<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Web-Engineer</title>
    <description>The latest articles on DEV Community by Web-Engineer (web-engine).</description>
    <link>https://dev.to/web-engine</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Forganization%2Fprofile_image%2F15308%2Fbe5eedc7-e176-4578-aea9-b10a5e0c9db5.png</url>
      <title>DEV Community: Web-Engineer</title>
      <link>https://dev.to/web-engine</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/web-engine"/>
    <language>en</language>
    <item>
      <title>Rescuing the Loupedeck: how we brought LoupixDeck to the Mac</title>
      <dc:creator>Craig Lawson</dc:creator>
      <pubDate>Mon, 16 Nov 2026 17:49:36 +0000</pubDate>
      <link>https://dev.to/web-engine/rescuing-the-loupedeck-how-we-brought-loupixdeck-to-the-mac-2cgi</link>
      <guid>https://dev.to/web-engine/rescuing-the-loupedeck-how-we-brought-loupixdeck-to-the-mac-2cgi</guid>
      <description>&lt;p&gt;On the corner of our desk sits a Loupedeck CT: a slab of knobs, buttons, touch keys and a big centre wheel. It's a lovely bit of hardware, and for a while it was close to becoming a very expensive paperweight. It's now one of the most useful tools we own, thanks to an open-source project called LoupixDeck. We've ended up contributing quite a lot to it.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F0hjmmjglkk6k2d8sb09q.webp" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F0hjmmjglkk6k2d8sb09q.webp" width="800" height="300"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  A brilliant controller with a software problem
&lt;/h2&gt;

&lt;p&gt;Loupedeck started in Helsinki, building hardware consoles for photo and video editing that grew into general-purpose controllers for streamers and creators. Logitech &lt;a href="https://ir.logitech.com/press-releases/press-release-details/2023/Logitech-Acquires-Loupedeck/" rel="noopener noreferrer"&gt;acquired Loupedeck in July 2023&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;In our experience, the software was never the best part of the package, and it didn't get lighter after the acquisition. Each update felt heavier, and for a control surface that should simply respond the moment you touch it, that matters.&lt;/p&gt;

&lt;p&gt;Then, in March 2025, Loupedeck &lt;a href="https://loupedeck.com/blog/important-update-end-of-loupedeck-sales-and-future-development-plans/" rel="noopener noreferrer"&gt;announced it would no longer sell the Loupedeck product line&lt;/a&gt;, switching its focus to Logitech's MX Creative Console. It said support for existing devices would continue, and that the two products share a backend from software release 6.0. Support for the legacy 5.8 software and its marketplace &lt;a href="https://loupedeck.com/us/blog/important-announcement-end-of-support-for-loupedeck-legacy-software-marketplace-5-8/" rel="noopener noreferrer"&gt;ended on 30 March 2026&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;If you own a Loupedeck Live or CT, the direction of travel is clear. The hardware's future is now in someone else's roadmap.&lt;/p&gt;

&lt;h2&gt;
  
  
  Enter LoupixDeck
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8lymno780vt4ctuk4w71.webp" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F8lymno780vt4ctuk4w71.webp" width="600" height="697"&gt;&lt;/a&gt;{.img-right}&lt;/p&gt;

&lt;p&gt;&lt;a href="https://loupixdeck.com/" rel="noopener noreferrer"&gt;LoupixDeck&lt;/a&gt; is an open-source alternative to the official software, created and maintained by RadiatorTwo. Its own site sums it up neatly: "Your control deck. No vendor software." It's MIT-licensed, built in C# with .NET 10 and Avalonia, and supports the Loupedeck Live, Live S and CT, along with Razer's Stream Controller and Stream Controller X.&lt;/p&gt;

&lt;p&gt;It does what a control deck should do, without the weight:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Profiles, pages and nested folders&lt;/li&gt;
&lt;li&gt;A visual macro editor&lt;/li&gt;
&lt;li&gt;Pages that switch automatically with the active application&lt;/li&gt;
&lt;li&gt;A command-line interface and socket automation, so other tools can drive it&lt;/li&gt;
&lt;li&gt;A virtual device mode for testing without hardware&lt;/li&gt;
&lt;li&gt;A plugin system with a built-in Plugin Store and a public SDK&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;It started life on Windows and Linux. That's where we came in, because at Web-Engineer we work on Macs.&lt;/p&gt;

