# Penpot 2.17: Background Blur, Smarter Design Tokens, and More > [!question] Lingering Questions > - What evidence backs the "faster propagation" and "more reliable" claims in the release video — no benchmark or defect-rate comparison is given, just the announcement. > - Does native background blur measurably degrade text contrast in practice, or only in a worst-case theoretical sense the release doesn't test? > - Is the transcript's "NCP" genuinely a caption error for MCP (Model Context Protocol), and if so, what does the actual assistant-integration feature do? > [!tip] Article Ideas > - The token-visibility and in-context-comment features are a clean comparison point for a piece on WordPress Global Styles / `theme.json` collaborative editing — Penpot's open-source design-token workflow vs. WordPress's structured-style workflow, both trying to make system-level decisions visible during ordinary editing. > [!link] Cross Reference > - [[3. Resources/Permanent Notes/In-Context AI Assistants Lower the Adoption Barrier More Than Standalone Apps]] — same "keep the tool inside the existing workspace" logic, applied to AI assistants rather than design review > - [[3. Resources/Permanent Notes/Open-Source Design Tools Reach Parity Through Community-Contribution Velocity]] — Penpot as a recurring case study in open-source design-tool maturation > - [[3. Resources/MOCs/Visual Thinking MOC]] > [!quote] Vocabulary > - *WebGL rendering path* — Penpot's GPU-accelerated canvas renderer, still beta and disabled by default in this release > - *Design token propagation* — how a change to a shared token (e.g. a typography scale) flows out to every layer that references it ## Bib - **Source**: [Penpot 2.17 release video](https://www.youtube.com/watch?v=0ZlNzSx-fic) — Penpot, YouTube - **Reader doc**: [[3. Resources/Readwise/Articles/Penpot 2.17 Background blur, smarter design tokens, and more]] ## Author's thesis Penpot 2.17 pairs one headline visual feature (native background blur) with several quieter workflow changes — in-context comments, wider WebGL coverage kept opt-in, clearer AI-assistant integration status, and more visible typography tokens — that the release itself frames as more consequential than the blur effect, though the announcement's performance and reliability claims aren't yet backed by published benchmarks. ## Key arguments (≤3) 1. **Native effects remove round-trip work** — background blur built into the canvas removes a reason to fake the effect with external tools or exported layers. 2. **Review friction drops when comments stay in context** — reviewers can discuss feedback without leaving the workspace, a small feature the release treats as a meaningful coordination win. 3. **Default-off betas separate testing from exposure** — WebGL reaches the prototype viewer but stays opt-in, letting the team collect real-project feedback without broad-rollout risk. ## Claim atoms ### [H1] Native effects remove round-trip design work > [!quote] ID:rw_01ky7v5vb08net08wqxzxptcky > "You can now blur everything that appears behind a liar directly inside Penpot." (caption error for "layer") Where/why: 00:19–00:29; useful when comparing native tooling against external-effect workarounds. ### [H2] In-context comments reduce review switching > [!quote] ID:rw_01ky7v5w772ytn2jz5b460y9n1 > "You can now access and use comments while you are designing without needing to leave your current workspace." Where/why: 01:01–01:20; design-review workflow — source of the Permanent note below. ### [H3] Default-off betas separate testing from exposure > [!quote] ID:rw_01ky7v1st8weqykmz1qq2m4xw4_h3 > "WebGL is still in beta, and it remains disabled by default." Where/why: 01:39–01:46; renderer rollout — a progressive-delivery example for high-impact infrastructure changes. ## Permanent candidates - **"Feedback moves faster when it stays inside the artifact being reviewed"** — the strongest generalizable claim (from H2): distinct from the vault's existing in-context-AI-assistant note (that one is about tool adoption; this is about review speed specifically). Promoted below. ## Processing checklist - [x] Connect into the relevant MOC — [[3. Resources/MOCs/Visual Thinking MOC]] - [x] Create Permanent note for the review-context claim - [ ] Verify whether the transcript's "NCP" is a caption error for MCP before citing the assistant-integration feature elsewhere ## Links - [[3. Resources/MOCs/Visual Thinking MOC]] — an open-source design tool's collaboration and token-visibility features, relevant to how visual work gets reviewed and shared - [[3. Resources/Permanent Notes/Review Feedback Moves Faster Inside the Work Context Than Outside It]] — the new Permanent note this source generated