8 Product Gaps People Are Still Complaining About This Week (June 1, 2026)

Eight product gripes surfaced from HN comments and X posts in the past 48 hours — none of them have a good answer yet. Each one is a specific, repeatable frustration, not a vague wish. Developer tooling, creator monetization, consumer hardware, and AI workflow all show up this week.

1. MCP has no stateless call option — and nobody's building one

"I wish there was better support for stateless calls." — @didibus, commenting on "MCP is dead?" discussion 1

The gripe: The Model Context Protocol requires stateful session management. For quick one-shot tool calls — a database lookup, a schema fetch, a simple REST call — spinning up a full MCP session is wasteful overhead. The commenter's actual point: the protocol design argument that matters isn't "should MCP exist" but "why does it only support stateful connections when most real agent calls don't need persistent state?"

What exists: UTCP (Universal Tool Calling Protocol) launched recently as a lighter alternative, explicitly marketing itself as eliminating "the wrapper tax." But it hasn't hit production-grade adoption yet, and its support for the existing MCP ecosystem is limited.2

The gap: MCP 1.x has no stateless call mode. No major AI provider has shipped one. This means every trivial tool invocation carries session setup cost — particularly painful at the edge, in serverless environments, and in high-throughput agent pipelines.

Feasibility: High. A stateless MCP-compatible transport wrapper could ship as a library. Spec the request/response envelope, skip the session handshake, return results inline. Real demand is there: the "MCP is dead?" thread generated 401 comments and 395 points.

Loading content card…

2. AI tab completion keeps touching code you told it not to touch

"The tab completions are sometimes overeager and I wish I had more control over them." — @dr_kiszonka, commenting in "AI job grief" thread 3

The gripe: The commenter is coping with long LLM waits by writing code with AI tab completions running alongside — but the completions are too aggressive. They override or suggest changes to code that the developer already decided on. There's no "zone of control" — no way to say "complete inside this function, but don't touch this class."

What exists: Cursor, Copilot, and similar tools let you enable/disable completions globally via a keyboard shortcut or settings toggle. That's a blunt instrument. None of them support region-level or semantic-boundary control — "complete within this function scope only," "don't suggest rewrites to frozen files," or "suppress completions while my cursor is inside a marked block."

The gap: Granular, region-aware completion suppression. The locked-code-region idea (preventing AI from rewriting specific AST nodes or file sections) has appeared in prior HN gripes — but no editor ships it as a first-class feature today.

Feasibility: Medium. Needs deep LSP integration to know semantic boundaries. An editor extension could start with file-level suppression via a marker comment (e.g. // ai-freeze-start) and ship in weeks. Full AST-boundary mode is harder but buildable.

3. Patreon is missing basic features and subscribers know it

"I pay for multiple Patreons, too. I wish Patreon wouldn't be such a shitty website/app. It's insane what basic features it's missing." — @blks, commenting on Meta subscriptions thread 4

The gripe: Someone paying for multiple creator subscriptions venting that Patreon, which has existed since 2013 and processes billions in creator payments, still feels like an MVP. No specifics given — which is the tell. When users can't even name the specific missing feature because there are too many to count, that's a product in trouble.

What exists: The Patreon alternatives market is crowded: Circle, Memberful, Substack, Ghost, Buy Me a Coffee, Ko-fi, and dozens of newer entrants.5 A Reddit thread from this week explicitly titled "Patreon sucks, I'm building a replacement" was getting responses from frustrated creators listing specific gaps: better notification system, per-post analytics, native community features, improved DM experience.6

The gap: What none of the alternatives have fully cracked: retaining the subscriber discovery and cross-creator browsing that Patreon's marketplace still provides, combined with a modern UX. Creators leave Patreon for better tools; subscribers stay because their creators are there. Closing that loop requires a destination, not just a payments widget.

Feasibility: High for a focused vertical slice (e.g., "Patreon for indie developers" with GitHub integration, changelog feeds, and release notes as subscription content). The infrastructure problem is solved. The product problem is differentiation.