&lt;h2&gt;
  
  
  Bringing it to the Mac
&lt;/h2&gt;

&lt;p&gt;We opened our first pull request for macOS support on Apple Silicon in September. Porting a hardware tool to a new operating system is rarely one big change. It's lots of small ones:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Finding the device.&lt;/strong&gt; On Linux, LoupixDeck discovers devices with &lt;code&gt;udevadm&lt;/code&gt;. On macOS we walk the &lt;code&gt;ioreg&lt;/code&gt; device tree instead, which reports vendor and product IDs in a different format and in different places.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Sending keystrokes.&lt;/strong&gt; Macros need to type and click on your behalf. On macOS that goes through &lt;code&gt;CGEventPost&lt;/code&gt;, and it needs the Accessibility permission. Without that permission, macOS drops the events silently, so the app has to check the permission and ask for it clearly.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Behaving like a Mac app.&lt;/strong&gt; The menu bar should say "LoupixDeck", the Dock icon should appear only while the window is open, and the Command key should be labelled properly.
&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fip5t74sc7vnz5qk5tvnw.webp" width="800" height="800"&gt;{.img-right}
Testing on a real CT turned up more work, so we also added support for the CT's centre wheel and its screen, and fixed the CT's side strips. Then came the kind of bugs you only find by using something every day. The serial port was being polled a thousand times a second, which wasted CPU while the deck sat idle. A device could stay blank after the computer woke from sleep. And a lost touch event could leave the grid stuck.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;So far, more than 30 of our pull requests have been merged into the project. macOS support is still labelled an experimental alpha in the project's README, and that's fair. It hasn't had the same mileage as Windows and Linux yet. But it's the setup we use every day.&lt;/p&gt;

&lt;h2&gt;
  
  
  Official Mac builds from our cloud
&lt;/h2&gt;

&lt;p&gt;Source code is only half the job. Most people want to download an app and drag it into Applications, not install the .NET SDK.&lt;/p&gt;

&lt;p&gt;LoupixDeck releases often. We counted ten releases in three weeks in October. Building Mac packages by hand at that pace wasn't realistic, so we automated it. With every release, our own cloud infrastructure builds the macOS packages on Apple Silicon hardware we run ourselves, and publishes them straight to the official release on GitHub. We build both Apple Silicon and Intel versions, and as registered Apple developers we sign and notarise every one. That means macOS can check the app came from us and hasn't been tampered with, so it opens without being blocked as coming from an unidentified developer. So far our pipeline has produced 82 builds for the project. The DMGs sit next to the Windows and Linux builds on each release, and both the project's README and &lt;a href="https://loupixdeck.com/" rel="noopener noreferrer"&gt;loupixdeck.com&lt;/a&gt; credit Web-Engineer as the sponsor of the macOS builds.&lt;/p&gt;

&lt;p&gt;This is the same kind of CI and release automation we build for clients. Here it happens to be pointed at an open-source project we love rather than a client's product.&lt;/p&gt;

&lt;h2&gt;
  
  
  A batch of plugins
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fufxl7gnd412s9aeqp5oo.webp" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fufxl7gnd412s9aeqp5oo.webp" width="800" height="800"&gt;&lt;/a&gt;{.img-right}&lt;br&gt;
Once the Mac side was running, we started writing plugins for the things on and around our own desks. So far we've published eight:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Claude&lt;/strong&gt;: open Claude, quick entry, usage gauges and a live state key for Claude Code sessions&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Slack&lt;/strong&gt;: unread badges, huddles, status and Do Not Disturb across workspaces&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Visual Studio Code&lt;/strong&gt;: run commands, see problems, git and debug state on the keys&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Mac Monitor&lt;/strong&gt;: live CPU, memory, thermal, network and disk keys&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Wacom&lt;/strong&gt;: display toggle and precision mode for Wacom tablets on macOS&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Govee&lt;/strong&gt;: lights, colours, scenes and segments over the local network&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;TP-Link Kasa&lt;/strong&gt;: smart plugs and power strips, with live on/off state&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Anthem&lt;/strong&gt;: receiver control, including volume, inputs, listening modes and Zone 2&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Each one deserves its own write-up, and we'll cover several of them in future posts.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why a web agency cares about a desk controller
&lt;/h2&gt;

