惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

量子位
Recent Announcements
Recent Announcements
D
Docker
V
V2EX
阮一峰的网络日志
阮一峰的网络日志
Vercel News
Vercel News
Microsoft Security Blog
Microsoft Security Blog
The GitHub Blog
The GitHub Blog
U
Unit 42
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
月光博客
月光博客
腾讯CDC
B
Blog
博客园_首页
罗磊的独立博客
D
DataBreaches.Net
IT之家
IT之家
酷 壳 – CoolShell
酷 壳 – CoolShell
L
LangChain Blog
aimingoo的专栏
aimingoo的专栏
MongoDB | Blog
MongoDB | Blog
GbyAI
GbyAI
Stack Overflow Blog
Stack Overflow Blog
M
MIT News - Artificial intelligence

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
Cx Dev Log — 2026-05-16
COMMENTERTHE · 2026-05-17 · via DEV Community

COMMENTERTHE9

The biggest concrete win today? Without a doubt, it was CX-218: a pivotal fix in lower_binary that took us from 8 arithmetic JIT parity passes, with 9 skips, to a much cleaner 13 passes and only 4 skips. This was no minor detail; it unblocked five more operations that had been silently sidestepping execution due to an unresolved type placeholder. Beyond that, the day was driven by determinism test coverage enhancements and some necessary site refinements.

The CX-218 Fix: Tackling SemanticType::Numeric in lower_binary

Whenever the lowering pass reaches an arithmetic expression—think 0 + 0 or 3 * 4—the result was defaulting to SemanticType::Numeric. This placeholder essentially waved a red flag, screaming "numeric type unresolved," until a fix could be devised. Previously, lower_value quietly resolved this to the target’s numeric_literal_ir_type(), defaulting to I64 on 64-bit systems. But lower_binary? Not so much. It was busy calling lower_type(result_ty), which cheerfully replied with UnsupportedSemanticType. Consequently, every test silently exited with a skip, leaving parity untested and unnoticed.

What changed? The fix mirrored lower_value by resolving SemanticType::Numeric directly to ctx.target.numeric_literal_ir_type() within lower_binary. A pragmatic move, even if it slightly duplicated logic across paths. If this tweak sounds like a band-aid, that's because it is. Until a centralized resolution emerges, this approach lets tests run and ensures zero parity failures.

Determinism Test Effort

Eight new PRs have made their way into submain. They are all gradually paving the way for our Phase 12 differential harness strategy, ensuring we remain on the cutting edge of determinism:

  • CX-197 (PR #238) — Improved determinism coverage.
  • CX-199 (PR #240) — Rebased CompoundAssign determinism tests.
  • CX-200 (PR #241) — Unary determinism tests, also rebased and fresh.
  • CX-202 (PR #243) — CodeRabbit’s logical OR fix freshens up CX-198.
  • CX-203 (PR #244) — While-in loop determinism tests, rebased cleanly.
  • CX-204 (PR #245) — Conditional branch merge patterns got some love.
  • CX-213 (PR #255) — Added documentation coverage following submain’s latest changes.
  • CX-214 (PR #256) — Determinism tests that cover complex pointer/array/casts for good measure.

These PRs showcase our revised strategy: each JIT instruction domain now demands a solid determinism test pass prior to implementation being folded into submain. The latest batch covered a broad array of operations: conditional branches, loops, various operators, and data constructs. Everything’s rebased, ready to merge, and straining at the leash, poised to boost our reliability matrix.

The Growing Submain Gap

Submain's pace has outstripped main by 164 commits, skyrocketing from just 40 a week ago. Main’s been sluggish, only seeing daily log updates, and its test matrix stands at 111 passes with a few notable failures. These are usually down to Result/try semantics and looping within functions. Submain, by all indications, clears all 117 tests currently identified. With such a divergence, integration's gonna be one tough cookie: an issue we’ve anticipated for a while, but no signs of a merge plan clear the horizon.

Elevating the Site

In a single day, five new commits polished the site to a gleam:

  • Documentation overhaul, aligned with the most recent interpreter/JIT operations, encompassing arrays, control-flow, and more.
  • Updated homepage status and voice, to reflect current capabilities.
  • Addressed mobile accessibility and fonts to improve user experience.
  • SEO and social metadata enhancements to prime us for public eyes.
  • Concise copy voice to round it all off.

This was no small feat. Aligning the docs with real Cx functionality (as opposed to past aspirational goals) was a massive step. Meanwhile, the SEO and accessibility tweaks hint that we’re getting ready for prime time.

What's Next

First up: push through the merges for CX-215 to CX-222 in submain. Eyes are particularly on CX-218 for its tangible code fixes, while the others bring in valuable determinism tests. The coming cycles promise more of the same: decisive determinism tests coupled with solid implementation once they pass muster.

But what of the submain-to-main merge? With a daunting 164 commits to reconcile, it looms large as our biggest conundrum. The site, freshly polished, seems to stand ready for its debut, assuming it's not waiting on the main branch to catch up. The path forward? Clear in parts but tantalizingly opaque in others.


Follow the Cx language project:

Originally published at https://cx-lang.com/blog/2026-05-16