Merge grpc-event-decompile-confirm: stock-client decompile + frame capture + event-parser fix

Three steps verifying the gRPC event read against the provided 2023 R2 client:

- Decompiled the stock managed client (Archestra.Historian.GrpcClient, HistorianAccess):
  confirms no hidden client-side difference. The stock client is gRPC-Web/HTTP-1.1 (same
  transport as ours), m_metadata is gzip-only, the ClientInterceptor is a no-op, and it
  presents no TLS client cert. The event-query orchestration lives in the native C++ core.
- Captured decrypted HTTP/1.1 frames of a native capture-event (50 rows) vs our SDK (0
  rows) through a TLS-terminating tee proxy. Found the native splits services across 5
  connections and runs the event query on a dedicated RetrievalService connection; tested
  replicating that (HISTORIAN_GRPC_EVENT_SPLIT_CHANNEL) -> still 0 rows, so the server
  correlates by session handle, not connection. Topology is not the gate.
- Verified the parse path against the provided client's real result buffer (50 events) and
  fixed a latent bug: the event-row buffer is version + rowCount + a one-time 0x1E header
  field then MARKERLESS rows; the parser wrongly treated 0x1E as a per-row marker and
  decoded only the first row of any multi-row buffer. This also affected the shipped WCF
  event read (identical v9 header). Fixed to a 10-byte header + markerless rows, accepting
  version 9 (WCF) and 11 (gRPC). The real 50-row buffer now decodes to exactly 50 events.

Net: every client-side angle for the gRPC zero-rows is exhausted (payload, transport,
metadata/cert, topology, data store) -> the gate is a server-internal per-connection
retrieval working-set. The parse path is now verified against real 2023 R2 event data on
both transports, and WCF event reads now correctly return all rows of a multi-row buffer.
gRPC event-row retrieval stays auth-solved / parse-verified / retrieval-server-gated.
328 offline tests pass.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01B6mcaT2PjRFKcogzp9UkfC

docs/reverse-engineering/grpc-event-query-capture.md

@@ -370,29 +370,34 @@ live-server access reference.
    retained for future connection-level probing. This eliminates the leading hypothesis and tightens the
    conclusion: the server scopes 0 events to our connection at a layer **above** the gRPC transport.
 
-2. **TLS client identity / certificate.** The native used `SecurityMode=TransportCertificate`. Determine
+2. ~~**TLS client identity / certificate.**~~ **DISPROVEN 2026-06-23 (decompile + capture).** The stock
-   whether it presents a **client certificate** the server uses to scope events (our SDK presents none —
+   client's `GrpcClientBase.InitializeBase` creates a bare `HttpClientHandler` and sets only
-   `AllowUntrustedServerCertificate=true`, server cert only). TEST: capture the TLS handshake (e.g.
+   `ServerCertificateCustomValidationCallback` — it **never adds a client certificate**. The TLS-tee
-   `SSLKEYLOGFILE` + Wireshark, or a decrypting proxy) for a native `capture-event` run and check the
+   capture (below) confirms `clientCert=none` on every native connection. So the native presents no client
-   Certificate message; if a client cert is presented, replicate it.
+   cert; this is not the gate.
 