&lt;p&gt;It's a fair question. LoupixDeck touches a surprising amount of what we do day to day: cross-platform desktop code, low-level hardware communication, operating-system integration, plugin APIs, network-connected devices and automated build pipelines. That's a good reminder that "web agency" undersells it. A lot of our work sits behind the website: the integrations, infrastructure and tooling that make things just work.&lt;/p&gt;

&lt;p&gt;It's also been a pleasure to work with an open-source maintainer who reviews quickly and cares about quality. Huge thanks to RadiatorTwo for building LoupixDeck and for welcoming our contributions.&lt;/p&gt;

&lt;p&gt;If you've got a Loupedeck in a drawer, give &lt;a href="https://loupixdeck.com/" rel="noopener noreferrer"&gt;LoupixDeck&lt;/a&gt; a try. The source is &lt;a href="https://github.com/RadiatorTwo/LoupixDeck" rel="noopener noreferrer"&gt;on GitHub&lt;/a&gt;. And if you've got a project that needs this mix of hardware, software and automation, &lt;a href="https://www.web-engineer.co.uk/contact" rel="noopener noreferrer"&gt;get in touch&lt;/a&gt;. We'd love to hear about it.&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>macos</category>
      <category>loupixdeck</category>
      <category>cicd</category>
    </item>
    <item>
      <title>Passwordless sign-in for DontSendMeACard, powered by Web-Engine Sentry</title>
      <dc:creator>Craig Lawson</dc:creator>
      <pubDate>Fri, 30 Oct 2026 17:05:56 +0000</pubDate>
      <link>https://dev.to/web-engine/passwordless-sign-in-for-dontsendmeacard-powered-by-web-engine-sentry-4j3g</link>
      <guid>https://dev.to/web-engine/passwordless-sign-in-for-dontsendmeacard-powered-by-web-engine-sentry-4j3g</guid>
      <description>&lt;p&gt;Signing in to DontSendMeACard.com now takes one field. You type your email address, a code lands in your inbox, you enter it, and you're in. There's no password to remember and no traffic lights to click. Behind that simple screen is a change we've wanted to make for a while: using Web-Engine Sentry, our own bot-protection service, to stop spam &lt;em&gt;and&lt;/em&gt; make sign-in faster for real people.&lt;/p&gt;

&lt;h2&gt;
  
  
  A bit of background
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9xt60jir4uxy5nq7gsyx.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F9xt60jir4uxy5nq7gsyx.png" width="800" height="731"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;&lt;a href="https://www.dontsendmeacard.com/" rel="noopener noreferrer"&gt;DontSendMeACard.com&lt;/a&gt; describes itself as "the online alternative to greeting cards". People send e-cards, often designed by charities, and donate what they would have spent on paper cards and postage. We've supported the platform for years. Like any site with public sign-up forms, it attracts bots.&lt;/p&gt;

&lt;h2&gt;
  
  
  When the captcha wasn't the problem, and wasn't the answer
&lt;/h2&gt;

&lt;p&gt;A few years ago the site's sign-up form started filling with junk accounts, even though Google reCAPTCHA v2 was in place. When we audited the history recently, the numbers were stark. Between September 2023 and August 2024, with reCAPTCHA v2 on the form, sign-ups matching our spam pattern ran at roughly 200 to nearly 1,000 a month.&lt;/p&gt;

&lt;p&gt;The uncomfortable finding was that the server was checking reCAPTCHA correctly. The bots were simply arriving with tokens that passed. That's an important lesson. A captcha integration can be implemented perfectly and still fail, because the challenge itself is no longer a reliable test of whether a visitor is human.&lt;/p&gt;

&lt;p&gt;A stop-gap was added in 2024 and the spam dropped sharply. But that check was never designed to prove a visitor was a person, and charities told us it wasn't the friendliest way to register. It needed replacing with something built for the job.&lt;/p&gt;

