is there an Rx method that already does this?

sliding_window.cs

namespace RxTest
{
    using System;
    using System.Collections.Generic;
    using System.Linq;

    using Chunk = System.Collections.Generic.List<int>;

    internal class Program
    {
        private static void Main(string[] args)
        {
            var stream = new[] { 10, 11, 12, 0, 1, 10, 11, 12, 13, 14, 15, 2, 4, 6, 7 };
            const int detla = 1;

            var groups = stream.Aggregate(
                new List<Chunk> { new Chunk() },
                (chunks, x) =>
                {
                    var lastChunk = chunks.Last();

                    if (!lastChunk.Any())
                    {
                        lastChunk.Add(x);
                    }
                    else
                    {
                        var lastItem = lastChunk.Last();
                        if (Math.Abs(lastItem - x) <= detla)
                        {
                            lastChunk.Add(x);
                        }
                        else
                        {
                            var newChunk = new Chunk { x };
                            chunks.Add(newChunk);
                        }
                    }

                    return chunks;
                }).ToList();

            groups.ForEach(
                chunk =>
                {
                    Console.WriteLine("");
                    chunk.ForEach(x => Console.Write(x + " "));
                });

            Console.ReadLine();
        }
    }
}

I'm looking for a better way to do this. Some more Rx-ish.

My exact scenario is that I have a stream of messages with timestamps. If a new message occurs within a given timespan since the previous message, I want to group it with that previous message. Otherwise, I start a new group.

Message[] -> Message[][]

In reality, my message group is a class with additional metadata and not just a Message[], but whatevs.