Prepare public beta: in-page confirmations, feedback and safe GPU errors

Keep existing Space access settings. Add recipe-import recovery, capability gates, GitHub issue drafts and bounded CUDA reference instructions; preserve simulator binaries.

BETA_TESTING.md

@@ -1,94 +1,61 @@
-# Small-group beta entry gate
-
-Current status: local UI/CPU testing is useful now. The refreshed runner interface
-requires hosted CPU and GPU qualification before inviting a broad audience. Do not describe it as a
-stable release or invite a broad audience to an unqualified build.
-
-## Protected beta address
-
-The Space must remain Protected. Share https://tiga-rce-ald-tao-gpu.hf.space/
-with testers. Do not share https://huggingface.co/spaces/TiGa-RCE/ald-tao-gpu
-as the public app entry: the repository page is restricted and can display 404
-to non-collaborators. HF documents the `.hf.space` address as the public embed
-URL for Protected Spaces. Source access is not required for anonymous CPU use.
-
-CPU processing and complete browser usability are separate acceptance checks.
-The interface permits GPU on the direct entry when the server reports it available;
-HF still enforces quota. Direct GPU may use the limited anonymous allowance.
-Optional HF sign-in identifies the visitor through an official PKCE flow using
-openid/profile only. Tokens stay in page memory; no personal-token field or app
-password is introduced. Verify HF identity — no GPU checks the same queue/client
-credential path before a run. Hosted OAuth/account-quota acceptance must still
-be qualified for each build; a completed anonymous run is not account attribution.
-
-Before remote testers receive the expanded interface:
-
-1. Freeze an allow-listed clean package and verify its manifest.
-2. Deploy it to the existing Protected test Space without changing visibility.
-3. Verify hosted CPU and one bounded GPU upload → run → download → deletion path
-   using disposable, user-owned/generated files. Check HF allowance before GPU.
-4. Confirm recipe persistence/import/export and batch completion/stop in both
-   Chrome and the in-app browser. Safari/Firefox/Edge and mobile widths then form
-   the human browser-zoo matrix, rather than a prerequisite for first local feedback.
-5. Freeze the beta build ID in each report and keep the previous clean build for
-   rollback. No invitation or repository permission change is part of this file.
-
-## Tester workflow
-
-- Use CPU first; it does not acquire a GPU lease. GPU is explicit and has no
-  automatic CPU fallback. A failed/quota-limited GPU call is not a successful run.
-- When HF rejects a staged GPU run for exhausted allowance, confirm the page
-  gives CPU/account guidance without a personal-token prompt, reports that the
-  ticket was deleted or is no longer pending, and allows another visitor to
-  prepare a request immediately.
-- Check that a loaded recipe changes the echoed effective settings and generation
-  results, not just the form. Reload and load again to check browser persistence.
-- Export/import a recipe backup. Imported data must not contain tracks, networks,
-  tokens or arbitrary fields. Invalid JSON must leave the existing library intact.
-- Run a two-step carry-forward batch and a reset-to-original batch. Export the
-  receipt; carry-forward input hash should match the preceding output hash.
-- Stop a longer batch. The current step may finish; later steps must not start.
-  A stop/error must remain visibly different from a complete batch.
-- Create/edit/export/re-import a small network. Inspect or edit a known weight
-  and verify the exported value. Modified artifacts must not inherit old scores.
-- Exercise default deletion, opt-in repeat download, manual delete, browser-only re-save,
-  result expiry, invalid files, missing files and connection failure.
-
-Current supported worker contract: Formula steering/acceleration, speed/raycast
-sensors, 2–4 layers with at most 16 nodes per layer, and bounded execution.
-Inspection can show larger networks, but that is not permission to train them.
-The track map previews selected tile layouts and terrain, including custom uploads, but does not show collision geometry. Star challenges are explicitly
-not evaluated. No driving improvement or real-game/leaderboard parity follows
-from a successful artifact lifecycle test. A bucket and broad network support
-remain separate work.
-
-## Reporting through GitHub
-
-Prepare read access plus issue filing, without write/merge/deployment privileges.
-The local issue-form draft is under `.github/ISSUE_TEMPLATE/runner_beta.yml`.
-It is not live until reviewed and committed/pushed to the intended repository.
-Repository invitations and collaborator permissions require named recipients;
-no invitations have been sent by this work.
-
-Reports should include browser/version, OS, build ID, CPU/GPU, reproduction steps,
-expected/actual behavior and a redacted screenshot or receipt. Do not attach a
-private trained network, a proprietary game-extracted track, tokens, job access
-keys, or full unrelated logs to an ordinary bug report. Use public/generated
-fixtures or describe the network shape and sensor/output contract.
-
-## Local 2026-10-13 candidate (not deployed)
-
-Content-addressed build IDs now appear in capabilities and exported run/batch receipts;
-reports include upstream simulator revision and packaged binary SHA-256 values.
-A single run has its own receipt export. `/feedback` supplies a private-feedback
-template; no public issue destination or invitation is activated.
-
-The native train-scratch adapter requires positive mutation even when continuing an
-uploaded network. The form and server now reject zero before native dispatch.
-Any GPU dispatch error attempts to discard its pending ticket; a missing ticket
-is not evidence that an already dispatched run stopped.
-
-This candidate requires an explicitly approved deployment followed by hosted
-CPU/GPU and real-account checks before tester invitations. Local checks and
-training experiments against the previous live build are recorded outside this
-package in the sibling sprint receipt. They do not qualify this candidate as deployed.
+# Public launch preflight
+
+This document defines the hosted qualification checks for the release package.
+Deployment status and build-specific acceptance belong in the release receipt;
+results from one build do not qualify another build.
+
+## Intended launch topology
+
+- The GPU Space app is publicly reachable at
+  https://tiga-rce-ald-tao-gpu.hf.space/.
+- The GPU Space source repository remains restricted.
+- The standalone CPU Space remains private. The public GPU app also offers an
+  anonymous CPU route, which does not acquire a GPU lease.
+
+App access and source-repository access are separate. Do not send users to the
+restricted source repository as the app entry point.
+
+## Required qualification for each deployed build
+
+1. Freeze the allow-listed package and verify its file manifest after source
+   and documentation changes.
+2. Deploy the package in the intended topology. Record the exact Space commit,
+   visibility, runtime stage, and build ID in the release receipt.
+3. On the deployed build, verify CPU upload → run → download → receipt
+   acknowledgement → server-side 404 using disposable generated files.
+4. Check included HF allowance before one bounded GPU fixture run. Verify the
+   CUDA report, candidate hash, download, and deletion; do not use paid overage.
+   A quota rejection must show CPU/account guidance and report whether the
+   pending ticket was discarded or already consumed.
+5. Verify optional HF sign-in and account attribution separately from an
+   anonymous GPU run. Do not ask users to paste personal access tokens.
+6. Exercise result expiry, opt-in repeat download, manual deletion, malformed
+   inputs, native failure, and temporary-file cleanup on the deployed build.
+7. Confirm recipe and batch behavior in supported browsers. A successful
+   artifact lifecycle check does not imply training quality or game parity.
+
+## Supported runner contract
+
+The worker accepts Formula steering/acceleration networks with two outputs,
+two to four layers, at most 16 nodes per layer, speed and raycast sensors, and
+bounded execution settings. Unsupported network shapes or sensors are not
+accepted for training. The track preview shows tile layout and terrain, not
+collision geometry. Star challenges are not evaluated.
+
+## CUDA reference check
+
+Use the deterministic generated network and bundled Replica B01 Oval track
+described in `CUDA_REFERENCE.md`. Keep input hashes, candidate output hash,
+settings, build ID, binary hashes, and deletion receipts together. The
+single fixture cannot establish general CUDA/CPU equivalence, original-game
+parity, or training quality.
+
+## Feedback
+
+`/feedback` serves a template only; it does not submit a report. The configured
+public destination is the [driving_sim_rs issue tracker](https://github.com/seeker-cyber-maker/driving_sim_rs/issues/new).
+Reports should identify the app URL and build ID, browser and version, operating
+system, CPU/GPU choice, reproduction steps, expected and observed behavior, and
+a redacted receipt or screenshot. Do not attach tokens, job access keys, private
+networks, proprietary tracks, or unrelated logs. Inspect receipts before
+sharing; batch receipts also include recipe and batch names.

