Changelog - TagPulse

v0.6.235 September 2026 Latest

New Clock Battery Check — A small coin battery (usually CR2032) on the motherboard keeps your PC's clock running while it is unplugged. When it runs out, the clock is wrong after every startup — causing browser security warnings, failed sign-ins and wrong file times that rarely get blamed on the real cause. TagPulse now recognises this pattern and tells you the likely fix: a battery that costs a couple of euros and takes a minute to replace. It waits until the problem repeats on different days, ignores one-offs like travel or a manual clock change, leaves dual-boot PCs and virtual machines alone, and reports it as information — never an alarm
Improved Time Events in Plain Language — Windows time and clock events in the Event Analyzer are now explained simply: what each one means, whether it matters, and when the clock battery is worth a look
Improved Fuller Motherboard Details — The motherboard panel now reads more of what your board and its firmware publish about themselves, including the SMBIOS specification level reported by the BIOS. How much appears varies from board to board, since each manufacturer decides how much to publish — but wherever the detail exists, TagPulse now shows it. It is the kind of thing you only want on the day you need it: checking whether a board supports a firmware feature, matching a machine against a manufacturer's advisory, or answering a question about a PC without opening the case
Fixed Application Details in Software Impact — Opening an application from the products list could show an empty page instead of its details. It now opens reliably, with the application's size shown in sensible units

v0.6.228 August 2026

New TagPulse Now Speaks Greek — The whole application is available in Greek as well as English: every screen, every alert, and the reasoning behind each one, switched whenever you like from the selector at the top of the dashboard. It is a large surface to translate, so if a corner of it still reads oddly to you, tell us and it will be right in the next release. And if you would like TagPulse in your own language, write to [email protected] — you do not need to be a professional translator, we provide the wording, the tools and the review, and everyone who helps is credited in the application, gets a licence on us and early access to every new version
New Show Hidden Alerts — A button at the top of the alerts panel brings back anything you have cleared away, so a finding you set aside in a busy moment is still there when you want a second look — or when you want to show a colleague what the machine reported that morning. It counts what is waiting behind it, and tells you on hover how long hidden findings are kept before they are removed automatically
Improved A Dashboard That Runs Light — The live view has been rebuilt around a single steady heartbeat: every chart, gauge and indicator now moves together on one shared rhythm, redrawn when your machine actually has something new to report. The motion is smoother and more even than before, and the whole page asks a small fraction of the processor it used to — most strikingly on high-refresh gaming displays, where a live monitoring window is at its most demanding. A tool that watches your PC should be the quietest thing running on it, so leave the dashboard open all day, plugged in or on battery, and it will simply keep out of your way
Improved A Full Week of Process History — The History page now keeps seven full days of detail for the machine in front of you — not only the shape of the graph, but which applications were behind every point on it. When someone tells you their PC "went odd on Tuesday afternoon", you can go straight back to Tuesday afternoon, land on the exact moment, and read what was genuinely using the machine while it happened. The hourly picture carries on out to ninety days beyond that week, and every chart now states plainly, underneath its legend, the exact period it is showing
Improved Fifteen Days of Alert History — Every finding TagPulse raises now stays on your dashboard for a fortnight rather than a week. A machine left running through a holiday, or a site you only get to every other week, still shows what it saw while nobody was there — including conditions that appeared and then settled by themselves. Those are marked as recovered, so at a glance you can tell what still wants your attention apart from what sorted itself out while you were away
Improved Checks That Wait for Real Work — Before TagPulse judges a fan or a drive to be running slow, it now confirms the machine was genuinely working at the time. A quiet machine spins its fans down and lets its drive rest — that is the hardware doing its job, not failing at it — so TagPulse holds its verdict until the machine is actually being asked for something

v0.6.217 August 2026

Improved The Full Windows Update Picture — Every machine on your fleet dashboard now reports its patch position in full: which updates are waiting, how many are already downloaded and ready to go in, when Windows last looked, when it last installed something, and whether a restart is still owed before that work counts as done. Windows 10, Windows 11 and Windows Server are all read the same way, so a mixed fleet of desktops and servers reads as one list. And when a machine reports nothing pending, TagPulse confirms that with Windows directly before presenting it as up to date — the same check a technician would make before signing a machine off
Improved Alerts That Respect Your Attention — An ongoing condition — a disk slowly filling up, memory that is always tight — is now told the way a good technician would tell it: once when it appears, one reminder the next day, and then TagPulse keeps watch quietly while the finding stays visible on your dashboard. It speaks up again the moment things genuinely get worse, or when a problem truly goes away and later returns. This memory survives restarts too, so a PC that is switched off every evening hears each finding just as calmly as one that never sleeps. Urgent hardware emergencies are, as always, reported the instant they happen

