy is the Android playbook applied to XR.
Also: Valve's Steam Frame announced, PICO (Project Swan flagship indicated for late 2026), Varjo for high-end enterprise/simulation, HoloLens 2 in maintenance mode with enterprise hardware like HMS SiNGRAY G2 positioned for that gap. ⚠️ And the sector is volatile — one source notes four VR studios closing within a single week in 2026.
⚠️ Smart glasses are the growth story, not headsets: reported at 42% of XR coverage, with Meta's Ray-Ban line including a display model. Displays are OLED-on-silicon near-term with MicroLED in development.
14.2 ⚠️ The genuinely actionable finding — OpenXR and WebXR
This is the part I'd act on regardless of how the hardware race resolves.
- ⚠️ OpenXR is the hedge. Applications built against it run on any conformant headset unmodified; Unity and Unreal both recommend it; Android XR supports OpenXR 1.1, which means the same rendering pipeline reaches Quest. ⚠️ A multi-platform program is increasingly the realistic answer, which is exactly the argument for an OpenXR-based, engine-portable strategy now.
- WebXR adoption reportedly grew 40% in 2026, with Interop 2026 proposing WebXR as a focus area — meaning browser vendors coordinating to close compatibility gaps. ⚠️ One codebase across Quest, Vision Pro, Galaxy XR and phone browsers is a real proposition now, with the usual web trade-offs in performance and device access.
§15. Resources
| Source | Why |
|---|---|
| LaValle, Virtual Reality | ⚠️ Free online, rigorous, the best foundation — perception, geometry, tracking |
| Jerald, The VR Book: Human-Centered Design for VR | ⚠️ §1 → xr-perceptual-constraints-displays-and-tracking, §9 → xr-platforms-audio-and-comfort and §10 → xr-ux-design-and-performance-budgeting in depth |
| Schmalstieg & Höllerer, Augmented Reality: Principles and Practice | The AR reference |
| Oculus/Meta Developer "Best Practices" | ⚠️ Hard-won comfort guidance; still the practical baseline |
| Apple HIG for visionOS | ⚠️ Excellent 3D interaction design thinking, platform-independent value |
| Khronos OpenXR specification | §7 → xr-platforms-audio-and-comfort |
| Abrash's blog and talks | ⚠️ The clearest writing on why XR is perceptually hard |
| Albert et al., "Latency requirements for foveated rendering in VR" | §1.1 → xr-perceptual-constraints-displays-and-tracking's gaze budget |
| IEEE VR / ISMAR / CHI proceedings | Where the comfort research lives |
§16. Quick Reference
16.1 Picker
| Need | Use |
|---|---|
| Multi-platform target | ⚠️ OpenXR + Unity or Unreal (§7 → xr-platforms-audio-and-comfort, §14.2) |
| Widest reach, one codebase, lowest cost | WebXR (§14.2) |
| Comfortable locomotion | ⚠️ Teleport + snap turn as defaults (§5.4 → xr-rendering-input-and-spatial-understanding) |
| Precise interaction | Controllers — don't default to hands (§5 → xr-rendering-input-and-spatial-understanding) |
| Low-fatigue interaction | ⚠️ Gaze + pinch; hands near the body (§5.3 → xr-rendering-input-and-spatial-understanding) |
| Free GPU headroom | ⚠️ Fixed foveated rendering (§4.3 → xr-rendering-input-and-spatial-understanding) |
| Halve CPU draw overhead | Single-pass/multiview stereo (§4.1 → xr-rendering-input-and-spatial-understanding) |
| Antialiasing in XR | ⚠️ MSAA, not TAA (§4.4 → xr-rendering-input-and-spatial-understanding) |
| AR content that stays put | ⚠️ Anchors, never world coordinates (§6 → xr-rendering-input-and-spatial-understanding) |
| Make AR objects look real | ⚠️ Occlusion + contact shadows (§6 → xr-rendering-input-and-spatial-understanding) |
| Presence per unit effort | ⚠️ Head-tracked spatial audio (§8 → xr-platforms-audio-and-comfort) |
16.2 Ship checklist
- [ ] Native frame rate hit without relying on reprojection? (§4.2 →
xr-rendering-input-and-spatial-understanding) - [ ] Frame-time histogram clean — no 1% spikes? (§11 →
xr-ux-design-and-performance-budgeting) - [ ] Profiled on device, thermally soaked 20+ minutes? (§11 →
xr-ux-design-and-performance-budgeting) - [ ] Tested with users who have never used XR? (§1.2 →
xr-perceptual-constraints-displays-and-tracking) - [ ] Comfort options: vignette, snap turn, teleport, seated mode, height? (§9 →
xr-platforms-audio-and-comfort) - [ ] Camera never moves without user input? (§1.2 →