CUDA_REFERENCE.md

@@ -0,0 +1,67 @@
+# Bounded CUDA reference check
+
+This protocol checks that the hosted CUDA path can accept a known small input,
+complete one bounded Formula training run, return a validated artifact, and
+delete the server copy after verified download. It is not a test of broad game
+parity, training quality, or general CPU/GPU equivalence.
+
+## Public-package fixture
+
+- Track: `builtin_tracks/tracks/Replica B01 Oval.track`, catalog ID
+  `787f73f5-3b8c-5d95-b0e3-0016ed076970`, SHA-256
+  `0df5755150c145a35edc6c8693f429e5b9ff4b12210ba67cbfcb932c58faf158`.
+- Network: deterministic, generated Formula network, shape `[6,4,2]`; one
+  speed sensor and five raycasts at `-60,-30,0,30,60` degrees with length 800;
+  outputs `[0,1]`; for zero-based indices, first-layer weight `W[i][j]` is
+  `(i+1)*(j+1)/100`, second-layer `W[i][j]` is `(i+1)*(j-0.5)/10`, and every
+  bias is zero. Serialize the JSON with sorted keys and compact separators.
+- Generated network bytes: 554; SHA-256
+  `d9851cb1eee40dee78fb8c4492caeac8f130b5543f6d4551bb2175c13c579c3b`.
+
+The following standard-library Python snippet is the complete network fixture
+generator. It prints the exact bytes and digest; `Name` is intentionally absent
+because the worker does not require it.
+
+```python
+import hashlib
+import json
+
+def generated_network():
+    shape = [6, 4, 2]
+    first = [[(i + 1) * (j + 1) / 100 for j in range(4)] for i in range(6)]
+    second = [[(i + 1) * (j - 0.5) / 10 for j in range(2)] for i in range(4)]
+    value = {
+        "Type": 0,
+        "Outputs": [0, 1],
+        "Sensors": [
+            {"$type": "speed"},
+            *[
+                {"$type": "raycast", "Degrees": angle, "Length": 800}
+                for angle in (-60, -30, 0, 30, 60)
+            ],
+        ],
+        "NeuralNetwork": {
+            "Shape": shape,
+            "Weights": [first, second],
+            "Biases": [[0.0] * 4, [0.0, 0.0]],
+        },
+    }
+    return json.dumps(value, sort_keys=True, separators=(",", ":")).encode()
+
+network = generated_network()
+print(len(network), hashlib.sha256(network).hexdigest())
+```
+
+Pair these network bytes with the unchanged `Replica B01 Oval.track` file from
+the package; do not rewrite the compressed track file.
+
+- Bounded run settings: population 1, generation 1, 120 ticks, mutation 0.01,
+  seed 42, one worker, default total-score reward. These are a smoke fixture,
+  not a quality recommendation.
+
+For a live CUDA qualification, first verify the target build ID and included HF
+allowance. Run the fixture once through CPU and once through GPU on the same
+build when CUDA quota permits. Record both reports, output SHA-256 values, binary
+hashes, candidate validation, and deletion receipts. A mismatch is a finding to
+investigate; do not widen the claim from this one track/network/settings tuple.
+Do not spend paid overage. No live GPU job was run to prepare this protocol.

DATA_POLICY.md

@@ -1,6 +1,6 @@
-# Data handling & deletion policy — private integration test
+# Data handling & deletion policy — public-release candidate
 
-Operator: TiGa (seeker-cyber-maker). Version: HF runner refresh 2026-10-12.
+Operator: TiGa (seeker-cyber-maker). Candidate version: 2026-10-13. This package is not a receipt that a hosted deployment or publication has occurred.
 