v0.6.204 August 2026

New A Welcome Note on Day One — A fresh installation now opens with a short message explaining what to expect: TagPulse is still learning what is normal for that particular PC, and the findings that work by comparison start appearing once it has gathered a couple of days of running time. Urgent hardware problems are never held back — overheating, a failing drive, memory errors, a stopped fan or a full disk are reported from the very first minute
New Digital Signature Badges — Every application and running process on the Software Impact page now carries a badge showing whether its file is digitally signed, and by whom. Publisher-verified software is marked at a glance, so anything unsigned stands out immediately — the kind of check that normally means right-clicking a file and digging through its properties, now visible on every entry at once
New Complete Software Picture — Software Impact now lists every installed application it can see, not only the ones busy at that moment. Quiet background software, Microsoft Store apps and portable programs all appear with their real publisher and product name, so nothing is using your machine unnoticed
New Resource History per Application — Each entry now opens to show its CPU and memory over the last 24 hours, with the low, average and high for the period. A single reading tells you what an application is doing right now; the trend tells you whether that is normal for it
New Installation Timeline — Installations are now grouped into the sessions they actually happened in, so you can see what arrived together — an application and its bundled components in one place, rather than a flat list of names and dates
New Process Family Detail — Expanding an application now shows which processes belong to it, which one is the main program, and what launched each of them. When something appears in your process list and you want to know where it came from, the answer is now on the page
Improved The True Cost of Multi-Process Applications — Modern browsers and chat apps run dozens of separate processes, and each one on its own looks small. TagPulse now combines every process an application is running into a single honest figure and shows how many there are, so a browser spread across 40 processes reports what it is really using in total
Improved Clearer Application Identity — Applications now show the name and publisher taken from the program's own file details, so entries read the way you would recognise them. Non-Latin and accented publisher names — Greek, French, and others — display correctly throughout
Improved Consistent Impact Scoring — The impact figure shown on each application, the category it is filed under, and the totals at the top of the page are now all derived from one single measurement, so every number on the screen agrees with the others
Improved Better Update Handling — Updates now arrive and install more smoothly and reliably on every machine

v0.6.195 July 2026

New Virtual-Memory Exhaustion Alert — TagPulse now spots when a machine's virtual memory is exhausted and names the likely cause — a page file that's too small or disabled, or one program eating far too much — with clear advice on the fix. Random crashes that used to look like a mystery now come with an explanation
New Protection for Memory-Starved PCs — When a computer has completely run out of virtual memory (a condition that makes any program crash at random), TagPulse now starts in a light mode instead of adding to the pressure — live monitoring stays on, the heavier checks pause, and everything resumes by itself the moment memory recovers. A calm note on the dashboard tells you what's happening and why
New VRM Temperature Watch — On motherboards that expose VRM (power-delivery) temperature sensors, TagPulse now keeps an eye on them against a conservative danger threshold — a quiet guardian for the components that feed power to your CPU
New Learning From Similar Hardware — Your machine's GPU temperatures, fan performance and disk responsiveness are now compared with the fleet-wide normal for the same hardware class. What's typical for your exact hardware stays quiet — and a genuine outlier stands out early
Improved Voltage Monitoring, Rebuilt Around Your Power Supply — Voltage alerts now measure your power supply's main rails (+12V, +5V, +3.3V) directly against their official specifications. CPU and GPU rails — which change voltage thousands of times a second by design — are tracked for insight but never alarmed. When a voltage warning appears now, it points at a real power problem, sustained and verified
Improved Finer Drive-Health Grading — SMART drive alerts now scale with the evidence: a couple of remapped sectors arrive as an early advisory, while serious counts or the drive's own failure prediction remain immediately critical. Machines holding an older critical are re-assessed automatically under the improved grading
Improved Liquid-Cooling Pump Intelligence — A pump reading 0 RPM is now judged by what actually matters: the temperature of the component it cools. Pumps parked by their fan curve stay quiet, while a genuinely failed pump is caught within minutes as heat builds
Improved Smarter Disk-Speed Monitoring — The "drive running slower than usual" check now understands busy moments: it measures against a fair baseline even on very fast drives, speaks up only when a slowdown persists through 15+ minutes of real activity — a backup, a game loading, or a Windows update won't trigger it — and mentions it at most once a day
Improved Smarter Fan Monitoring — Fan alerts are now more deliberate: a fan is watched once it has proven it really spins, momentary sensor blips are ignored, and chipset fans that park by design are treated as normal — while a genuinely stopped fan on a warm system still raises an immediate emergency
Improved Steadier Alerts While Gaming — Activity detection now holds the gaming state through the natural ebb and flow of a session, so performance advisories no longer interrupt mid-game over the normal rhythm of loading screens, menus and quiet moments
Improved Hardware Events, Verified First — A Windows "fatal hardware error" record is now cross-checked against whether the system actually crashed. If your PC sailed on unharmed, you see a calm recovery note instead of an emergency
Improved Considerate Database Maintenance — Heavy database cleanup now checks that the system has memory to spare before running, and quietly waits for a better moment on a struggling machine
Improved Sturdier Through Sleep and Restarts — Waking from sleep is recognized for what it is, alert memory now survives restarts, and the monitoring engine degrades gracefully instead of stopping when a sensor source misbehaves
Fixed Steadfast Cloud Connectivity — Connection handling now side-steps IPv6 and DNS hiccups automatically, and a slow or unreachable network can never stall the monitoring service itself
Fixed Smoother Updates on Busy PCs — The self-update process now waits patiently for the right moment on a busy system and always cleans up after itself

