SShortSingh.
Back to feed

Midnight Blockchain Can Partially Succeed: What Developers Must Know

0
·7 views

Unlike Ethereum or Solana, Midnight's transaction model allows a transaction to partially succeed — with some sections committing to the ledger even if others fail. Every Midnight transaction passes through three sequential stages: well-formedness, a guaranteed phase, and a fallible phase. If a transaction fails during the guaranteed phase it is rejected entirely, but a failure in the fallible phase still preserves any state changes made during the guaranteed phase. Critically, fees for all phases are collected upfront during the guaranteed phase and are not refunded if fallible work fails. Developers building on Midnight must account for this partial-success model to avoid unintended half-updated states and unexpected fee losses.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

A Simple Three-Part Structure to Explain Technical Issues in English

Non-native English speakers often struggle to communicate technical problems clearly when trying to include too many details at once. A structured approach—conclusion, reason, example—can make explanations easier to follow for both the speaker and the listener. The method starts with a clear recommendation, follows with the single most important supporting reason, and ends with a concrete, observable example. This framework reduces reliance on complex vocabulary by anchoring explanations to specific triggers and outcomes rather than broad architectural descriptions. Practicing the structure aloud, then repeating it without notes, helps reinforce the habit of delivering information in a logical order.

0
ProgrammingDEV Community ·

Figma Releases 10 Community AI Skills to Streamline Design Workflows

Figma has published a set of 10 AI-powered skills for its design community, aimed at automating and improving various stages of the design process. Among the notable tools is a skill that scans designs to identify animation opportunities and another that expands a single screen into a full range of UI states, including loading, error, and edge cases. A component handoff skill converts components into shareable canvas documentation, while an analysis skill audits components for duplicates, orphans, and naming inconsistencies. The collection also includes a design review skill that categorizes feedback using a traffic light system, flagging critical blockers, suggested fixes, and elements that are working well. The release reflects Figma's continued push to integrate AI assistance directly into collaborative design workflows.

0
ProgrammingDEV Community ·

Fragmented Project Files Hide Decision History and Cause Team Confusion

When a project spans multiple tools, critical context — like why a pricing model changed — often gets buried across disconnected files. By the end of a typical work week, a team may have a research doc, spreadsheet, slide deck, interface sketch, and chat log with no clear record of their order or relationship. This fragmentation means no team member can easily tell which decision overrides another or trace the reasoning behind key choices. The gap between where decisions are made and where their rationale lives is a common source of rework and confusion. Consolidating work into a single shared space where each file retains its history could help teams avoid losing context as projects evolve.

0
ProgrammingDEV Community ·

Developer Documents 369 Verified Merged Pull Requests Across 33 External GitHub Repos

Developer Aniruddha Adak has publicly documented 369 verified pull requests merged into external GitHub repositories across 33 projects, spanning from January 2024 to August 2026. The figure is derived from a total of 398 lifetime merged pull requests, with 29 self-owned repository merges excluded from the count. Of his 918 total pull requests submitted, fewer than half were merged, a rate he describes as normal for open-source contribution. The bulk of his merges — 306 — came in 2024 during high-volume participation including Hacktoberfest, while 2025 and 2026 saw fewer but more substantial contributions to widely used production tools. All 369 merges are individually linked and independently verifiable via the GitHub search API.