 Files are processed only for the requested job, not collected for our own
 training, analytics or public examples. Inputs remain in the browser until Run.

FEEDBACK.md

@@ -1,8 +1,9 @@
-# Supervised beta feedback draft
+# Feedback report template
 
 This form is a local template. It does not submit a message or upload files.
-Share a completed copy with the person who invited you, using the agreed private channel.
-A public issue destination has not been activated for this candidate.
+Use the public issue form at https://github.com/seeker-cyber-maker/driving_sim_rs/issues/new
+to report a problem. Review every attachment before posting. The issue tracker is
+for product reports; do not include confidential material.
 
 - App URL:
 - Build ID (from an exported run or batch receipt):

FRONTEND_SOURCE.json

@@ -1,10 +1,10 @@
 {
   "source": "Runner r14 single-engine browser UI; selected .track tile renderer added from simulator atlas",
   "source_sha256": "65d1e5f2c4e19b4951b828e82510005ab99a5270cf3dcb9f4eaa0fa1af8c40f8",
-  "output_sha256": "315058a08c31ec512108473904f559087080edb684c02f4d594cbe5032c77791",
+  "output_sha256": "aeb602565bf2da9d40cb6c3aae0b6e713b89e79a81557fed3e7f3dd8d16f195c",
   "scope": "Frontend only; no engine, private data, track assets, API calls or storage copied",
-  "status": "Local beta sprint candidate; not deployed",
+  "status": "Public beta release candidate; local verification pending hosted qualification",
   "track_atlas_sha256": "d97c8900e3e5fb17734b5bd61d1f4f4b196d9be30af9848e2e273ae7e4079ae4",
-  "client_sha256": "cc18e5ce4405ccc7cecc3ab6cc6d7b2abddcc6cfac69f47825c546b8d7c1bcf1",
+  "client_sha256": "b6c2e0439bc1e4cf55c5e751e497e842e1e911b229f7480d7b1e04dea246536f",
   "track_map_sha256": "ab5374aa49c7ccec90a787ee8c56cd71beb56b903bf2bbc322783e6f43c009f0"
 }

README.md

@@ -8,8 +8,10 @@ hf_oauth: true
 hf_oauth_expiration_minutes: 60
 ---
 
-Protected CPU/ZeroGPU training beta. The native Formula runner uses the matched NVIDIA CUDA simulator/library pair from Zorro909/driving_sim_rs at commit d8ab74c25df70b6a86ed101080e33d5fb785ea6f. The newer b90b8af source was checked, but its published library targets AMD HIP and its added five-wheel snowmobile support is outside this Formula-only beta. Open the public app at https://tiga-rce-ald-tao-gpu.hf.space/; the source repository remains restricted. CPU runs anonymously. GPU is explicitly selected and subject to available HF allowance; optional HF sign-in uses the official OAuth flow and needs no personal token entry.
+Public GPU app package. Its intended topology is a publicly reachable app at https://tiga-rce-ald-tao-gpu.hf.space/ with the GPU Space source repository restricted; the standalone CPU Space remains private. Deployment status and build-specific qualification are recorded in the release receipt. The native Formula runner uses the matched NVIDIA CUDA simulator/library pair from Zorro909/driving_sim_rs at commit d8ab74c25df70b6a86ed101080e33d5fb785ea6f. The newer b90b8af source was checked, but its published library targets AMD HIP and its added five-wheel snowmobile support is outside this Formula-only package. CPU runs anonymously. GPU is explicitly selected and subject to available HF allowance; optional HF sign-in uses the official OAuth flow and needs no personal token entry.
 
 The selected track map previews tile layout and asphalt/dirt/ice directly from the selected file. Twenty-eight reconstructed editor-format tracks are included. Settings-only recipes and batches save in the visitor's browser. Formula steering/acceleration networks within the current worker limits are supported.
 
 Application-controlled inputs and intermediate files are temporary. Outputs are held in memory for at most five minutes and deleted sooner after verified download by default; a single-run visitor may explicitly keep the result until that five-minute deadline. This is not a provider-wide zero-data-retention guarantee. Read DATA_POLICY.md and TRACK_LIBRARY.md. Upstream code and license: https://github.com/Zorro909/driving_sim_rs. Open upstream PRs 19 and 21 are not included.
+
+Feedback issues: https://github.com/seeker-cyber-maker/driving_sim_rs/issues/new.