xr-perceptual-constraints-displays-and-tracking, §10 →xr-ux-design-and-performance-budgeting) - [ ] Text legible on device, not just in the editor? (§10 →
xr-ux-design-and-performance-budgeting) - [ ] UI body-locked or diegetic, not head-locked? (§10 →
xr-ux-design-and-performance-budgeting) - [ ] Audio head-tracked? (§8 →
xr-platforms-audio-and-comfort) - [ ] Tracking-loss and relocalization failure handled gracefully? (§3 →
xr-perceptual-constraints-displays-and-tracking, §6 →xr-rendering-input-and-spatial-understanding) - [ ] AR: anchors used, occlusion and contact shadows present? (§6 →
xr-rendering-input-and-spatial-understanding) - [ ] Accessibility: seated, one-handed, subtitles, non-stereo fallback? (§9 →
xr-platforms-audio-and-comfort) - [ ] Guardian/boundary interaction sane? (§9 →
xr-platforms-audio-and-comfort)
§17. Method
§1–§11 → xr-perceptual-constraints-displays-and-tracking, xr-rendering-input-and-spatial-understanding, xr-platforms-audio-and-comfort, xr-ux-design-and-performance-budgeting rest on perceptual research and stable engineering practice — LaValle,
Jerald, the platform best-practice documents, and the IEEE VR / ISMAR literature.
⚠️ Human physiology does not version, which is why §1 → xr-perceptual-constraints-displays-and-tracking is the most durable part of this
document and the part I'd read first.
Scoped to complement: rasterization, PBR and the rendering equation sit in a graphics
reference; SLAM, VIO and multi-view geometry in a computer-vision reference. ⚠️ §3 → xr-perceptual-constraints-displays-and-tracking and
§6 → xr-rendering-input-and-spatial-understanding deliberately point at those rather than restating them.
Two searches were run in August 2026: the platform and hardware landscape, and the perceptual thresholds (motion-to-photon, cybersickness factors, VAC, foveated rendering latency).
Confidence. High in §1 → xr-perceptual-constraints-displays-and-tracking — the 20 ms MTP consensus, the four primary cybersickness
contributors, the VAC consequence list, and the 42–91 ms foveated-rendering latency
tolerance all come from peer-reviewed sources and are consistent across them.
⚠️ The §1.1 → xr-perceptual-constraints-displays-and-tracking point that gaze latency and head latency are different budgets by roughly 3×
is worth internalizing and is frequently conflated in practitioner writing.
High in §2–§11 → xr-perceptual-constraints-displays-and-tracking, xr-rendering-input-and-spatial-understanding, xr-platforms-audio-and-comfort, xr-ux-design-and-performance-budgeting, which are established practice.
⚠️ §14 is explicitly the weak section and is flagged as such in place. The sourcing is XR trade press, vendor blogs, and development agencies who sell XR services — several of the comparison articles returned are marketing for consultancies. The structural claims are corroborated across independent sources (OpenXR as the portability layer; Android XR rising; Quest holding installed-base advantage; smart glasses growing faster than headsets). ⚠️ The specific figures — the 40% WebXR growth, the 38% coverage share, the 15–25% PolySpatial overhead, prices and roadmap dates — are single-source or vendor-adjacent, and I have attributed rather than asserted them. Verify before committing budget.
§14.2 is the part I'd act on: ⚠️ the OpenXR argument holds regardless of which hardware vendor wins, which is precisely what makes it the safe architectural bet.
04 同仓库更多 Skills
white-paper-writing
Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks "how should I structure this" for a long-form document.
prose-technical-mechanics
Use when editing or writing prose at a technical/mechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.
prose-mechanics
Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate/Jamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix "AI-sounding" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.
narrative-structure
Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or "how should I structure this story/pitch/presentation" request.
legal-brief-writing
Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic/neurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.
journal-publishability-criteria
Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs/paper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.