Loading content card…

4. Smart TVs can't fall back from AV1 even when they're choking on it

"I wish there was an option to still use VP9 / H.264 on those systems (even limited to 1080p)." — @weiliddat, commenting in "Dav2d" thread 7

The gripe: Smart TVs technically "support" AV1, but the hardware decoder is too weak for real-time 1x playback at any meaningful resolution. The result: 1-3 second lag on any UI interaction while watching YouTube in AV1, and some TVs automatically dropping to 480p at 1.25x speed. There's no user-accessible toggle to downgrade to VP9 or H.264 and get smooth performance back.

What exists: YouTube has historically supported codec-specific URL parameters, and browser extensions like h264ify let desktop users force H.264 playback. But smart TV apps — YouTube, Netflix, Prime — expose no such settings. The TV's YouTube app decides the codec; you don't.

The gap: A codec preference setting in smart TV streaming apps, or a router-level tool that intercepts DASH manifest requests and rewrites codec preference headers before they reach the TV. Neither exists as a polished consumer product.

Feasibility: Medium for a router-side tool (requires network-level MITM on HTTPS, which is non-trivial). Lower feasibility for a TV-app solution — would require app developer cooperation. The bigger opportunity might be a browser extension equivalent for Android TV / Fire TV via sideloaded APK.

5. Builders who ship daily want a tool that turns their laptop activity into content

"Posting on X daily as a full time builder is hard. Some days deep in code. Some days debugging for hours. I want a tool that tracks my laptop activity and turns it into posts automatically. Someone build this. Or I will." — @quiora301, X/Twitter 8

The gripe: Building in public requires consistency, but most days a developer's work doesn't produce visible artifacts — no shipped feature, no PR merged, no blog post. Yet the work happened: six hours of debugging a specific race condition, reading three papers, iterating through five prompt drafts. The builder wants that turned into content without manually narrating it.

What exists: ActivityWatch tracks application and window focus time locally.9 Wakatime tracks coding time by language and project for developers. Neither transforms activity into narrative content. Tools like Buffer and Hootsuite require you to write the post; they only handle scheduling and distribution.

The gap: An activity-aware content draft tool that observes what you actually did (terminal commands, file diffs, browser tabs, app focus), infers a narrative ("spent 4 hours on auth bug, fixed it by X"), and drafts a post in your voice. The inference + voice-matching + one-click publish pipeline doesn't exist as a coherent product.

Feasibility: High. Local screen/activity telemetry is solved (ActivityWatch, Wakatime). LLM summarization is cheap. The risky part is privacy: users will not install this if it sends raw activity data to a cloud. Self-hosted or local-model deployment is the only viable distribution path. Strong signal: the tweet itself got bookmarked multiple times; this is a felt need in builder communities.

6. LibreOffice on Wayland still requires a manual env var to be usable

"Needed: Why doesn't it recognize Wayland and perform properly when scrolling?? Or act through XWayland by itself, rather than me having to search for a solution for an hour?" — @zelphirkalt, commenting in Microsoft Office thread 10

The gripe: On Debian 13 with KDE, LibreOffice using Wayland directly produces 1-second scroll lag on every page redraw. The fix — setting QT_QPA_PLATFORM=xcb to force XWayland — requires finding a forum post, understanding the Qt platform abstraction layer, and setting an env var before launch. None of that is user-accessible. The person is a programmer and it still took an hour.

What exists: LibreOffice has partial Wayland support, and multiple distributions ship env-var workaround documentation.11 But auto-detection and graceful fallback are not implemented — LO doesn't benchmark its own rendering performance at startup and decide which backend to use.

The gap: A startup-time performance probe in LibreOffice (or a distro-level wrapper) that renders a test frame, measures redraw latency, and automatically falls back to XWayland if Wayland-native is too slow. This is a solved pattern in graphics drivers; it just hasn't been applied here.

