The step that reports success either way
Encryption is a mechanical transformation. You hand it a block of text and a key, and it produces output. It has no way of knowing whose key that is, whether anybody holds the matching half, or whether the person you have in mind has ever seen it. If the key is well formed, the operation succeeds. That success is a statement about arithmetic and about nothing else.
So the condition on this card is not that you encrypted. It is that the key you encrypted to is one the far end can undo. Those are different claims and only one of them is checked by the tool in front of you. The output of a correct operation and the output of a useless one look the same: same armour, same shape, same absence of complaint. There is no visual difference to notice, which is why noticing is not the method.
There are a few ordinary ways the key turns out to be the wrong one. It was collected from somewhere other than where the listing points, and belongs to somebody real who is not the intended reader. It is an old key that has been replaced. Or, most commonly and most quietly, your own key was selected in a list of keys and the block is now something only you can read. In all three cases the failure is invisible locally and only becomes visible at the far end, by which point the contents have already left.
Two conditions that have to be settled first
The delivery address is written the way the post wants it has to be finished before this card can begin. What you seal is normally the address block, and sealing is not a step you get to redo cheaply: correcting a mistake means producing a second block, sending it, and hoping the right one is the one that gets used. Finality is the reason the edge runs in this direction rather than the other.
The listing is one you have actually read supplies everything about the key itself. Whether anything is expected to be encrypted, which key is the correct one, and where that key is published are all stated in the listing text and are rarely stated anywhere else. A key found by some other route may be genuine and still be the wrong genuine key, which is a distinction worth holding onto. The same distinction appears in a different setting on You can tell a mirror from a copy of the market.
What an unreadable block does to the waiting
One card sits below this one, You can wait without doing anything, and the mechanism connecting them is more direct than it looks.
That card asks whether you can leave a situation alone until the day you already named. Everything about an unreadable block is designed to make that impossible. If you suspect the block failed, you cannot rest, because there is a plausible action available and it feels urgent. If you do not suspect it, the silence has no explanation, and unexplained silence produces exactly the same behaviour: a second message, then a third, then a decision to resend the details some other way. That last step is the expensive one, because the plaintext version of everything you carefully sealed then travels through whichever channel felt convenient at the time.
The knock on effect continues. Waiting without acting feeds Escrow is still holding when the problem appears, and the actions people take to relieve the feeling of waiting are the actions that change the order state. A block nobody can read is therefore not a self contained failure. It is an engine that produces further failures for as long as it goes unresolved.
| What follows from an unreadable block | Why it follows |
|---|---|
| Waiting stops being possible | Silence has no explanation and a plausible action is always available |
| The address work is lost as well | The block that failed is the block that contained it |
| A second copy goes out by another route | The obvious fix undoes the reason for the first attempt |
| The order state moves while you wait | Actions taken to relieve waiting are actions on the order |
How the invisible failure spends the following days
You paste, you select a key, the tool reports that it worked, and you send the result. Nothing in that sequence gives you a reason to look twice. On your screen it is indistinguishable from every successful version of the same operation you have ever performed.
At the other end somebody opens it and gets nothing. Whether they write back to say so depends entirely on how they work, and you cannot assume they will. So what you experience is a gap, and the gap is where the misdiagnosis happens. You go looking for the fault in the address, then in the connection, then in the market, then in whether the account is behaving properly. The one thing you do not examine is the step that told you it succeeded, because there is nothing on the screen inviting you to. If this is false describes the general form of this: the loudest symptom is usually several cards away from the false condition.
Two checks that do not depend on anybody replying
The first is the direct one. Encrypt something harmless to the same key, send it, and ask whether it came out readable. A pass is a plain confirmation that quotes the harmless text back at you. A fail is no answer, or a report that it could not be opened, or a friendly reply that never mentions the test.
The second check needs nobody at all. Take the output you just produced and try to decrypt it yourself. If you can read it back, it was encrypted to a key whose private half you hold, which means it went to you and not to them. That is a genuine local test, it costs a few seconds, and it catches the most common misfire on this card without sending anything anywhere.
Neither a tool guide nor a safety claim
It is not a tutorial in any tool, and no tool is named here on purpose. It is also not a claim that encrypting something protects you, which is a much larger question this map does not answer. It asks one thing: can the recipient undo what you just did.