Support

Something is not behaving. Start here.

This page takes a few minutes and needs nothing but you — nothing to install, nothing to prepare, nothing to look up first. It ends with something to do, or with a plain statement that the problem is ours and how to tell us about it.

Start with what the product said

If the product said something, its exact wording is the most useful thing you have. Two different problems can produce messages that read almost the same, and the wording is often the only thing that tells them apart. Copy it if you still have it, and work from what you remember if you do not — that is a normal place to start, not a worse one.

If there was no message and something is simply absent, that is answerable too.

Two more things carry most of the answers below, so keep them in mind as you read: what you were doing immediately before, and what you expected to happen instead.

Narrow it

These are the questions the message did not ask. Any one of them can be most of the answer.

• Is the rest of the product still updating, or has everything stopped?
• Does it happen every time, or did it happen once?
• Did it start after you changed something — a setting, a browser, a device, a network?
• Were you doing something unusual, or the thing you always do?

One message can cover more than one situation, and that is usually deliberate. Where two situations would have you do different things, the product is meant to say which one you are in. Where they would have you do the same thing, it says one thing rather than naming a cause it cannot see from where it is standing, so a line that covers several causes is often the finished answer. Where it leaves you genuinely unsure which of two things to do, that is a gap in the product, and worth telling us about.

It will be one of four

Saying which one you are in is most of the answer, because it tells you what kind of result is available.

Something you can change, or another program can. A permission not granted, a device not selected, a browser that does not offer the capability, a network not letting the product reach what it needs. Sometimes it sits in another piece of software altogether — a driver installer, rig-control software, your operating system. These have a next action, and it is one you can take. Where it belongs to another project, they are the ones who can fix it, and their documentation will be current in a way ours could not be; the notes you have already made answer their first questions too.

A source that is not answering, or is old. These products depend on outside sources and say which. When one is unavailable the product will usually say so. What is worth knowing is what that source covers, so you can tell whether what is missing mattered for what you were doing — and that is in your product’s guide, below.

A limit that is real and documented. The guide states it, and nothing is broken. Recognizing one is worth as much as a fix, because a limit that reads like a fault keeps you looking for a cause that is not there.

Ours. Some problems have no action on your side at all. If you have worked down this page and nothing above fits, that is the answer, and it is a finished one rather than a failure to find the real cause.

Read your product’s guide

Each guide says, for the things that product shows, what the source covers, how current it is, and what it will not tell you. Start with what it says about its own messages: where that product puts them, which of its shorter wordings cover more than one situation, and which of them are not failures at all. A message about a source being unavailable is often fully explained by a limit the guide already states.

Each guide keeps that under Support in its own navigation. Read yours for what your product says, and come back here for what to do about it.

If it is ours, report it

Reporting is a real next step rather than a consolation, and nothing else changes the cause. Your product’s guide says where to report.

What makes a report worth reading is the exact message, what you were doing immediately before, what you expected instead, and roughly when it happened. A theory about the cause is welcome but is not needed; if you have one, put it beside what you saw rather than in place of it, because what you saw is the part nobody else can supply.

An automatic error report is not your report. When a product runs into an error it recognizes, it can send one to a third-party error service on its own — which service, and what it carries, is on that product’s privacy page. That describes a fault the product noticed. It does not say what you were trying to do, what you expected, or which thing on the screen was the wrong one, and plenty of problems produce none at all because nothing broke as far as the product could tell. A result that confused you, an empty panel, a limit nobody told you about: none of those look like an error from the inside.

So the two are not substitutes, and there is no ticket for you to come back and check either way. Reporting is your decision, and nothing here depends on it.

If you would rather ask an AI assistant

This means a chat assistant you already use somewhere — in your phone, your browser, your search engine, or an office program, or one you reach at its own address. If none of those is part of how you already work, skip this section; everything above works with nothing but you.

A second page holds an AI assistant to how these products behave, so you can hand one the address and be asked the questions instead of reading them. It carries how to sort what you tell it, what it may not conclude, what it may not stand in for, and what it must never ask you for. Nothing on it is kept from you.

Send it this, which is a request rather than a bare link. An assistant handed only an address has been asked for nothing, and is likelier to tell you about the page than to use it.

Its first reply tells you whether it worked. One that has read the instructions asks which of the three products you were using and what happened, one question at a time. If it summarizes the page instead, or offers a diagnosis before asking you anything, you are getting its own answers rather than the instructions. Some assistants cannot open a link at all, and not every one of those says so plainly. None of that costs you anything — this page is still here, and it is the better route.

If it cannot open a link, paste the pages in instead. It needs both this page and the instructions, so that is two presses rather than one, and it works where nothing else does.

What to leave out

An assistant runs somewhere else, on someone else’s computer. Whatever you type into one has left your machine and cannot be taken back. None of this applies to reading this page; all of it applies there.

• The pairing code a product shows when it connects to something on your own machine. It is yours, and it is not needed to work out what went wrong.
• Your exact position. Your grid square is enough for any question about coverage or conditions. Latitude and longitude are not needed.
• Anyone else’s callsign, position, or equipment. What you saw on the map belongs to the operator who reported it, and nothing about your problem needs it.
• Anything from another window. If you send a picture, make it the part that matters rather than the whole screen. A full screen carries every other window, tab, and notification on it.

The instructions tell an assistant the same thing, and tell it not to ask you for any of these. If it asks anyway, it has stepped outside them, and you can say no without losing anything.