&lt;h2&gt;
  
  
  Step one: Sentry on every front door
&lt;/h2&gt;

&lt;p&gt;We started by switching the site's three public entry points to Sentry: charity sign-up, donor account set-up and sign-in. One setting controls all three.&lt;/p&gt;

&lt;p&gt;A few design choices matter here:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Server-to-server confirmation.&lt;/strong&gt; The browser widget's result isn't trusted on its own. The site's server confirms each result directly with Sentry before it creates an account or signs anyone in. A missing, forged or reused result is refused.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Choice of challenge.&lt;/strong&gt; Visitors aren't forced into one kind of test. For example, they can solve a visual puzzle or verify with an emailed code. Email verification needs no puzzle and doesn't depend on being able to see one, which makes it a more accessible route for many people.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Resilience.&lt;/strong&gt; The widget loads from a backup CDN if the primary one is unavailable, and confirmation falls back to a second API host. We've written about &lt;a href="https://www.web-engineer.co.uk/blog/failover-beyond-dns" rel="noopener noreferrer"&gt;how we built failover into Sentry&lt;/a&gt; before, so we won't repeat it here.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Before release we tested the obvious attacks locally: posting with no result, a forged result, a result replayed a second time and a result that claims to come from the wrong host. Each one was refused.&lt;/p&gt;

&lt;h2&gt;
  
  
  Step two: turning the human check into the sign-in
&lt;/h2&gt;

&lt;p&gt;Once Sentry was verifying email addresses on the sign-in page, a question followed naturally. If we've already proved someone controls an inbox, why also ask for a password?&lt;/p&gt;

&lt;p&gt;So the next day we flipped the sign-in page around:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Enter your email.&lt;/strong&gt; That's the only field.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;Sentry emails you a one-time code.&lt;/strong&gt;&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Enter the code.&lt;/strong&gt; The form submits as soon as Sentry accepts it. There's no extra "Sign in" button to hunt for.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;You're signed in to the account whose address Sentry verified. The address typed into the page is never trusted on its own. If someone verifies one address and then tries to submit another, the sign-in is refused. Other safeguards follow:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;Codes are &lt;strong&gt;single-use&lt;/strong&gt; and &lt;strong&gt;time-limited&lt;/strong&gt;. A code is accepted for 15 minutes.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Every sign-in starts a fresh session&lt;/strong&gt;, whichever route you used.&lt;/li&gt;
&lt;li&gt;Matching is &lt;strong&gt;exact&lt;/strong&gt;, so addresses that only look similar can't be swapped in.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Passwords haven't disappeared. A small &lt;strong&gt;"Try another way"&lt;/strong&gt; link under the form leads to the familiar email-and-password sign-in and password reset, for anyone who prefers them. The new sign-in also sits behind a configuration switch, so it can be turned off without a deploy if it ever needs to be.&lt;/p&gt;

&lt;p&gt;Here's the shape of the server-side check, as &lt;strong&gt;illustrative pseudo-code only&lt;/strong&gt;. It isn't Sentry's API:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight php"&gt;&lt;code&gt;&lt;span class="c1"&gt;// Illustrative only: sign in as the address that was verified,&lt;/span&gt;
&lt;span class="c1"&gt;// never the address the form claims.&lt;/span&gt;
&lt;span class="nv"&gt;$receipt&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;confirmWithVerificationService&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nv"&gt;$request&lt;/span&gt;&lt;span class="o"&gt;-&amp;gt;&lt;/span&gt;&lt;span class="n"&gt;token&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt; &lt;span class="c1"&gt;// server to server&lt;/span&gt;
&lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="o"&gt;!&lt;/span&gt;&lt;span class="nv"&gt;$receipt&lt;/span&gt; &lt;span class="o"&gt;||&lt;/span&gt; &lt;span class="nv"&gt;$receipt&lt;/span&gt;&lt;span class="o"&gt;-&amp;gt;&lt;/span&gt;&lt;span class="n"&gt;used&lt;/span&gt; &lt;span class="o"&gt;||&lt;/span&gt; &lt;span class="nv"&gt;$receipt&lt;/span&gt;&lt;span class="o"&gt;-&amp;gt;&lt;/span&gt;&lt;span class="nf"&gt;isExpired&lt;/span&gt;&lt;span class="p"&gt;())&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nf"&gt;refuse&lt;/span&gt;&lt;span class="p"&gt;();&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="nv"&gt;$user&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nf"&gt;findOpenAccountByExactEmail&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nv"&gt;$receipt&lt;/span&gt;&lt;span class="o"&gt;-&amp;gt;&lt;/span&gt;&lt;span class="n"&gt;verifiedEmail&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
&lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="o"&gt;!&lt;/span&gt;&lt;span class="nv"&gt;$user&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="nf"&gt;offerRegistrationOrTryAnotherWay&lt;/span&gt;&lt;span class="p"&gt;();&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;span class="nf"&gt;startFreshSession&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nv"&gt;$user&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;The key line is the lookup. The account comes from what was &lt;em&gt;verified&lt;/em&gt;, not from what was &lt;em&gt;typed&lt;/em&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Codes that look like they belong
&lt;/h2&gt;