v0.6.186 July 2026

New Hybrid-Laptop Graphics — On laptops with both a built-in graphics chip and a separate NVIDIA or AMD card, TagPulse now names each one correctly and reports the card that's actually doing the work — so your GPU usage and temperature are accurate instead of the idle card sitting at 0%
New Stopped-Fan Detection — A fan that stops spinning is now caught the moment it happens, with smart checks so silent semi-passive builds don't trigger false alarms
Improved Steadier Sensors on Laptops — An updated sensor engine with broader hardware compatibility, plus graceful handling on laptops where the firmware or another monitoring app holds the sensors — your core temperatures keep reporting instead of dropping out
Improved Smarter Voltage Monitoring — Every power rail now learns its own normal range for idle and load separately, so healthy voltages are never flagged while a genuine problem still stands out
Improved True Severity, Always — Drive-cable errors and SSD-wear notices now arrive at their real severity instead of instant emergency, and vendor-specific SSD counters can no longer trigger a false "SSD end of life"
Improved Steadier Battery Health Readings — Battery health is now judged on the 90-day trend, so a single odd reading after waking from sleep no longer colors the picture — only a confirmed decline gets reported
Improved International PCs, Fully Covered — Greek and other non-Latin file paths and device names are now fully supported in security checks and sensor monitoring
Improved Calmer Event Notices — Windows Event Log findings (hardware events, file system, USB, driver issues) now arrive as informational notices — real measured problems still alert at full severity
Improved Self-Healing Service — Crash-loop protection, a safe degraded mode, and a watchdog that revives a stuck component instead of letting monitoring silently stop
Improved Leaner Engine — The analysis engine was restructured into focused modules and stripped of redundant background work — lower CPU use and less database activity, same intelligence
Improved Honest Numbers Everywhere — The AI status page now reports true learning progress and false-positive rates, and new machines no longer appear under "unknown" hardware groups on the fleet dashboard
Fixed True Restart Reporting — Normal Windows shutdowns and restarts no longer show up as service crashes on the fleet dashboard
Fixed Alerts That Stay Until Fixed — A critical alert (like a failing drive) no longer vanishes after a reboot — it stays until the problem is resolved, and clears itself the moment the hardware genuinely recovers
Fixed No More Ghost Alerts — Old Windows event history can no longer resurface as fresh alerts after a restart, and repeated events are counted honestly instead of doubled
Fixed No Freezes on Slow Internet — A slow or unreachable cloud connection can no longer stall alert processing on the PC
Fixed Rare Stability Issues — Fixed rare crashes when the dashboard refreshed during heavy analysis, and a silent exit that could occur at service shutdown

v0.6.178 June 2026

New Hardware Fault Detection — Catches faulty RAM, CPU or bus errors the instant Windows logs them — the hidden cause of random crashes and freezes
New Slower-Boot Early Warning — Notices when your PC starts taking longer to boot, before it gets frustrating
New Calm From the Start — A new machine stays quiet until the engine has learned what's normal — no false alarms on day one. Real hardware faults still alert instantly
Improved A Sharper Self-Learning Engine — Much better at telling a real problem from everyday activity — noticeably fewer false alarms, proven across our live fleet
Improved Smarter Bottleneck Alerts — Flags a slowdown only when it's genuinely sustained — a passing Windows Update or scan won't trip a critical alarm
Improved Clearer Memory Alerts — Names the exact program hogging your memory
Improved Fair Software Blame — A suspect app is judged on its own footprint, so a gaming session can't get an app update blamed
Improved Smoother History — The history scroll bar follows your drag and survives switching time ranges
Fixed Rock-Solid Stability — Fixed a rare case where monitoring could silently stop on slow-to-query machines
Fixed True Alert Severity — Warnings no longer show up as critical on the dashboard
Fixed Accurate History Charts — Graphics-driver glitches can no longer poison your performance charts
Fixed Trustworthy Voltage Alerts — No more false "out of spec" voltage warnings on healthy PCs