RELEASE_MANIFEST.json

@@ -2,19 +2,20 @@
   "backend": "cpu+zerogpu",
   "files": {
     ".gitattributes": "c2ad16718fc51c888d32efc88c8f6e23ebe5fcf6f48e2954f896ba9770f24d72",
-    "BETA_TESTING.md": "9d6447c876a2b2b1d3dbb7637858c79a4c669d8582d8cc2a6a335ea8cc4297c1",
+    "BETA_TESTING.md": "c2d4bc2eff0f87144a862282f20585868c95533103221f5d42726242b2eeba21",
     "BINARY_PROVENANCE.json": "1116e1757afc6c8629cb2d957e3d77380fe4f7e3186c3832d16a5a08abce6aa8",
-    "DATA_POLICY.md": "0daf1b835bf829c0974e6b566cb5a605c6f52dba4517cfd54bed044e8eaf4264",
-    "FEEDBACK.md": "17127b693eb52cccf1707be24b9bced82b77c0eec174b1f696cc9a3feb33dd67",
-    "FRONTEND_SOURCE.json": "d9bf9402693f514956ec99432b6da2ccf73f17c7c26dd400123b6b9fb27bf31e",
+    "CUDA_REFERENCE.md": "56a07521fc3d6acd11e3b3c110293d1a5ff710d1b7d7b93115e7b98d5ee7c256",
+    "DATA_POLICY.md": "38da17f8528ede1ef1850e98832f75ba35ebdc86a0cde87e27674eb0ff9c7a7f",
+    "FEEDBACK.md": "f95b8f6afbf4a360976c0edca5d52f5b3a0b668c9a944d4947b840a7c53a122a",
+    "FRONTEND_SOURCE.json": "4a6b79b38053a0db37eb36b6e7209a14ad7f9f505c373003b89f950fd0ff4654",
     "GODOT_LICENSE": "fb2b3bb82c60f1e76841ba2122eb902f8e39c151476fe21ce10c9a05d1045304",
     "IMPORTER_BUILD_RECEIPT.json": "4ba5c533cc617105478e4d85c83f456a7dfa68b1388049624619871ba9e42e9d",
     "LICENSE": "53d5851b6440feeae36f72a5f57fcab37858c48b7084765ece4b0c54b9a28ed8",
-    "README.md": "caf4a2d7ac5ab949f9d9b9f5adb75c598c5669b23a040751041fc0709dc9b79f",
+    "README.md": "b41c614f81339e6392a018ba7133211a42e0416eb03544a6d54ac34efc7daa81",
     "STORAGE_BOUNDARY.md": "9a1ffb5854230690dba3f0895f9a8e570716de47be2c553de13a6892dd740015",
-    "TRACK_LIBRARY.md": "7e4829c5e0933bfb317c21d7441ad0ba2ee00ab7f9f1635549452d504aa0777b",
+    "TRACK_LIBRARY.md": "66b3656ed84f137d5b6116a0075731cd467818f555cca0969084aff67875319f",
     "UPSTREAM.json": "40a2f31fa9abc11e47e247c56909b93b587d1f5be524ab158b24f1f9eed93765",
-    "app.py": "271c13c152e916c704c9c311a4a2800ef98e3787e6557be9eab8c851de0e1fb4",
+    "app.py": "162486abd6ac7c9281c8e75c15e33373e0fddba2baf6a799dfd54d333fd575d2",
     "app_release.py": "7482559c667a665fd3cb6aa8af6c78422a822b515e14589db8306f7fd4860a8b",
     "assets/formula_template.json": "80f745df82bb7be45af52785fd582460bcddc63c5a4e1d8375d85b3122db4c6b",
     "assets/track-atlas.json": "d97c8900e3e5fb17734b5bd61d1f4f4b196d9be30af9848e2e273ae7e4079ae4",
@@ -52,12 +53,12 @@
     "builtin_tracks/tracks/Replica R08 Demolition Derby.track": "6f11a9ffbcb477c9d19964566b1b612d2a94675605246e93e21f27bbbdb357ef",
     "builtin_tracks/tracks/Replica R09 Eau Noir.track": "3ec0a78478cca1a4f0d2d98ac81f40783928adf271b70f0c32d295a6da2cdaa9",
     "builtin_tracks/tracks/Replica The Champions Trial.track": "a3bfb76422affc774b44c739b0636d08acb901fa7615b9adcb52c069f79834e1",
-    "client.js": "cc18e5ce4405ccc7cecc3ab6cc6d7b2abddcc6cfac69f47825c546b8d7c1bcf1",
+    "client.js": "b6c2e0439bc1e4cf55c5e751e497e842e1e911b229f7480d7b1e04dea246536f",
     "delivery.py": "07453999c3831a16e8b1fcc2b4ae5fb28beb7d3952a3fae8de65d16eb6f11e5b",
     "gpu_tickets.py": "b81aadb5ed42cec9212d44da7e7bbb7b413c5bd517383cac0b49ec9b4d659c92",
-    "hf-auth.js": "b59691ce2c517f193a5d5141538ebe61c961bb163cefa1fcf09b414c388984c1",
+    "hf-auth.js": "d5cf17b961d1f2efa9d4648d33ba86a148eff0d542c74a121710c4df35c92936",
     "hf_identity.py": "6993d0238ada835606dbdc3275c17a3fe5a5cac5ad2156ab78230acf2d051eba",
-    "index.html": "315058a08c31ec512108473904f559087080edb684c02f4d594cbe5032c77791",
+    "index.html": "aeb602565bf2da9d40cb6c3aae0b6e713b89e79a81557fed3e7f3dd8d16f195c",
     "requirements.txt": "9b532747e51bd052e82090ece860d6dafe027407f4512dc74272c148ea104c22",
     "server.py": "de98db424b96a4fbab7ce5b98ee42a94e846e2b3f317593e2346c5f87489ede8",
     "track-library.js": "351ec4c026b49420ce963f4310cb000d9e987fd28e6dac5ecd0176509340f6e6",

TRACK_LIBRARY.md

@@ -23,7 +23,8 @@ input. Upload a custom track remains available. Selecting a built-in clears the
 custom file selection; choosing an upload switches back to upload mode. Built-in
 selection survives completed runs in this page, but is not written to cookies,
 localStorage, recipes, or account storage. Network selection is still explicit.
-The renderer remains a placeholder.
+The canvas preview shows tile layout and asphalt, dirt, and ice terrain. It does
+not show collision geometry or prove that the replica matches native physics.
 