&lt;p&gt;A one-time code is only reassuring if people trust the email it arrives in. A message from an unfamiliar sender with someone else's branding looks a lot like phishing.&lt;/p&gt;

&lt;p&gt;DontSendMeACard is the first site to use a new Sentry option for its own mail settings and branding. The code emails now arrive under the DontSendMeACard name, with its logo and its own wording, and that can be updated from our cloud platform in moments, without touching the site's code. The DSMAC team are already refining the copy, including a clear reminder that staff will never ask you for your code.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it adds up to
&lt;/h2&gt;

&lt;p&gt;For visitors, sign-in is now one field and one code. For the site, every way in passes through a check that's confirmed on the server rather than trusted from the browser. And because email verification needs no puzzle, that check is easier for more people to complete.&lt;/p&gt;

&lt;p&gt;We'll be honest about one thing: it's early. The change has only just gone live, and we'll review the sign-up spam figures over the coming weeks. Moving from reCAPTCHA to Sentry was a deliberate choice, made because the data showed reCAPTCHA wasn't stopping these bots. We'll share how it holds up.&lt;/p&gt;

&lt;h2&gt;
  
  
  Want the same for your site?
&lt;/h2&gt;

&lt;p&gt;Web-Engine Sentry is WordPress-first, but it ships with an API and examples, which is exactly how we integrated it into DontSendMeACard's custom PHP platform. If spam is getting through your forms, or your sign-in is more effort than it needs to be, &lt;a href="https://www.web-engineer.co.uk/contact" rel="noopener noreferrer"&gt;get in touch&lt;/a&gt;. We're happy to talk through what would work for you.&lt;/p&gt;

</description>
      <category>spamprotection</category>
      <category>webenginesentry</category>
      <category>casestudy</category>
      <category>passwordless</category>
    </item>
    <item>
      <title>Failover beyond DNS</title>
      <dc:creator>Craig Lawson</dc:creator>
      <pubDate>Thu, 15 Oct 2026 15:11:32 +0000</pubDate>
      <link>https://dev.to/web-engine/failover-beyond-dns-1e36</link>
      <guid>https://dev.to/web-engine/failover-beyond-dns-1e36</guid>
      <description>&lt;p&gt;So you've built a service and it's running in the cloud. It sits in a Kubernetes cluster, it's designed for high availability, and every component has a replica. Then the data centre loses its connection to the internet, and none of that helps.&lt;/p&gt;

&lt;p&gt;It happens more often than we'd like to admit. Data centres with multiple independent points of presence still have outages. On 20 October 2025, a DNS fault inside AWS's US-EAST-1 region cascaded into hours of disruption for a long list of services that had nothing to do with each other, apart from where they were hosted. Cloudflare has had more than one bad day too, including an outage of its 1.1.1.1 DNS resolver in July 2025 and a network-wide failure that November. When a large share of the web sits on a handful of providers, one provider's bad afternoon becomes everyone's.&lt;/p&gt;

