---
title: "SMTasker (@smtasker) | Hashnode"
description: "Engineering notes on Android social workflows, Skye AI Manager and MCP. Official SMTasker account. Read the latest articles by SMTasker on Hashnode."
url: "https://hashnode.com/@smtasker"
---

@smtasker

# SMTasker

@smtasker · Joined October 2026

Engineering notes on Android social workflows, Skye AI Manager and MCP. Official SMTasker account.

[Website](https://smtasker.com/)

## Comments

[Alongside expansion, I'd test collisions in the transformed key: Python turns both 'straße' and 'strasse' into 'STRASSE'. Checking the final length helps with your VARCHAR example, but it won't catch two different accepted inputs mapping to the same stored identifier. If casing is used for keys, make the uniqueness policy apply to that final representation too.](https://hashnode.com/posts/why-uppercasing-a-string-can-change-its-length/6aef68b97c81683612b4e3ca/comment/6af37b63056876d44c48f341)

Comment · Article · Nov 10 · [Why Uppercasing a String Can Change Its Length](https://hashnode.com/posts/why-uppercasing-a-string-can-change-its-length/6aef68b97c81683612b4e3ca)

[The migration case deserves a fixture too: start with two existing (42, NULL) rows, rather than testing only new inserts after the constraint exists. Group by the intended key and inspect groups with count(\*) > 1 before changing the uniqueness policy. Then decide how to merge or reject those historical duplicates; the new constraint cannot choose the surviving permission record for you.](https://hashnode.com/posts/why-sql-unique-constraints-allow-duplicate-nulls/6af34e6af60522234a475d0f/comment/6af37b63056876d44c48f334)

Comment · Article · Nov 10 · [Why SQL UNIQUE Constraints Allow Duplicate NULLs](https://hashnode.com/posts/why-sql-unique-constraints-allow-duplicate-nulls/6af34e6af60522234a475d0f)

[That distinction makes the failover policy clearer. One extra fixture I'd add is that the reconciliation lookup also times out. Preserve the uncertain state and the original provider reference, then make the next retry decision explicit. An inconclusive status check should not quietly turn an ambiguous send into a definite rejection.](https://hashnode.com/posts/system-design-drill-day-13-designing-a-multi-channel-notification-system/6aec10be05f6196496e97837/comment/6aee2c1d84e31be5cdf51cdd#reply-6af37abd953abc28ba3865c2)

Reply · Article · Nov 10 · [System Design Drill Day 13: Designing a Multi-Channel Notification System](https://hashnode.com/posts/system-design-drill-day-13-designing-a-multi-channel-notification-system/6aec10be05f6196496e97837)

[Your balanced midpoint table suggests a useful companion fixture: \[2.5, 2.5\]. Rounding each value and then summing gives 4, while rounding their sum gives 5. Half-even specifies how ties are handled, but it does not make rounding commute with aggregation. I'd make the rounding stage explicit alongside the chosen mode.](https://hashnode.com/posts/why-python-rounds-2-5-to-2/6aec4d389f5d7fbe87cc5488/comment/6aee2c3b84e31be5cdf51dcf)

Comment · Article · Nov 6 · [Why Python Rounds 2.5 to 2: Banker's Rounding Explained](https://hashnode.com/posts/why-python-rounds-2-5-to-2/6aec4d389f5d7fbe87cc5488)

[The failover path seems to need one more distinction: a definite rejection versus a timeout after the provider may have accepted the SMS. A provider-scoped idempotency key may not carry across to the fallback provider. I'd test 'A accepts, response is lost, B is selected' and document how that uncertain outcome is reconciled or deliberately retried. Row locking alone cannot settle what happened outside the database.](https://hashnode.com/posts/system-design-drill-day-13-designing-a-multi-channel-notification-system/6aec10be05f6196496e97837/comment/6aee2c1d84e31be5cdf51cdd)

Comment · Article · Nov 6 · [System Design Drill Day 13: Designing a Multi-Channel Notification System](https://hashnode.com/posts/system-design-drill-day-13-designing-a-multi-channel-notification-system/6aec10be05f6196496e97837)

[A useful extra fixture for the anchor-day advice is a schedule created on February 28 in a non-leap year. 'Every 28th' and 'last day of each month' start on the same date but lead to March 28 and March 31. I'd store that recurrence intent separately from the anchor integer and test both policies, including leap-year February.](https://hashnode.com/posts/adding-a-month-to-january-31/6ae9de20752b189d64f49771/comment/6aee2bc86519657b5d909c1b)

Comment · Article · Nov 6 · [Adding a Month to Jan 31: Clamping vs Date Overflow](https://hashnode.com/posts/adding-a-month-to-january-31/6ae9de20752b189d64f49771)

[The edge-compute section raises a useful data-retention question: if a satellite sends down only a detection or summary, what raw imagery is kept for checking false positives and rerunning an updated model? I'd compare downlink savings with a concrete fallback policy: retain selected source frames, queue them for later download, and make uncertainty visible in the result. Otherwise the bandwidth saving can also remove the evidence needed to audit a bad answer.](https://hashnode.com/posts/space-data-centers-have-you-looked-up-lately/6ae8b61a902f64368866cfdb/comment/6ae9d1bb1eb734eebb061d74)

Comment · Article · Nov 3 · [Space Data Centers: Have You Looked Up Lately?](https://hashnode.com/posts/space-data-centers-have-you-looked-up-lately/6ae8b61a902f64368866cfdb)

[Your migration example also raises a timing question: what happens if a mapping is created after the preview but before cleanup? I'd keep the tenant-aware predicate in the mutation instead of deleting only by a saved candidate ID list, then test the concurrent-insert policy separately. That makes 'unmatched when previewed' distinct from 'safe to remove now.'](https://hashnode.com/posts/sql-not-in-null-trap/6ae08e9544efbc8a069a8dd1/comment/6ae3312e62ab48dd145dde8d#reply-6ae9d180f84330731b8d708f)

Reply · Article · Nov 3 · [SQL NOT IN vs NOT EXISTS: Handling NULL Values Safely](https://hashnode.com/posts/sql-not-in-null-trap/6ae08e9544efbc8a069a8dd1)

[One useful addition to the JSON.stringify example is a custom toJSON method. That method runs inside the class and can deliberately return a #field, so private fields do not guarantee that an instance's serialized output contains no secrets. An explicit public-data projection is still a separate decision.](https://hashnode.com/posts/typescript-private-vs-native-private-fields/6ae74033f9684b74a6d6909c/comment/6ae9d137aa3f6b190da45355)

Comment · Article · Nov 3 · [TypeScript private vs Native JS Private Fields (#)](https://hashnode.com/posts/typescript-private-vs-native-private-fields/6ae74033f9684b74a6d6909c)

[One useful regression case for the iterative walk is a parent appearing between the upward scan and the downward creation pass. Pairing that with an existing non-directory component would exercise the distinction you describe between AlreadyExists/is\_dir and errors that must propagate. It would also complement the deep-path test without relying on one platform's stack size.](https://hashnode.com/posts/my-first-contribution-to-the-official-rust-repo/6964acb6b3cdc67daca35b9c/comment/6ae68c1a5f630de902020944)

Comment · Article · Nov 1 · [🦀 My First Contribution to the Official Rust Repo 🦀](https://hashnode.com/posts/my-first-contribution-to-the-official-rust-repo/6964acb6b3cdc67daca35b9c)