 The public app exposes only catalog-listed IDs and hash-checked replica bytes.
 No directory browsing, caller-selected paths, remote track URLs or extraction

app.py

@@ -17,6 +17,7 @@ from fastapi.responses import FileResponse,JSONResponse
 from app_release import ROOT,cpu_worker,verify_package
 from gpu_tickets import InputTickets
 from server import create_app
+from delivery import CapacityExceeded
 from worker import run_native,track_input,network_input
 from training_config import validate_config
 from build_info import build_identity
@@ -90,8 +91,16 @@ def build_app(cuda_call):
         if not admission.acquire(blocking=False):raise ValueError('GPU worker busy')
         try:
             track,network,automatic,*config=tickets.take(ticket)
-            payload,report=cuda_call(track,network,*config)
-            return {**base.state.results.publish(payload,automatic),'backend':'gpu','report':report}
+            try:
+                payload,report=cuda_call(track,network,*config)
+                return {**base.state.results.publish(payload,automatic),'backend':'gpu','report':report}
+            except CapacityExceeded:
+                raise ValueError('Result capacity reached; no result was published.') from None
+            except Exception:
+                # Gradio logs callback exceptions. Keep native/provider details out
+                # of both logs and the browser-facing error while preserving the
+                # UI's separate pending-ticket cleanup check.
+                raise ValueError('GPU processing failed; no result was published.') from None
         finally:admission.release()
 
     @app.api(name='visitor_identity',api_visibility='undocumented')

client.js

@@ -5,6 +5,7 @@ let busy = false;
 let localResult = null;
 let capabilitiesReady = false;
 let gpuTransport = null;
+let gpuAvailable = false;
 let gradioClient = null;
 let stagedNetwork = null;
 let batchActive = false;
@@ -29,7 +30,7 @@ function buttons() {
   window.HFAuth?.setBusy(busy || !!job);
   window.setTrainingBusy?.(busy || !!job);
   window.setEditorBusy?.(busy || !!job);
-  $('run').disabled = busy || !!job || !capabilitiesReady || !networkSource();
+  $('run').disabled = busy || !!job || !capabilitiesReady || !networkSource() || ($('backend').value==='gpu'&&!gpuAvailable);
   $('backend').disabled = busy || !!job || !capabilitiesReady;
   $('download').disabled = busy || !job;
   $('erase').disabled = busy || !job;
@@ -38,11 +39,20 @@ function buttons() {
   $('forgetLocal').disabled = busy || !localResult;
   $('track').disabled=busy || !!job;
   $('network').disabled=busy || !!job;
-  if($('batch_start'))$('batch_start').disabled=busy || !!job || !capabilitiesReady;
+  if($('batch_start'))$('batch_start').disabled=busy || !!job || !capabilitiesReady || ($('backend').value==='gpu'&&!gpuAvailable);
   if($('batch_stop'))$('batch_stop').disabled=!batchActive || batchStopRequested;
   if($('run_export'))$('run_export').disabled=busy || !runReceipt;
   if($('batch_export'))$('batch_export').disabled=busy || !batchReceipt;
 }
+$('backend').addEventListener('change',buttons);
+const feedbackLink=$('feedbackLink');
+feedbackLink?.addEventListener('click',()=>{
+  const url=new URL('https://github.com/seeker-cyber-maker/driving_sim_rs/issues/new');
+  const browser=navigator.userAgent||'Not reported';
+  const body=['Build: '+(appBuild?.build_id??'Not reported'),'Browser and OS: '+browser,'Selected backend: '+$('backend').value.toUpperCase(),'','What happened?'].join('\n');
+  url.searchParams.set('title','ALD beta problem report');url.searchParams.set('body',body);
+  feedbackLink.href=url.href;
+});
 function saveLocal() {
   if(!localResult)return;
   const link=document.createElement('a'),url=URL.createObjectURL(localResult);
@@ -197,7 +207,7 @@ $('download').onclick = async () => {
   finally {busy=false; buttons();}
 };
 $('erase').onclick=async()=>{
-  if(!confirm('Delete remaining server-side result and revoke download access? Local files are unaffected.'))return;
+  if(!await window.appConfirm('Delete remaining server-side result and revoke download access? Local files are unaffected.',{heading:'Delete server result',confirmLabel:'Delete result',danger:true}))return;
   busy=true; buttons();
   try {await request('/api/jobs/'+job.id,{method:'DELETE',headers:headers()}); job=null; say('Server result deleted. Local files are unaffected.');}
   catch(error){say(error.message);}
@@ -286,8 +296,10 @@ async function connect(){
       $('buildLabel').textContent='Build: '+(appBuild?.build_id?.slice(0,25)??'Unavailable — include the app URL in feedback');
       $('buildLabel').title=appBuild?.build_id??'';
     }
+    gpuAvailable=capabilities.gpu===true;
     gpuTransport=capabilities.gpu_transport??null;
-    $('gpuOption').disabled=!capabilities.gpu;
+    $('gpuOption').disabled=!gpuAvailable;
+    if(!gpuAvailable)$('backend').value='cpu';
     $('backendNote').textContent=capabilities.gpu?(hostedStandalone?'CPU is anonymous. GPU may use a limited anonymous allowance or your optional HF account allowance.':'CPU is anonymous. GPU requires an available HF allowance.'):'CPU training is ready; no sign-in or GPU allowance needed.';
     $('gpuAccount').hidden=!capabilities.gpu;
     $('computeBlock').hidden=!capabilities.gpu;
@@ -295,7 +307,7 @@ async function connect(){
     say('Connected. Choose your track and network, then run the test.');
   } catch(error) {
     $('connectionRecovery').hidden=false;
-    capabilitiesReady=false; buttons();
+    capabilitiesReady=false;gpuAvailable=false;gpuTransport=null;$('backend').value='cpu';$('gpuOption').disabled=true;buttons();
     $('backendNote').textContent='Connection unavailable — training is disabled.';
     say('Could not reach this app. Retry the public beta app address: https://tiga-rce-ald-tao-gpu.hf.space/. The HF repository page is restricted in Protected mode. No run was started.');
   } finally {clearTimeout(timeout);$('retry').disabled=false;}

hf-auth.js

@@ -40,7 +40,7 @@
   node('hfSignIn').onclick=async()=>{
     if(locked||exchanging||!ready)return;
     if((node('track').files.length||node('network').files.length) &&
-       !confirm('HF sign-in navigates away. Selected files and unsaved editor changes are lost; saved recipes remain. Continue, then reselect files?'))return;
+       !(await window.appConfirm('HF sign-in navigates away. Selected files and unsaved editor changes are lost; saved recipes remain. Continue, then reselect files?',{heading:'Continue to HF sign-in',confirmLabel:'Continue to sign in'})))return;
     exchanging=true;render();
     try{
       const state={};