&lt;p&gt;The uncomfortable lesson is that your hardware can be perfectly healthy and you can still be offline. Spread yourself across data centres with failover between them, and a DNS problem can still undo the whole arrangement.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why this mattered to us
&lt;/h2&gt;

&lt;p&gt;&lt;a href="https://www.web-engineer.co.uk/work/we-sentry" rel="noopener noreferrer"&gt;Web-Engine Sentry&lt;/a&gt; (Sentry for short) is our site-protection plugin. It cuts spam by making a site hard work for bots while keeping it easy for humans. It's built for &lt;a href="https://wordpress.org/plugins/web-engine-sentry/" rel="noopener noreferrer"&gt;WordPress&lt;/a&gt; first, and it ships with an &lt;a href="https://www.web-engineer.co.uk/doc/sentry-api" rel="noopener noreferrer"&gt;API&lt;/a&gt; and examples so it can protect any site. Sentry gives sites several ways to check that a visitor is real: captcha, email verification and mobile verification. As soon as a site puts one of those checks in front of its contact form or login page, our uptime becomes their uptime. If we're unreachable, their visitors can't get in touch or sign in.&lt;/p&gt;

&lt;p&gt;So "it's in the cloud" wasn't a good enough answer. We needed a plan that keeps the service running wherever possible, whatever fails underneath it.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option one: throw hardware at it
&lt;/h2&gt;

&lt;p&gt;You can solve most availability problems with hardware, if you have the budget. Large cloud providers do exactly that, and they price it in. For most teams, though, a purely hardware-based answer is expensive, and you're still bound by the physical limits of someone else's estate.&lt;/p&gt;

&lt;p&gt;The conventional approach would be to host in two geographic regions and use DNS failover to switch between them. It works, but it isn't instant. Once the primary goes down, health checks have to notice, records have to change, and resolvers around the world have to let go of cached answers. Short TTLs help, but not every resolver honours them. That leaves a window, often a few minutes, where requests fail. For a verification step sitting in front of someone's checkout, a few minutes is a long time.&lt;/p&gt;

&lt;h2&gt;
  
  
  Option two: make the software smarter
&lt;/h2&gt;

&lt;p&gt;The alternative is to let the client decide. Give the consumer of the service an ordered list of endpoints, add some basic error handling, and when a request fails, move to the next one straight away.&lt;/p&gt;

&lt;p&gt;The real advantage is that those endpoints can be completely independent: different hosting, different networks, even different DNS providers. Failover no longer waits for a record to propagate. It happens on the next request, in milliseconds. Our weak point becomes our own code and the service itself, rather than the rigid physical constraints of our suppliers, and our own code is something we can test, improve and ship.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;DNS failover vs client-side fallback&lt;/strong&gt;&lt;br&gt;
&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Faac7q1a9nn9f6si2r75c.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Faac7q1a9nn9f6si2r75c.png" width="800" height="294"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;The idea in its simplest form:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="c1"&gt;// Simplified illustration, not the Sentry SDK itself&lt;/span&gt;
&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;ENDPOINTS&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;
  &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;https://a.example-primary.net&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;https://b.example-secondary.org&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;https://c.example-tertiary.io&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
&lt;span class="p"&gt;];&lt;/span&gt;

&lt;span class="k"&gt;async&lt;/span&gt; &lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;callService&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;path&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="nx"&gt;options&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;{})&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;for &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;base&lt;/span&gt; &lt;span class="k"&gt;of&lt;/span&gt; &lt;span class="nx"&gt;ENDPOINTS&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="k"&gt;try&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
      &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="nx"&gt;res&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="nf"&gt;fetch&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;base&lt;/span&gt; &lt;span class="o"&gt;+&lt;/span&gt; &lt;span class="nx"&gt;path&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
        &lt;span class="p"&gt;...&lt;/span&gt;&lt;span class="nx"&gt;options&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
        &lt;span class="na"&gt;signal&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nx"&gt;AbortSignal&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;timeout&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="mi"&gt;3000&lt;/span&gt;&lt;span class="p"&gt;),&lt;/span&gt;
      &lt;span class="p"&gt;});&lt;/span&gt;
      &lt;span class="k"&gt;if &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;res&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nx"&gt;ok&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="nx"&gt;res&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
    &lt;span class="p"&gt;}&lt;/span&gt; &lt;span class="k"&gt;catch &lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="nx"&gt;e&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
      &lt;span class="c1"&gt;// network error or timeout: try the next endpoint&lt;/span&gt;
    &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="k"&gt;throw&lt;/span&gt; &lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nc"&gt;Error&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="s1"&gt;All endpoints unavailable&lt;/span&gt;&lt;span class="dl"&gt;'&lt;/span&gt;&lt;span class="p"&gt;);&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  Decentralise by design
&lt;/h2&gt;