-3. **HTTP/2-level capture.** The byte[]/handle capture is RPC-payload only. Capture the actual HTTP/2
+3. ~~**HTTP/2-level / connection-frame capture.**~~ **DONE 2026-06-23 — topology difference found, tested,
-   frames (HEADERS/SETTINGS/stream IDs, connection reuse) for the native run vs ours — via a
+   NULL.** Built a TLS-terminating tee proxy (`artifacts/.../httpcap/`, gitignored: self-signed server
-   TLS-decrypting mitm on the loopback forward — to see any connection-level header/affinity our capture
+   cert, forwards through the loopback tunnel, logs decrypted HTTP/1.1 + gRPC-Web both ways) and ran a
-   can't see.
+   **native `capture-event` (returns 50 rows) and our SDK diagnostic (0 rows) through the same
-
+   proxy/upstream**. Note: the stock client is gRPC-Web/HTTP-1.1 (not HTTP/2 — `alpn` empty), so the
-2. **TLS client identity / certificate.** The native used `SecurityMode=TransportCertificate`. Determine
+   capture is HTTP/1.1 framing. Findings:
-   whether it presents a **client certificate** the server uses to scope events (our SDK presents none —
+   - **Connection topology differs.** The native opens **5 TLS connections, one per service** —
-   `AllowUntrustedServerCertificate=true`, server cert only). TEST: capture the TLS handshake (e.g.
+     `HistoryService` (ExchangeKey/OpenConnection/Register/EnsureTags), `StatusService` (×2), and
-   `SSLKEYLOGFILE` + Wireshark, or a decrypting proxy) for a native `capture-event` run and check the
+     **`RetrievalService` (the event query: GetRetrievalInterfaceVersion → StartEventQuery → GetNext →
-   Certificate message; if a client cert is presented, replicate it. **Lower-probability after #1: the
+     EndEventQuery) on its own dedicated connection**. Our SDK collapses **every service onto one
-   plain-HTTP/2 path presents no client cert either, yet auth + registration still succeed and the gate
+     connection**. (Matches the decompile: stock has a separate `GrpcClientBase` per service.)
-   persists — so the gate is not at the TLS-identity layer the cert would affect.**
+   - **Framing differs** (benign): native uses `content-length` + `Expect: 100-continue`; SDK uses
-
+     `transfer-encoding: chunked`. The server accepts both (our `StartEventQuery` returns a valid handle),
-3. **HTTP/2-level capture.** The byte[]/handle capture is RPC-payload only. Capture the actual HTTP/2
+     so framing is not the gate. No extra/hidden header on either side; `clientCert=none` throughout.
-   frames (HEADERS/SETTINGS/stream IDs, connection reuse) for the native run vs ours — via a
+   - **TESTED the topology hypothesis (`HISTORIAN_GRPC_EVENT_SPLIT_CHANNEL=1`):** ran
-   TLS-decrypting mitm on the loopback forward — to see any connection-level header/affinity our capture
+     `StartEventQuery`/`GetNext`/`EndEventQuery` on a **dedicated RetrievalService connection** (no
-   can't see.
+     re-handshake, reusing the session handle — exactly mirroring native conn4), registration staying on
+     the main connection. **Result: still `0B00000000001E000000` (0 rows), `QH=1063`.** Splitting the
+     event query onto its own connection — the one concrete structural difference the capture revealed —
+     **does not make rows flow.** So the server correlates by session handle, not by connection, and the
+     topology is **not** the row-scoping gate. The `CreateHttp2`/`SPLIT_CHANNEL` switches + the
+     `httpcap` proxy are retained as diagnostics.
 
 4. ~~**Server-side ground truth.**~~ **ANSWERED 2026-06-23 (DISPROVES the data-scoping premise).** Via
    the SOCKS→SQL relay (read-only; `artifacts/.../sqlschema/`, gitignored), dumped the full event schema
@@ -420,15 +425,77 @@ live-server access reference.
    `Microsoft.Data.SqlClient`; authenticate with the server's SQL login, not the domain Historian acct —
    creds in the gitignored creds file).
 