index.html

@@ -1,8 +1,34 @@
 <!doctype html><html lang="en"><meta charset="utf-8"><meta name="viewport" content="width=device-width, initial-scale=1"><title>AI Learns To Drive</title>
 <style>
 :root{color-scheme:dark;font:16px system-ui;background:#0c1420;color:#e5edf5}body{max-width:1400px;margin:auto;padding:22px}h1{font-size:32px;margin:0}h2{font-size:21px;margin:0 0 14px}small,.hint{color:#a2b4c8;line-height:1.6}.grid{display:grid;grid-template-columns:minmax(360px,.9fr) minmax(0,1.25fr);gap:24px}.card{background:#152233;border:1px solid #2a3d52;border-radius:16px;padding:22px;margin-top:20px}label{display:block;margin:12px 0 5px}input,select,button{font:inherit;background:#0c1725;color:#e5edf5;border:1px solid #49627c;border-radius:7px;padding:9px}input[type=number],select{width:100%;box-sizing:border-box}input[type=file]{max-width:100%;box-sizing:border-box}button{cursor:pointer}button.primary{background:#62dfb1;color:#102522;font-weight:700}button:disabled{opacity:.4;cursor:default}.row{display:grid;grid-template-columns:1fr 1fr;gap:15px}.reward{display:grid;grid-template-columns:1.7fr 1fr 70px 30px;gap:5px;margin:8px 0}.reward input{width:70px;box-sizing:border-box}a{color:#7ce4c3}canvas{width:100%;background:#09111b;border-radius:10px}#status{white-space:pre-wrap;line-height:1.7}table{width:100%;text-align:left;border-collapse:collapse;font-size:14px}td,th{padding:9px;border-bottom:1px solid #2a3d52}.pill{display:inline-block;background:#284539;border-radius:20px;padding:5px 12px;color:#91efc6;margin:6px 0}#error{color:#ffb5a2;white-space:pre-wrap}.toolbar{display:flex;gap:12px;align-items:center;flex-wrap:wrap;padding-bottom:16px;border-bottom:1px solid #2a3d52;margin-bottom:15px}.run-card{position:sticky;top:20px}details{margin-top:18px;border-top:1px solid #2a3d52;padding-top:14px}summary{cursor:pointer;font-weight:600;color:#c6d9e9}details[open]>summary{margin-bottom:12px}[hidden]{display:none!important}#map_note{font-size:14px}#runs{max-height:380px;overflow:auto}.table-scroll{overflow:auto}@media(max-width:850px){.grid{grid-template-columns:minmax(0,1fr)}.card{min-width:0}.run-card{position:static}body{padding:16px}}
-.tabs{display:flex;flex-wrap:wrap;gap:10px;margin:20px 0}.tabs button[aria-selected=true]{background:#62dfb1;color:#102522}.library{display:grid;grid-template-columns:repeat(auto-fit,minmax(260px,1fr));gap:20px}.library input[type=text]{width:100%;box-sizing:border-box}.queue-row{display:flex;gap:8px;margin:8px 0}.queue-row select{flex:1}.pane{min-width:0}.map{min-height:300px;background:#09111b;border-radius:12px;padding:10px}.map canvas{display:block;width:100%;height:auto;max-height:320px;object-fit:contain}.map-legend{display:flex;gap:18px;flex-wrap:wrap;margin:10px 0;color:#c6d9e9;font-size:14px}pre{white-space:pre-wrap;overflow-wrap:anywhere}input[type=text]{width:100%;box-sizing:border-box}#libraries .row label>input,#libraries .row label>select{display:block;margin-top:6px}#editor_parameters input{max-width:240px} .eyebrow{color:#62dfb1;font-size:12px;letter-spacing:.18em;font-weight:800}.intro{font-size:19px;color:#b9cddd;max-width:820px;margin:10px 0 0}header{padding:12px 0 4px}header h1{font-size:clamp(30px,4vw,45px);letter-spacing:-.03em;margin-top:7px}.about{max-width:850px;font-size:14px;margin-top:15px}.about p{line-height:1.65}.card{box-shadow:0 10px 35px #050a1280}.tabs{position:sticky;top:0;z-index:3;background:#0c1420eF;padding:12px 0;border-bottom:1px solid #2a3d52}.tabs button{border-radius:12px;padding:11px 15px}.map-status{margin:7px 0;color:#dbe9f1;font-size:14px}#run{background:#62dfb1;color:#102522;font-weight:800;width:100%;margin-top:12px;padding:13px}.privacy-note{font-size:13px;line-height:1.5;color:#a2b4c8}.keep-result{display:flex;gap:10px;align-items:flex-start;padding:12px;border:1px solid #3b566b;border-radius:10px;background:#102030;line-height:1.4}.keep-result input{margin:3px 0 0;flex:none} .terrain{display:inline-block;width:11px;height:11px;border-radius:3px;margin-right:4px;background:#607e96}.terrain.dirt{background:#b89561}.terrain.ice{background:#b9d9ed}.terrain.start{background:#66e5aa}@media(max-width:850px){.map{min-height:0}}</style>
-<header><div class="eyebrow">AI LEARNS TO DRIVE · TRAINING LAB</div><h1>Train, inspect, repeat.</h1><p class="intro">Pick a track and network, run a bounded training session, and keep your recipes in this browser.</p><details class="about"><summary>About this beta and its data handling</summary><p>Simulator by <a href="https://github.com/Zorro909/driving_sim_rs">Zorro909</a>; interface and hosting by <a href="https://github.com/seeker-cyber-maker/driving_sim_rs">TiGa</a>. Inspired by the <a href="https://store.steampowered.com/app/3312030/AI_Learns_To_Drive/">game by ActivationFunction</a>. Built-in tracks are reconstructed replicas. Star challenges are not evaluated here.</p><p>Uploaded files are used for the requested run. Results are held briefly in application memory and deleted after verified download or expiry. Browser recipes contain settings only. <a href="/policy" target="_blank" rel="noopener">Read the data policy</a>.</p></details><p><a href="/feedback" target="_blank" rel="noopener">Beta feedback template</a> · Export a run receipt when reporting a problem.</p><p id="buildLabel" class="hint">Build: connecting…</p></header>