v0.6.166 June 2026

New Fleet-Smart Alerts — Every performance alert is now sanity-checked against thousands of machines with similar hardware doing the same activity. A gaming PC running its GPU at full power is recognized as perfectly normal — you only hear about things that are genuinely unusual for a machine like yours
New Thermal Trend Watch — TagPulse now tracks how warm your machine runs compared to its own normal at the same workload, laying the groundwork to catch dust build-up, tired fans or aging thermal paste long before they become a real problem
Improved Clearer Disk Health Alerts — Early warnings ("keep an eye on this drive") are now clearly separated from genuine failures, alerts name the exact drive model, and the same issue won't notify you again every day
Improved Smarter Self-Learning Engine — The on-device AI now learns from richer system signals (memory pressure, system activity, GPU temperature) and trains more reliably in the background
Improved Richer Machine Details — System information and installation details now appear consistently for every machine in the fleet dashboard
Improved Calmer Notifications — Duplicate alerts are filtered in the cloud, and hardware alerts no longer point at an unrelated busy app as the likely cause
Fixed Network Monitoring on Multi-Adapter PCs — Machines with several network connections (Wi-Fi + Ethernet, VPN adapters) now report their true combined network activity

v0.6.162 June 2026

New Failing Disk Alerts — When a drive shows real signs of trouble (bad sectors or the drive's own failure prediction), TagPulse now raises a clear alert on the PC and on the fleet dashboard — not just on the Disk page
Improved Game-Friendly Security Checks — Miner detection now recognizes digitally signed games and apps on any drive, so demanding games are never mistaken for something suspicious
Improved Calmer App Impact Alerts — An app is only flagged when the system is genuinely under heavy load and that app is the main cause
Improved More Reliable Alert Times — Alert timestamps are now set by the cloud, so a PC with a wrong clock can't produce confusing dates
Fixed Smarter Disk Speed Alerts — Disk speed warnings are now based on a real before-and-after speed test of your drive, so you only see one when the drive itself has genuinely slowed down — not during normal light use
Fixed Speed Test Tidy-Up — The disk speed test now removes its temporary TagPulse folder from every drive and USB stick when it finishes

v0.6.161 May 2026

New Restart & Shutdown Detection — New method to accurately recognize when a machine restarts or shuts down, so online status and uptime stay correct
Improved Faster Fleet Dashboard — A faster, smoother fleet management UI/UX experience
Improved Intel Core i9-14900KF — Better handling of brief single-core temperature spikes to prevent false thermal readings

v0.6.158 May 2026

New TagIO Remote Desktop Support — Detect TagIO installs on fleet machines and one-click connect
New USB Storage Lock/Unlock — Block or re-enable USB mass-storage devices (thumb drives, external HDDs) fleet-wide from the dashboard — keyboards, mice and other USB peripherals stay unaffected
New Friendlier Fleet Management UI — Refreshed UI/UX across fleet views for a cleaner, faster workflow
New Delete Machines from Fleet — Remove decommissioned machines directly from the dashboard
Fixed Minor UI display issues across the application

v0.6 January 2026

New Pulsie — Your PC health companion mascot with tips and insights
New Event Analyzer — Browse and analyze Windows Event Viewer logs without the complexity
New Crypto Miner Detection — Detect unauthorized miners with trusted-process whitelist and one-click kill
New Fleet Remote Management — Trigger Windows updates, winget installs, and manage startup remotely
New Software Impact Analysis — Correlate app activity with resource usage and temperature spikes
New Startup Apps Manager — View and toggle Windows startup items from TagPulse
Improved History Dashboard — Local + cloud data with synchronized temperature graphs
Improved Fleet Dashboard Speedtest — Built-in network speed tests across all machines
Improved Country & VM Detection — Better geo-location and VMware/Hyper-V detection for CHE calibration
Fixed SQLite database locking issues causing slowdowns
Fixed Crashes on ASUS computers with specific motherboard configurations
Fixed Fan degradation detection false positives
Fixed Event Viewer upload flooding on certain systems
Fixed Cloud sync error handling (4xx/5xx) and slow-internet timeouts
Fixed Windows special characters in usernames causing sync failures

v0.5 2025

New CHE (Computer Help Engine) — AI engine that learns your PC's behavior patterns
New Fleet Dashboard — Monitor and manage multiple PCs from a single web dashboard
New Cloud Sync — Hardware data synced to the cloud for remote access and history
New Smart Alerts — AI-driven notifications for temperature, disk health, and anomalies
Improved Complete UI redesign with modern React dashboard
Improved Lower CPU and memory usage across the board

v0.4 2024

New S.M.A.R.T. Disk Monitoring — Read SSD/HDD health data and predict failures
New Process Monitor — Real-time view of CPU, memory, and GPU usage per process
New Auto-Update System — Silent background updates with rollback support
Improved Sensor polling accuracy and reduced overhead
Fixed Various stability fixes and crash recovery improvements