-**Conclusion (after #1 disproven + #4 answered).** Three independent angles are now exhausted: client
+### Stock managed client decompiled (2026-06-23) — confirms no hidden client-side difference
-payload (byte-identical), transport (native HTTP/2 == gRPC-Web, both 0 rows), and data store (global,
+
-unscoped, 71,332 events the engine serves via INSQL but withholds from our gRPC connection). The gate is
+Closing the gap that prior cycles left: the zero-rows conclusion had leaned on **wire capture**
-a **server-internal per-connection retrieval working-set** that a pure-managed client cannot reconstruct
+(`instrument-grpc-nonstream`, which only hooks `byte[]` params on `Grpc*Client` methods) — blind to gRPC
-by matching wire bytes, transport, or data. The remaining angles (#2 client-cert, #3 HTTP/2-frame
+metadata/headers, interceptors, channel options, and any non-`byte[]` call. Read the **stock managed
-capture) are low-probability — #1 showed auth+registration succeed with no client cert over plain HTTP/2
+client source directly** (`histsdk-2023r2-analysis/decompiled/Archestra.Historian.GrpcClient` +
-and the gate still holds. **gRPC event-row retrieval stands documented as auth-solved /
+`HistorianAccess`; the pure-managed assemblies decompile cleanly even though the mixed-mode
-retrieval-server-gated**; `ReadEventsAsync` over gRPC keeps the honest no-row throw, and event reads use
+`aahClientManaged.dll` crashes ILSpy). Findings:
-the WCF transport.
+
+- **`GrpcClientBase.InitializeBase` builds the same channel we do.** `GrpcWebHandler((GrpcWebMode)0,
+  HttpClientHandler)` with `HttpVersion = 1.1` — i.e. **the stock client speaks gRPC-Web over HTTP/1.1,
+  the same transport as our SDK.** This *corrects the premise of hypothesis #1*: there was never a native
+  `Grpc.Core` HTTP/2 path to differ from — the stock client that returns 50 rows is itself gRPC-Web. The
+  HTTP/2 disproof's *conclusion* stands (and is reinforced: identical transport on both sides).
+- **`m_metadata` passed to every RPC (incl. `StartEventQuery`/`GetNextEventQueryResultBuffer`) is only
+  `grpc-internal-encoding-request: gzip`** — exactly our header set. No connection-id, session token, or
+  auth header rides in gRPC metadata. The **`ClientInterceptor` is a no-op** (`LogCall` is empty; both
+  unary overloads just invoke the continuation). So the "invisible per-connection metadata/header" blind
+  spot is **confirmed empty** — there is no hidden client-side identity the `byte[]` capture missed.
+- **The event-read query orchestration is genuinely not in managed code.** `CreateEventQuery` /
+  `EventQuery.StartQuery` / `MoveNext` are not in the managed `HistorianAccess`; the managed
+  `GrpcRetrievalClient.StartEventQuery` is a thin one-RPC stub. The query logic lives in the native
+  C++/CLI `HistorianClient` core (the mixed-mode part ILSpy can't decompile) — consistent with the
+  working-set being native/server-side, not a managed step we could read and replicate.
+
+So **every client-controllable layer is now confirmed identical by reading the stock source**, not just
+by wire match: request bytes, transport, channel options, gRPC metadata, interceptor. The remaining
+difference is below the managed surface (native core) / server-side.
+
+**Conclusion (after #1–#4 + stock client decompiled + TLS-tee capture).** Every angle is now exhausted:
+- **client payload** — byte-identical (IL capture + decompile);
+- **transport** — stock client is *also* gRPC-Web/HTTP-1.1; native HTTP/2 makes no difference, both 0 rows;
+- **client metadata/interceptor/channel** — decompiled: identical gzip-only header, no-op interceptor, no
+  client cert; the TLS-tee capture confirms no hidden header and `clientCert=none`;
+- **connection topology** — the native splits services across 5 connections and queries on a dedicated
+  RetrievalService connection; replicating that (`SPLIT_CHANNEL`) still returns 0 rows → the server
+  correlates by session handle, not connection;
+- **data store** — global, unscoped; 71,332 events the engine serves via INSQL but withholds from our
+  gRPC connection.
+
+The gate is a **server-internal per-connection retrieval working-set** that a pure-managed client cannot
+reconstruct by matching wire bytes, transport, metadata, topology, or data — and the establishing logic is
+in the native `HistorianClient` C++ core, not in any decompilable managed step or observable on the wire.
+**gRPC event-row retrieval stands documented as auth-solved / retrieval-server-gated**; `ReadEventsAsync`
+over gRPC keeps the honest no-row throw, and event reads use the WCF transport. Diagnostics retained for
+any future server-side investigation: the `httpcap` TLS-tee proxy, the `CreateHttp2` / `SPLIT_CHANNEL`
+switches, the `EventReadDiagnostic` test, and the `capture-event` harness (native, returns rows).
+
+### Verify the parse path against the provided client's real data (2026-06-23) — found + fixed a latent bug
+
+Used the provided 2023 R2 client as an **oracle**: the `capture-event` harness returns 50 real events
+(verified live + through the `httpcap` proxy), and the `instrument-grpc-nonstream` rewrite captured the
+exact `GetNextEventQueryResultBuffer.result` buffer the stock client received — **63,192 bytes, version
+`0x0B` (11), rowCount 50** (25 `Alarm.Set` + 25 `Alarm.Clear`). Fed that real buffer through our
+`HistorianEventRowProtocol.Parse` to verify the read path decodes genuine gRPC event data, and it
+**exposed a latent parser bug**:
+
+- The real row buffer is `version(2) + rowCount(4) + headerField(4, =0x1E)` then **markerless rows**
+  (`rowFormat(2)=7 + filetime(8) + 8×u16 slots + compact-ascii type + propCount + props`). Our parser
+  wrongly treated the one-time `0x1E` field as a **per-row marker** and re-consumed `[marker+format]`
+  every row — so it parsed only the **first** row of any multi-row buffer and stopped. This is **not
+  gRPC-specific**: the captured **WCF v9** buffer has the identical `0900 <rowCount> 1E000000 0700 …`
+  header, so the shipped WCF event read had the same latent multi-row truncation.
+- **Fix:** read a 10-byte buffer header (skip the `0x1E` field once) and parse markerless rows; accept
+  container version **9 (WCF) and 11 (gRPC)**. Verified: the real 50-row buffer now decodes to exactly 50
+  events, ending cleanly at end-of-buffer (`Parse_RealStockClientCapture_DecodesAllEvents`, gated on
+  `HISTORIAN_EVENT_CAPTURE_NDJSON`); plus a synthetic v11 golden test. 328 offline tests pass.
+
+So the **parse path is now verified against the provided client's real event data** — the one remaining
+gap is strictly the server delivering rows to our gRPC connection (the working-set gate above). If that
+were ever opened, the decoded events would now flow through correctly on both transports.
 