+.tabs{display:flex;flex-wrap:wrap;gap:10px;margin:20px 0}.tabs button[aria-selected=true]{background:#62dfb1;color:#102522}.library{display:grid;grid-template-columns:repeat(auto-fit,minmax(260px,1fr));gap:20px}.library input[type=text]{width:100%;box-sizing:border-box}.queue-row{display:flex;gap:8px;margin:8px 0}.queue-row select{flex:1}.pane{min-width:0}.map{min-height:300px;background:#09111b;border-radius:12px;padding:10px}.map canvas{display:block;width:100%;height:auto;max-height:320px;object-fit:contain}.map-legend{display:flex;gap:18px;flex-wrap:wrap;margin:10px 0;color:#c6d9e9;font-size:14px}pre{white-space:pre-wrap;overflow-wrap:anywhere}input[type=text]{width:100%;box-sizing:border-box}#libraries .row label>input,#libraries .row label>select{display:block;margin-top:6px}#editor_parameters input{max-width:240px} .eyebrow{color:#62dfb1;font-size:12px;letter-spacing:.18em;font-weight:800}.intro{font-size:19px;color:#b9cddd;max-width:820px;margin:10px 0 0}header{padding:12px 0 4px}header h1{font-size:clamp(30px,4vw,45px);letter-spacing:-.03em;margin-top:7px}.about{max-width:850px;font-size:14px;margin-top:15px}.about p{line-height:1.65}.card{box-shadow:0 10px 35px #050a1280}.tabs{position:sticky;top:0;z-index:3;background:#0c1420eF;padding:12px 0;border-bottom:1px solid #2a3d52}.tabs button{border-radius:12px;padding:11px 15px}.map-status{margin:7px 0;color:#dbe9f1;font-size:14px}#run{background:#62dfb1;color:#102522;font-weight:800;width:100%;margin-top:12px;padding:13px}.privacy-note{font-size:13px;line-height:1.5;color:#a2b4c8}.keep-result{display:flex;gap:10px;align-items:flex-start;padding:12px;border:1px solid #3b566b;border-radius:10px;background:#102030;line-height:1.4}.keep-result input{margin:3px 0 0;flex:none} .terrain{display:inline-block;width:11px;height:11px;border-radius:3px;margin-right:4px;background:#607e96}.terrain.dirt{background:#b89561}.terrain.ice{background:#b9d9ed}.terrain.start{background:#66e5aa}@media(max-width:850px){.map{min-height:0}}dialog#appConfirm{box-sizing:border-box;width:min(92vw,440px);border:1px solid #54718d;border-radius:16px;padding:24px;background:#152233;color:#e5edf5;box-shadow:0 24px 80px #000b}dialog#appConfirm::backdrop{background:#02060bcc}#appConfirm h2{margin:0 0 10px}#appConfirm p{line-height:1.55;color:#c4d3e1;white-space:pre-wrap}#appConfirm .actions{display:flex;justify-content:flex-end;gap:10px;flex-wrap:wrap;margin-top:22px}#appConfirm button{min-width:100px}#appConfirm button.danger{background:#8e3b37;color:#fff;border-color:#d7837e}</style>
+<dialog id="appConfirm" aria-labelledby="appConfirmTitle" aria-describedby="appConfirmMessage">
+  <h2 id="appConfirmTitle">Confirm action</h2>
+  <p id="appConfirmMessage"></p>
+  <div class="actions"><button id="appConfirmCancel" type="button">Cancel</button><button id="appConfirmAccept" type="button">Continue</button></div>
+</dialog>
+<script>
+(() => {
+  const dialog=document.getElementById('appConfirm'),title=document.getElementById('appConfirmTitle');
+  const message=document.getElementById('appConfirmMessage'),cancel=document.getElementById('appConfirmCancel');
+  const accept=document.getElementById('appConfirmAccept');
+  let finish=null,returnFocus=null;
+  dialog.addEventListener('close',()=>{
+    const resolve=finish;finish=null;
+    if(resolve)resolve(dialog.returnValue==='confirm');
+    if(returnFocus?.isConnected)returnFocus.focus();returnFocus=null;
+  });
+  cancel.addEventListener('click',()=>dialog.close('cancel'));
+  accept.addEventListener('click',()=>dialog.close('confirm'));
+  window.appConfirm=(text,{heading='Confirm action',confirmLabel='Continue',cancelLabel='Cancel',danger=false}={})=>{
+    if(finish) return Promise.resolve(false);
+    returnFocus=document.activeElement;title.textContent=heading;message.textContent=text;
+    accept.textContent=confirmLabel;cancel.textContent=cancelLabel;accept.classList.toggle('danger',danger);
+    return new Promise(resolve=>{finish=resolve;dialog.showModal();cancel.focus();});
+  };
+})();
+</script>
+<header><div class="eyebrow">AI LEARNS TO DRIVE · TRAINING LAB</div><h1>Train, inspect, repeat.</h1><p class="intro">Pick a track and network, run a bounded training session, and keep your recipes in this browser.</p><details class="about"><summary>About this beta and its data handling</summary><p>Simulator by <a href="https://github.com/Zorro909/driving_sim_rs">Zorro909</a>; interface and hosting by <a href="https://github.com/seeker-cyber-maker/driving_sim_rs">TiGa</a>. Inspired by the <a href="https://store.steampowered.com/app/3312030/AI_Learns_To_Drive/">game by ActivationFunction</a>. Built-in tracks are reconstructed replicas. Star challenges are not evaluated here.</p><p>Uploaded files are used for the requested run. Results are held briefly in application memory and deleted after verified download or expiry. Browser recipes contain settings only. <a href="/policy" target="_blank" rel="noopener">Read the data policy</a>.</p></details><p><a id="feedbackLink" href="https://github.com/seeker-cyber-maker/driving_sim_rs/issues/new" target="_blank" rel="noopener">Report a problem on GitHub</a> · Opens an editable draft and never submits it. Review details before posting; export a run receipt separately if it helps explain the issue.</p><p id="buildLabel" class="hint">Build: connecting…</p></header>
