https://rawcull.netlify.app/tags/concurrency/Recent content in Concurrency on RawCullHugoenFri, 27 Mar 2026 09:26:38 +0100https://rawcull.netlify.app/blog/2026/03/26/swift-concurrency-in-rawcull/Thu, 26 Mar 2026 00:00:00 +0000https://rawcull.netlify.app/blog/2026/03/26/swift-concurrency-in-rawcull/<h1 id="swift-concurrency-in-rawcull">Swift Concurrency in RawCull<a class="td-heading-self-link" href="#swift-concurrency-in-rawcull" aria-label="Heading self-link"></a></h1> <p>A summarized document about Concurrency in RawCull.</p> <hr> <h2 id="1--why-concurrency-matters-in-rawcull">1 Why Concurrency Matters in RawCull<a class="td-heading-self-link" href="#1--why-concurrency-matters-in-rawcull" aria-label="Heading self-link"></a></h2> <p>RawCull is a macOS photo-culling application that works with Sony A1 ARW raw files. A single RAW file from the A1 can be 50–80 MB. When you open a folder with hundreds of shots, the app must scan metadata, extract embedded JPEG previews, decode thumbnails, and manage a multi-gigabyte in-memory cache — all while keeping the UI perfectly fluid and responsive at 60 fps. Without concurrency that would be impossible.</p>https://rawcull.netlify.app/blog/2026/03/17/concurrency-model/Tue, 17 Mar 2026 00:00:00 +0000https://rawcull.netlify.app/blog/2026/03/17/concurrency-model/<h1 id="concurrency-model--rawcull">Concurrency Model — RawCull<a class="td-heading-self-link" href="#concurrency-model--rawcull" aria-label="Heading self-link"></a></h1> <blockquote> <p><strong>Files covered:</strong></p> <ul> <li><code>RawCull/Model/ViewModels/RawCullViewModel.swift</code></li> <li><code>RawCull/Views/RawCullSidebarMainView/extension+RawCullView.swift</code></li> <li><code>RawCull/Actors/ScanFiles.swift</code></li> <li><code>RawCull/Actors/ScanAndCreateThumbnails.swift</code></li> <li><code>RawCull/Actors/ExtractAndSaveJPGs.swift</code></li> <li><code>RawCull/Actors/ThumbnailLoader.swift</code></li> <li><code>RawCull/Actors/RequestThumbnail.swift</code></li> <li><code>RawCull/Actors/SharedMemoryCache.swift</code></li> <li><code>RawCull/Actors/DiskCacheManager.swift</code></li> <li><code>RawCull/Actors/SaveJPGImage.swift</code></li> <li><code>RawCull/Enum/SonyThumbnailExtractor.swift</code></li> <li><code>RawCull/Enum/JPGSonyARWExtractor.swift</code></li> </ul> </blockquote> <hr> <h2 id="overview">Overview<a class="td-heading-self-link" href="#overview" aria-label="Heading self-link"></a></h2> <p>RawCull uses Swift structured concurrency (<code>async</code>/<code>await</code>, <code>Task</code>, <code>TaskGroup</code>, and <code>actor</code>) across four primary flows:</p> <table> <thead> <tr> <th>Flow</th> <th>Entry point</th> <th>Core actor(s)</th> <th>Purpose</th> </tr> </thead> <tbody> <tr> <td>Catalog scan</td> <td><code>RawCullViewModel.handleSourceChange(url:)</code></td> <td><code>ScanFiles</code></td> <td>Scan ARW files, extract metadata, load focus points</td> </tr> <tr> <td>Thumbnail preload</td> <td><code>RawCullViewModel.handleSourceChange(url:)</code></td> <td><code>ScanAndCreateThumbnails</code></td> <td>Bulk-populate the thumbnail cache for a selected catalog</td> </tr> <tr> <td>JPG extraction</td> <td><code>extension+RawCullView.extractAllJPGS()</code></td> <td><code>ExtractAndSaveJPGs</code></td> <td>Extract embedded JPEG previews and save to disk</td> </tr> <tr> <td>On-demand thumbnails</td> <td>UI grid + detail views</td> <td><code>ThumbnailLoader</code>, <code>RequestThumbnail</code></td> <td>Rate-limited, cached per-file thumbnail retrieval</td> </tr> </tbody> </table> <p>The two long-running operations (thumbnail preload and JPG extraction) share a <strong>two-level task pattern</strong>:</p>