 **2 of 3 layers cleared** (key exchange + client key); the 3rd (token construction) is localized to a
 specific managed method, pending dnlib extraction. ExchangeKey + the v8 serializer are committed; the

src/AVEVA.Historian.Client/Grpc/HistorianGrpcEventOrchestrator.cs

@@ -281,9 +281,21 @@ internal sealed class HistorianGrpcEventOrchestrator
         HistorianEventFilter? filter,
         CancellationToken cancellationToken)
     {
-        var retrievalClient = new GrpcRetrieval.RetrievalService.RetrievalServiceClient(connection.Channel);
+        // HTTP/2-frame capture (grpc-event-query-capture.md #3) showed the stock client runs the event
+        // query on a DEDICATED RetrievalService TLS connection, separate from the HistoryService
+        // connection that opened+registered the session (correlated only by the session handle); our SDK
+        // collapses every service onto one connection. Opt in via HISTORIAN_GRPC_EVENT_SPLIT_CHANNEL=1 to
+        // run StartEventQuery/GetNext/EndEventQuery on their own connection (mirrors native conn4: no
+        // re-handshake, just the existing handle), to test whether topology is the row-scoping gate.
+        bool splitChannel = string.Equals(
+            Environment.GetEnvironmentVariable("HISTORIAN_GRPC_EVENT_SPLIT_CHANNEL"), "1", StringComparison.Ordinal);
+        HistorianGrpcConnection rconn = splitChannel ? HistorianGrpcChannelFactory.Create(_options) : connection;
+        try
+        {
+
+        var retrievalClient = new GrpcRetrieval.RetrievalService.RetrievalServiceClient(rconn.Channel);
         GrpcRetrieval.GetRetrievalInterfaceVersionResponse retrievalVersion = retrievalClient.GetRetrievalInterfaceVersion(
-            new GrpcRetrieval.GetRetrievalInterfaceVersionRequest(), connection.Metadata, Deadline(), cancellationToken);
+            new GrpcRetrieval.GetRetrievalInterfaceVersionRequest(), rconn.Metadata, Deadline(), cancellationToken);
         HistorianServerVersionGate.Validate(HistorianServiceInterface.Retrieval, retrievalVersion.UiVersion, _options);
 
         // Version 6 envelope: the stock 2023 R2 client sends v6 (the WCF path's v5 request is accepted
@@ -306,7 +318,7 @@ internal sealed class HistorianGrpcEventOrchestrator
                 UiQueryRequestType = HistorianEventQueryProtocol.QueryRequestTypeEvent,
                 BtRequest = ByteString.CopyFrom(requestBuffer)
             },
-            connection.Metadata,
+            rconn.Metadata,
             Deadline(),
             cancellationToken);
 
