Notes on systems, telecoms infrastructure and problems worth explaining. https://ncartmell.co.uk/writing/ 2026-09-17T12:00:00+01:00 Nathan Cartmell https://ncartmell.co.uk/ © 2026 Nathan Cartmell https://ncartmell.co.uk/writing/the-estate-is-the-hard-part.html 2026-09-17T12:00:00+01:00 2026-09-17T12:00:00+01:00

Rewriting a provisioning server from Python to Kotlin, and cutting thousands of desk phones over to it by DNS. The rewrite was the easy half.

https://ncartmell.co.uk/writing/whitespace-was-load-bearing.html 2026-09-17T11:59:00+01:00 2026-09-17T11:59:00+01:00

Removing newlines from a valid XML document took out desk phone directories across an entire handset range. The XML stayed valid; the line length did not.

https://ncartmell.co.uk/writing/nonces-dont-fit-in-one-process.html 2026-09-17T11:58:00+01:00 2026-09-17T11:58:00+01:00

Digest authentication worked fine on a single Python process. Running several instances behind a load balancer turned a local variable into a distributed systems problem.

https://ncartmell.co.uk/writing/six-vendors-one-api.html 2026-09-17T11:57:00+01:00 2026-09-17T11:57:00+01:00

You can abstract the transport. You cannot abstract the semantics. Notes on writing one provisioning interface across six manufacturers' firmware.

https://ncartmell.co.uk/writing/cheaper-for-the-server.html 2026-09-17T11:56:00+01:00 2026-09-17T11:56:00+01:00

Splitting a monolith stopped any one service holding a whole directory in memory. It also made large directories slow to assemble — and the slowest client sets the limit.

https://ncartmell.co.uk/writing/everyone-checks-in-at-once.html 2026-09-17T11:55:00+01:00 2026-09-17T11:55:00+01:00

Give every phone the same check-in interval and they synchronise. The obvious fix — random jitter — would have rebooted the estate on every cycle.

https://ncartmell.co.uk/writing/writing-to-both.html 2026-09-17T11:54:00+01:00 2026-09-17T11:54:00+01:00

You cannot move a live platform to a new database in one step. Dual writes are how you make the migration boring — and most of the work is in the failure handling.

https://ncartmell.co.uk/writing/expected-failures-are-not-errors.html 2026-09-17T11:53:00+01:00 2026-09-17T11:53:00+01:00

Rolling tracing and structured logging across a platform. The instrumentation was the easy half; making the output mean something took longer.

https://ncartmell.co.uk/writing/the-address-that-has-to-be-right.html 2026-09-17T11:52:00+01:00 2026-09-17T11:52:00+01:00

Two overlapping models had grown up for the address that determines where a 999 call is routed. Consolidating them meant deciding which one had ever been true.

https://ncartmell.co.uk/writing/porting-a-number.html 2026-09-17T11:51:00+01:00 2026-09-17T11:51:00+01:00

Moving a phone number between providers is a regulated, document-backed handover between competitors, and the failure mode is a business losing the number on its van.

https://ncartmell.co.uk/writing/someone-elses-phone-bill.html 2026-09-17T11:50:00+01:00 2026-09-17T11:50:00+01:00

Toll fraud costs real money by the minute. The useful controls turned out to be graduated ones — risk scores that trigger verification rather than walls that block.

https://ncartmell.co.uk/writing/the-socket-is-not-the-call.html 2026-09-17T11:49:00+01:00 2026-09-17T11:49:00+01:00

Building real-time calling into a mobile app means the operating system, your app and a websocket all hold an opinion about whether a call is happening. They disagree.

https://ncartmell.co.uk/writing/presence-is-not-a-boolean.html 2026-09-17T11:48:00+01:00 2026-09-17T11:48:00+01:00

The little light showing whether a colleague is free looks like one bit. It is a composite of registration, call state, do-not-disturb and queue availability, delivered to every client.

https://ncartmell.co.uk/writing/the-rewrite-that-was-never-a-project.html 2026-09-17T11:47:00+01:00 2026-09-17T11:47:00+01:00

Building a chat client inside a phone app over three years, and moving it from UIKit to SwiftUI in a single commit nobody scheduled.

https://ncartmell.co.uk/writing/the-store-is-part-of-the-system.html 2026-09-17T11:46:00+01:00 2026-09-17T11:46:00+01:00

Shipping one codebase as several branded apps through two app stores. Release is not a deployment step, it is a dependency with its own deadlines and its own opinions.

https://ncartmell.co.uk/writing/the-tool-your-colleagues-live-in.html 2026-09-17T11:45:00+01:00 2026-09-17T11:45:00+01:00

Maintaining the internal portal a company runs on. Your users sit near you, nothing is ever prioritised until it breaks, and it is the union of every other system you own.

https://ncartmell.co.uk/writing/one-person-several-accounts.html 2026-09-17T11:44:00+01:00 2026-09-17T11:44:00+01:00

Single sign-on assumes an identity maps to an account. In a white-label platform it often maps to four, and the login flow has to cope before the user has proved anything.

https://ncartmell.co.uk/writing/hiding-a-button-is-not-a-permission.html 2026-09-17T11:43:00+01:00 2026-09-17T11:43:00+01:00

Adding granular permissions to a portal that grew up without them, and keeping the interface honest about what a user can actually do.

https://ncartmell.co.uk/writing/reports-are-a-scheduling-problem.html 2026-09-17T11:42:00+01:00 2026-09-17T11:42:00+01:00

Building call reporting end to end. The queries are the easy part; the hard parts are recurrence, drilldowns and deciding what a number means.

https://ncartmell.co.uk/writing/a-catalogue-is-a-query.html 2026-09-17T11:41:00+01:00 2026-09-17T11:41:00+01:00

Selling hardware, numbers and contract addons through partners. What you can order depends on who you are, what you already have, and how long you signed up for.

https://ncartmell.co.uk/writing/decommissioning-is-the-milestone.html 2026-09-17T11:40:00+01:00 2026-09-17T11:40:00+01:00

Replacing the helpdesk your support team works in, without them stopping. The migration is the easy half; turning the old one off is the part that counts.