 <nav class="tabs" aria-label="Lab sections"><button data-tab="drive" aria-selected="true">1 · Drive & view</button><button data-tab="training" aria-selected="false">2 · Training runs & presets</button><button data-tab="batch" aria-selected="false">3 · Batches</button><button data-tab="editor" aria-selected="false">4 · Network editor</button></nav><div id="drive" class="pane"><div class="grid"><section class="card"><h2>Simulation</h2><button id="run" disabled>Run training</button><p class="hint">Select or create a network below to enable training. Review settings on tab 2 before running.</p><p>Supports Formula steering/acceleration, speed/raycast sensors, and 2–4 layers of up to 16 nodes.</p>
 <p id="computeBlock"><label for="backend">Compute</label><select id="backend" disabled><option value="cpu">CPU — no GPU allowance</option><option id="gpuOption" value="gpu" disabled>GPU — uses GPU allowance</option></select></p><p id="backendNote" class="hint">Checking available backends…</p>
 <div id="connectionRecovery" hidden><button id="retry" type="button">Reconnect to app</button><p class="hint">Checks server availability. Does not restart training, sign you in, or reset GPU allowance.</p></div>
@@ -249,10 +275,10 @@
   window.getBatchPlan=()=>window.BatchTraining.plan(read(),q('batch_slot').value);
   const action=fn=>async()=>{try{await fn();}catch(error){message('Not completed: '+error.message);}};
   q('recipe_apply_draft').onclick=action(()=>{applySettings(JSON.parse(q('recipe_draft').value));q('preset').value='custom';q('reward_preset').value='custom';message('Draft applied to the training form. No run started.');});
-  q('recipe_save').onclick=action(()=>{
+  q('recipe_save').onclick=action(async()=>{
     const book=read(),slot=q('recipe_slot').value;
     const next=api.validate({...book,recipes:[...book.recipes.filter(r=>r.slot!==slot),{slot,name:q('recipe_name').value,settings:currentSettings()}]});
-    if(book.recipes.some(r=>r.slot===slot)&&!confirm('Replace recipe slot '+slot+' on this device?'))return;
+    if(book.recipes.some(r=>r.slot===slot)&&!(await window.appConfirm('Replace recipe slot '+slot+' on this device?',{heading:'Replace saved recipe',confirmLabel:'Replace'})))return;
     store.save(next);message('Saved slot '+slot+' in this browser.');
   });
   q('recipe_load').onclick=action(()=>{
@@ -266,14 +292,14 @@
     message('Recipe export requested. No tracks, networks, results or credentials are included by the application.');
   });
   q('recipe_import').onchange=action(async()=>{
-    const file=q('recipe_import').files[0];if(!file)return;if(file.size>32768)throw Error('Recipe file exceeds 32 KB');
+    const input=q('recipe_import'),file=input.files[0];if(!file)return;input.value='';if(file.size>32768)throw Error('Recipe file exceeds 32 KB');
     const book=api.parse(await file.text());
-    if(!confirm('Replace this browser’s recipe library with '+book.recipes.length+' imported recipes? Export a backup first if needed.'))return;
+    if(!(await window.appConfirm('Replace this browser’s recipe library with '+book.recipes.length+' imported recipes? Export a backup first if needed.',{heading:'Import recipe library',confirmLabel:'Replace library'})))return;
     if(!store)throw Error('Browser storage unavailable');store.save(book);
     message('Imported '+book.recipes.length+' recipes locally. No run was started.');q('recipe_import').value='';
   });
-  q('recipe_clear').onclick=action(()=>{
-    if(!confirm('Clear saved recipes and batches for this app on this device? Server run data and downloaded files are unaffected.'))return;
+  q('recipe_clear').onclick=action(async()=>{
+    if(!(await window.appConfirm('Clear saved recipes and batches for this app on this device? Server run data and downloaded files are unaffected.',{heading:'Clear local recipes and batches',confirmLabel:'Clear local data',danger:true})))return;
     if(!store)throw Error('Browser storage unavailable');store.clear();message('Local recipes cleared. Server data and downloads are unaffected.');
   });
   q('batch_save').disabled=false;q('batch_load').disabled=false;
@@ -282,11 +308,11 @@
   q('queue_add').onclick=()=>{
     const text=q('batch_recipe_sequence').value.trim();q('batch_recipe_sequence').value=text?text+', '+q('batch_add_recipe').value:q('batch_add_recipe').value;
   };
-  q('batch_save').onclick=action(()=>{
+  q('batch_save').onclick=action(async()=>{
     const book=read(),slot=q('batch_slot').value;
     const runs=q('batch_recipe_sequence').value.split(',').map(s=>s.trim().toUpperCase());
     const next=api.validate({...book,batches:[...book.batches.filter(b=>b.slot!==slot),{slot,name:q('batch_name').value,carry:q('carry').checked,runs}]});
-    if(book.batches.some(b=>b.slot===slot)&&!confirm('Replace saved batch slot '+slot+'?'))return;
+    if(book.batches.some(b=>b.slot===slot)&&!(await window.appConfirm('Replace saved batch slot '+slot+'?',{heading:'Replace saved batch',confirmLabel:'Replace'})))return;
     store.save(next);message('Batch '+slot+' saved locally.');
   });
   q('batch_load').onclick=action(()=>{