@@ -331,7 +343,7 @@ internal sealed class HistorianGrpcEventOrchestrator
                 {
                     nextResponse = retrievalClient.GetNextEventQueryResultBuffer(
                         new GrpcRetrieval.GetNextEventQueryResultBufferRequest { UiHandle = session.ClientHandle, UiQueryHandle = queryHandle },
-                        connection.Metadata,
+                        rconn.Metadata,
                         EventPollDeadline(),
                         cancellationToken);
                 }
@@ -384,7 +396,12 @@ internal sealed class HistorianGrpcEventOrchestrator
         }
         finally
         {
-            EndEventQuerySafely(retrievalClient, connection, session.ClientHandle, queryHandle);
+            EndEventQuerySafely(retrievalClient, rconn, session.ClientHandle, queryHandle);
+        }
+        }
+        finally
+        {
+            if (splitChannel) { rconn.Dispose(); }
         }
     }
 

src/AVEVA.Historian.Client/Wcf/HistorianEventRowProtocol.cs

@@ -10,10 +10,10 @@ namespace AVEVA.Historian.Client.Wcf;
 /// native event read (instrument-wcf-readmessage record 24, two rows for Alarm.Set + Alarm.Clear):
 ///
 /// <code>
-/// UInt16 version = 9
+/// UInt16 version = 9 (WCF) | 11 (2023 R2 gRPC)
 /// UInt32 rowCount
+/// UInt32 headerField = 0x1E             // ONE buffer-level field (NOT a per-row marker)
 /// rowCount × Row {
-///   UInt32 rowMarker = 0x1E
 ///   UInt16 rowFormat = 7
 ///   Int64  eventTimeUtcFiletime
 ///   UInt16 × 8                          // purpose unclear (slot offsets?)
@@ -42,10 +42,23 @@ namespace AVEVA.Historian.Client.Wcf;
 internal static class HistorianEventRowProtocol
 {
     public const ushort EventRowProtocolVersion = 9;
-    public const uint RowMarker = 0x0000001Eu;
+
-    public const ushort RowFormatV9 = 7;
+    /// <summary>
-    private const int HeaderSize = 6;
+    /// 2023 R2 gRPC returns the event-row buffer with container version <c>11</c> instead of the
-    private const int RowFixedHeaderSize = 4 + 2 + 8 + 16;
+    /// 2020 WCF <c>9</c>. The row layout is otherwise <b>byte-identical</b> (verified against a captured
+    /// stock-client read: header <c>0B00 &lt;rowCount&gt; 1E000000</c> then markerless rows, 50
+    /// Alarm.Set/Alarm.Clear rows decoded clean to end-of-buffer; the WCF v9 capture has the same
+    /// <c>0900 &lt;rowCount&gt; 1E000000</c> header). Accept both, exactly as the interface-version gate
+    /// accepts History 11 and 12.
+    /// </summary>
+    public const ushort EventRowProtocolVersionGrpc = 11;
+
+    /// <summary>Constant buffer-level field following <c>rowCount</c> (observed <c>0x1E</c>). NOT a
+    /// per-row marker — it appears exactly once, before the first row.</summary>
+    public const uint BufferHeaderField = 0x0000001Eu;
+    public const ushort RowFormat = 7;
+    private const int BufferHeaderSize = 2 + 4 + 4;   // version + rowCount + headerField
+    private const int RowFixedHeaderSize = 2 + 8 + 16; // rowFormat + filetime + 8×UInt16 slots
 
     private const byte ValueTypeBool = 0x02;
     private const byte ValueTypeGuid = 0x10;
@@ -55,25 +68,26 @@ internal static class HistorianEventRowProtocol
 
     public static IReadOnlyList<HistorianEvent> Parse(ReadOnlySpan<byte> buffer)
     {
-        if (buffer.Length < HeaderSize)
+        if (buffer.Length < 6)
         {
             return [];
         }
 
         ushort version = BinaryPrimitives.ReadUInt16LittleEndian(buffer[..2]);
-        if (version != EventRowProtocolVersion)
+        if (version != EventRowProtocolVersion && version != EventRowProtocolVersionGrpc)
         {
             return [];
         }
 
         uint rowCount = BinaryPrimitives.ReadUInt32LittleEndian(buffer.Slice(2, 4));
-        if (rowCount == 0)
+        if (rowCount == 0 || buffer.Length < BufferHeaderSize)
         {
             return [];
         }
 