&lt;p&gt;Real resilience comes from breaking the application into its key parts and giving each one its own availability strategy. Once you treat each component as a separate service, you can tune how each one fails over. Here's how that played out for Sentry.&lt;/p&gt;

&lt;h3&gt;
  
  
  DNS
&lt;/h3&gt;

&lt;p&gt;We spread our endpoints across several domains, each managed by a different DNS provider, with multiple endpoints behind each domain. If one provider has a bad day, the others are unaffected. Relatively short TTLs mean we can repoint routes quickly when we need to.&lt;/p&gt;

&lt;h3&gt;
  
  
  Frontend
&lt;/h3&gt;

&lt;p&gt;The browser script loads from whichever CDN endpoint is available. The frontend knows about the alternatives, so if a script fails to load it moves on to the next one. That gives us geographically separate fallbacks managed entirely on the visitor's side. Once it finds an endpoint that works, it sticks with it for the rest of the session, so visitors don't pay the cost of a failed attempt twice.&lt;/p&gt;

&lt;h3&gt;
  
  
  Backend
&lt;/h3&gt;

&lt;p&gt;On the site's own server, the Sentry SDK acts as a gateway to the available service endpoints. That pushed us to treat each backend as a separate node from the start and to design a proper synchronisation strategy between them. We had to decide early what the service provides and how to deliver it in a decentralised way, rather than bolting resilience on later.&lt;/p&gt;

&lt;h3&gt;
  
  
  Data
&lt;/h3&gt;

&lt;p&gt;Each node also runs its own separate database, holding a specific set of tables duplicated across several sites. Our APIs keep these in sync with the master. Some of that data is read-only reference data; the rest lets a node keep tracking activity on its own. That's enough for the service to keep running even when the primary site is unavailable. When the master comes back, each node syncs what it recorded back to it.&lt;/p&gt;

&lt;p&gt;Behind everything sits a highly available database cluster, and the nodes treat it as the ultimate source of truth. On paper, that cluster should never go down. We deliberately don't assume it won't. Making no promises about its availability gives us another layer of fault tolerance, and forced us to plan up front how data is reconciled once everything is back online.&lt;/p&gt;

&lt;h2&gt;
  
  
  The result
&lt;/h2&gt;

&lt;p&gt;None of this is exotic. It's a fallback list, some careful error handling, and the discipline to assume every layer can fail. Together those give a site running Sentry verification that keeps working through regional outages, DNS failures and the occasional very bad day in a data centre.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F4aukbr6dh7f0m91p8f1q.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F4aukbr6dh7f0m91p8f1q.png" width="800" height="334"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  Try it, and tell us what you think
&lt;/h2&gt;

&lt;p&gt;You can install Sentry from the &lt;a href="https://wordpress.org/plugins/web-engine-sentry/" rel="noopener noreferrer"&gt;WordPress plugin directory&lt;/a&gt;, or read more about &lt;a href="https://www.web-engineer.co.uk/work/we-sentry" rel="noopener noreferrer"&gt;Web-Engine Sentry&lt;/a&gt; and the &lt;a href="https://www.web-engineer.co.uk/doc/sentry-api" rel="noopener noreferrer"&gt;Sentry API&lt;/a&gt; to use it on any site.&lt;/p&gt;

&lt;p&gt;How do you handle failover in your own services? Have you been caught out by DNS? Join the discussion and let us know.&lt;/p&gt;

</description>
      <category>sentry</category>
      <category>highavailability</category>
      <category>failover</category>
      <category>dns</category>
    </item>
  </channel>
</rss>