Feasibility: Medium. LibreOffice is open source — a patch could ship, but the project's contributor pipeline is slow. The indie opportunity is a launch wrapper script or a distro package hook that auto-sets the env var based on hardware capability detection. A 100-line shell script would fix this for millions of Linux users.

7. Nobody explains what passkey account recovery actually looks like

"The FAQs on the page says there are 'secure and reliable recovery options' when the device with the passkey is lost. I wish there was more explanation for that because trying to help people recover Microsoft accounts when they don't have the password is mostly when I have used the SMS codes." — @moberley, commenting in "Microsoft to stop sending SMS codes" thread 12

The gripe: Microsoft is deprecating SMS codes and replacing them with passkeys. But the recovery documentation is vague — "secure and reliable recovery options exist" without describing what they are, in what scenarios they work, and what happens when a user loses their only passkey device and has no fallback. The person asking helps non-technical users recover accounts; for them, this matters.

What exists: Microsoft has a formal account recovery flow via the Microsoft Entra admin portal and a user-facing ACSR (Account Self-Cancellation and Recovery) form.13 But the consumer-facing support page doesn't surface this clearly, and the documentation gap between "enterprise admin" and "normal person locked out" is enormous.

The gap: A clear, scenario-based passkey recovery guide that covers the most common failure cases: lost phone, stolen device, upgraded phone with broken passkey sync, dead security key. Not a tech spec — a decision tree. No consumer-focused product category exists for "passkey recovery UX consulting," but the educational content gap is real and no major platform fills it well.

Feasibility: High as a content product. A single well-indexed "how to recover your account when your passkey device is gone" guide per major platform (Microsoft, Apple, Google) would rank immediately and serve millions. A SaaS angle: a recovery-as-a-service tool that pre-configures recovery codes and stores them in a trusted escrow, similar to 1Password Emergency Kit but specifically for passkey fallback.

8. The MLB Ballpark app doesn't cover minor league baseball

"I like the mlb ballpark app but I wish it also tracked minor league games/tickets/stats or that there was at least an milb version that existed." — @sawbathrmtrap04, X/Twitter 14

The gripe: Minor league baseball fans — particularly prospect watchers — have no first-party app parity with the MLB experience. Live scores, pitch-by-pitch Gameday, ticket purchasing, and push notifications don't exist in a polished official MiLB experience that mirrors the MLB Ballpark app.

What exists: An official MiLB app does exist in both the App Store and Google Play.15 But reviews and community feedback consistently describe it as a severely stripped-down experience compared to the major league counterpart. A third-party app called "Prospects" tracks player movement and stats for affiliated MiLB players, but it's not a full game-day experience.16

The gap: A unified MLB + MiLB experience for prospect-following fans. The data exists — box scores, pitch data, transactions — but the consumer product layer is thin. The specific opportunity: a fan app built around prospect tracking, featuring game-day push alerts keyed to specific players (not just teams), cross-level career stats, and call-up/option notifications.

Feasibility: High. Minor league game data is available through MLB's Stats API, which has public documentation. The product work is UI and notification architecture, not data access. Strong demand signal exists in dedicated subreddits and fantasy baseball communities that already manually track this information.

Loading content card…

Source note: HN Algolia API used for HN comments. Individual HN comments don't carry per-comment engagement counts; parent thread upvote scores are used as a proxy demand signal.

8 Product Gaps People Are Still Complaining About This Week (June 1, 2026)

1. MCP has no stateless call option — and nobody's building one

2. AI tab completion keeps touching code you told it not to touch

3. Patreon is missing basic features and subscribers know it

4. Smart TVs can't fall back from AV1 even when they're choking on it

5. Builders who ship daily want a tool that turns their laptop activity into content

6. LibreOffice on Wayland still requires a manual env var to be usable

7. Nobody explains what passkey account recovery actually looks like

8. The MLB Ballpark app doesn't cover minor league baseball

References