         List<HistorianEvent> events = new(checked((int)rowCount));
-        int cursor = HeaderSize;
+        // Skip the single buffer-level header field (0x1E); rows follow with NO per-row marker.
+        int cursor = BufferHeaderSize;
         for (uint rowIndex = 0; rowIndex < rowCount; rowIndex++)
         {
             if (!TryReadRow(buffer, ref cursor, out HistorianEvent? row))
@@ -95,19 +109,13 @@ internal static class HistorianEventRowProtocol
             return false;
         }
 
-        uint marker = BinaryPrimitives.ReadUInt32LittleEndian(buffer.Slice(cursor, 4));
+        ushort format = BinaryPrimitives.ReadUInt16LittleEndian(buffer.Slice(cursor, 2));
-        if (marker != RowMarker)
+        if (format != RowFormat)
         {
             return false;
         }
 
-        ushort format = BinaryPrimitives.ReadUInt16LittleEndian(buffer.Slice(cursor + 4, 2));
+        long filetime = BinaryPrimitives.ReadInt64LittleEndian(buffer.Slice(cursor + 2, 8));
-        if (format != RowFormatV9)
-        {
-            return false;
-        }
-
-        long filetime = BinaryPrimitives.ReadInt64LittleEndian(buffer.Slice(cursor + 6, 8));
         DateTime eventTimeUtc = DateTime.FromFileTimeUtc(filetime);
         int afterFixedHeader = cursor + RowFixedHeaderSize;
 

tests/AVEVA.Historian.Client.Tests/HistorianEventRowProtocolTests.cs

@@ -90,6 +90,69 @@ public sealed class HistorianEventRowProtocolTests
         Assert.Equal(500, evt.Properties["priority"]);
     }
 
+    [Fact]
+    public void Parse_Version11GrpcHeader_ParsesRowsIdenticalToV9()
+    {
+        // 2023 R2 gRPC returns the event-row buffer with container version 11; the per-row layout is
+        // byte-identical to the WCF v9 format. The parser must accept both (verified against a captured
+        // stock-client read of 50 Alarm.Set/Alarm.Clear rows whose header began 0B00 .. 1E000000 0700).
+        DateTime t1 = new(2026, 6, 23, 13, 34, 14, DateTimeKind.Utc);
+        DateTime t2 = t1.AddSeconds(10);
+
+        byte[] header = BuildHeader(2u, HistorianEventRowProtocol.EventRowProtocolVersionGrpc); // version 11
+        byte[] buffer = Concat(header, BuildRow(t1, "Alarm.Set", []), BuildRow(t2, "Alarm.Clear", []));
+        IReadOnlyList<HistorianEvent> events = HistorianEventRowProtocol.Parse(buffer);
+
+        Assert.Equal(2, events.Count);
+        Assert.Equal("Alarm.Set", events[0].Type);
+        Assert.Equal(t1, events[0].EventTimeUtc);
+        Assert.Equal("Alarm.Clear", events[1].Type);
+        Assert.Equal(t2, events[1].EventTimeUtc);
+    }
+
+    // Verification against the PROVIDED 2023 R2 client: parse the real GetNextEventQueryResultBuffer
+    // result the stock client received (50 events), proving our read path decodes genuine gRPC event
+    // data. The capture carries customer identity so it is gitignored — point HISTORIAN_EVENT_CAPTURE_NDJSON
+    // at the captured ndjson to run; the test skips cleanly otherwise (no fixture committed).
+    [Fact]
+    public void Parse_RealStockClientCapture_DecodesAllEvents()
+    {
+        string? ndjson = Environment.GetEnvironmentVariable("HISTORIAN_EVENT_CAPTURE_NDJSON");
+        if (string.IsNullOrWhiteSpace(ndjson) || !File.Exists(ndjson))
+        {
+            return; // gated: no capture available
+        }
+
+        byte[]? resultBuffer = null;
+        foreach (string line in File.ReadLines(ndjson))
+        {
+            if (!line.Contains("GetNextEventQueryResultBuffer.result.out")) continue;
+            int i = line.IndexOf("\"Base64\":\"", StringComparison.Ordinal);
+            if (i < 0) continue;
+            i += "\"Base64\":\"".Length;
+            int j = line.IndexOf('"', i);
+            resultBuffer = Convert.FromBase64String(line.Substring(i, j - i));
+            break;
+        }
+
+        Assert.NotNull(resultBuffer);
+        ushort version = BinaryPrimitives.ReadUInt16LittleEndian(resultBuffer.AsSpan(0, 2));
+        uint rowCount = BinaryPrimitives.ReadUInt32LittleEndian(resultBuffer.AsSpan(2, 4));
+        Assert.Equal(HistorianEventRowProtocol.EventRowProtocolVersionGrpc, version); // real gRPC buffer is v11
+
+        IReadOnlyList<HistorianEvent> events = HistorianEventRowProtocol.Parse(resultBuffer);
+
+        // Our parser decodes every row the stock client received.
+        Assert.Equal((int)rowCount, events.Count);
+        Assert.All(events, e =>
+        {
+            Assert.False(string.IsNullOrEmpty(e.Type));
+            Assert.NotEqual(default, e.EventTimeUtc);
+        });
+        // Sanitized cross-check: only the generic AVEVA event types (no customer fields asserted).
+        Assert.All(events, e => Assert.Contains(e.Type, new[] { "Alarm.Set", "Alarm.Clear" }));
+    }
+
     [Fact]
     public void Parse_UnknownTypeMarker_KeepsRawBytesInPropertyBag()
     {
@@ -120,11 +183,15 @@ public sealed class HistorianEventRowProtocolTests
         Assert.Equal("Alarm.Set", events[0].Type);
     }
 
