Betas can update themselves now, and devices behind older docks come back
You can get the next beta from inside the app. Settings has a new Updates
section with an "Include beta versions" switch. Turn it on and pre-releases
arrive through the normal update prompt instead of you going to the releases
page. You still had to fetch this one by hand, sorry, that is the last time.
Two things it deliberately does not do. Turning the switch back off stops
future betas but does not move you off a beta you are already running, the
next stable release does that. And if you installed WhatCable with Homebrew
you will see a warning, because brew upgrade can put the stable back over a
beta without asking.
Devices plugged into a TB3-era dock or display are attributed to a port
again. LG UltraFine 5K, CalDigit TS3+ and anything else of that generation
carries its own PCIe USB controller instead of tunnelling USB the way a modern
dock does. WhatCable classified those devices as tunnelled with no port, so as
soon as you had two Thunderbolt things connected, everything behind the dock
quietly vanished from the port card. Found because someone emailed me about it
and then ran the test kit so I could see their setup. If you own a dock of that
vintage, this is the build to try it on.
A charger-only cable no longer flashes "10 Gbps" as you plug it in. During
the first moment of a connection the port controller briefly reports a
SuperSpeed link that then turns out not to exist. WhatCable printed every
instant it was handed, so a speed appeared and vanished. It now waits for
something to corroborate the claim: a real SuperSpeed device that actually
reached the port. Settled readings are unchanged. This is the one The Verge
caught (issue #181).
The popup closes when you tell it to. A regression I introduced in beta.2:
the re-anchoring added in beta.2 could race a click-to-close and pop the window
straight back open. Worth a hands-on check if you run with the watts readout
on, since that was what triggered it.
Disconnect notifications say what left. They used to read "USB device
disconnected" with no name attached, which is not much of a notification.
Thanks to Rob for reporting that one.
Numbers follow your region. If you write 81,8% rather than 81.8%, the app
now agrees with you, on screen, in the CLI and in the widget. Exports and
--json deliberately stay dot-formatted so scripts do not break. Also Rob.
Pro: the charging resistance estimate is actually visible now. It shipped
in beta.2 rendering only on the live card, and a laptop's charging port is
normally the contracted one, so it appeared nowhere. It also says what it is
doing: measuring, waiting for the charging load to vary, or a result. Still
labelled experimental, still no good/bad verdict, because that needs bench
calibration I have not done yet.
The terms allow sharing what WhatCable tells you. Section 11 read as a ban
on showing output to a supplier or a customer. It never meant that. You can
share a reading as long as you describe it accurately.
Italian has been through a thorough pass from @bovirus, covering the charging
diagnostics, the power negotiation lines, the display and cable wording, and
the Pro screens. Traditional Chinese has another round of refinement from
@jimmyorz. The test-kit probes now flag a dump that got cut short by a cable
moving mid-read, so a truncated capture stops looking like a machine with
fewer devices.
The dock attribution and the popup close are the two to lean on. If a device
shows up under the wrong port, or the popup misbehaves again, tell me what was
plugged in at the time.