-    private static byte[] BuildHeader(uint rowCount)
+    private static byte[] BuildHeader(uint rowCount) => BuildHeader(rowCount, HistorianEventRowProtocol.EventRowProtocolVersion);
+
+    private static byte[] BuildHeader(uint rowCount, ushort version)
     {
-        byte[] header = new byte[6];
+        // version(2) + rowCount(4) + the single buffer-level header field (0x1E). Rows are markerless.
-        BinaryPrimitives.WriteUInt16LittleEndian(header.AsSpan(0, 2), HistorianEventRowProtocol.EventRowProtocolVersion);
+        byte[] header = new byte[10];
+        BinaryPrimitives.WriteUInt16LittleEndian(header.AsSpan(0, 2), version);
         BinaryPrimitives.WriteUInt32LittleEndian(header.AsSpan(2, 4), rowCount);
+        BinaryPrimitives.WriteUInt32LittleEndian(header.AsSpan(6, 4), HistorianEventRowProtocol.BufferHeaderField);
         return header;
     }
 
@@ -141,13 +208,13 @@ public sealed class HistorianEventRowProtocolTests
             propertyBlockSize += propertyBlocks[i].Length;
         }
 
-        byte[] row = new byte[4 + 2 + 8 + 16 + eventTypeBytes.Length + 2 + propertyBlockSize];
+        // Markerless row: rowFormat(2) + filetime(8) + 8×UInt16 slots(16) + type + propCount + props.
+        byte[] row = new byte[2 + 8 + 16 + eventTypeBytes.Length + 2 + propertyBlockSize];
         Span<byte> span = row;
-        BinaryPrimitives.WriteUInt32LittleEndian(span[..4], HistorianEventRowProtocol.RowMarker);
+        BinaryPrimitives.WriteUInt16LittleEndian(span[..2], HistorianEventRowProtocol.RowFormat);
-        BinaryPrimitives.WriteUInt16LittleEndian(span.Slice(4, 2), HistorianEventRowProtocol.RowFormatV9);
+        BinaryPrimitives.WriteInt64LittleEndian(span.Slice(2, 8), eventTimeUtc.ToFileTimeUtc());
-        BinaryPrimitives.WriteInt64LittleEndian(span.Slice(6, 8), eventTimeUtc.ToFileTimeUtc());
         // 16 bytes of zeroed slot ushorts left as-is.
-        int eventTypeOffset = 4 + 2 + 8 + 16;
+        int eventTypeOffset = 2 + 8 + 16;
         eventTypeBytes.CopyTo(span[eventTypeOffset..]);
         BinaryPrimitives.WriteUInt16LittleEndian(span.Slice(eventTypeOffset + eventTypeBytes.Length, 2), propertyCount);
         int cursor = eventTypeOffset + eventTypeBytes.Length + 2;