Before sharing a Base64 support example, decode it and review the text it actually contains. Changing an email field is not enough if the same address remains inside a message. This walkthrough checks both places, then rebuilds and verifies a small fictional example for public use.
All names, addresses and events below are invented. Prepare a separate fictional reproduction before putting data into a public tool or support thread; review real records only in an environment approved for that data.
Start with the decoded record you intend to share
Suppose a fictional export produces this one-line JSON record:
{"email":"maya@workshop.example","message":"Export for maya@workshop.example stopped at row 18."}
The email address occurs twice: once in a named field and once in free text. Keep this readable copy beside the encoded sample during the exercise. A field list alone would miss the second occurrence.
Base64 represents data as text for storage or transfer, as described in MDN’s Base64 overview. It does not make the underlying address confidential. RFC 4648 security considerations specifically discuss information being revealed when people mistake base encoding for protection.
Recover the address from the encoded example
- Open the Base64 Encoder and Decoder and choose Encode text to Base64.
- Paste the one-line fictional JSON above into Input, without an extra newline.
- Leave URL-safe alphabet and Wrap output at 76 characters unchecked, then select Convert.
The current tool produces:
eyJlbWFpbCI6Im1heWFAd29ya3Nob3AuZXhhbXBsZSIsIm1lc3NhZ2UiOiJFeHBvcnQgZm9yIG1heWFAd29ya3Nob3AuZXhhbXBsZSBzdG9wcGVkIGF0IHJvdyAxOC4ifQ==
Select Use output as input. The tool switches to decoding and restores the original JSON, including both copies of the address. No password or secret key is requested. An address being absent from the visible encoded characters is therefore not a privacy check.
Catch the incomplete edit before rebuilding the example
Now change only the email field. The partly edited record becomes:
{"email":"reader@example.com","message":"Export for maya@workshop.example stopped at row 18."}
Encode this version and decode it again using the same controls. The decoded email field has changed, but the message still contains maya@workshop.example. Encoding after an incomplete edit preserves that incomplete edit.
For our public reproduction, rebuild the message using only the fictional details needed to describe the export symptom:
{"email":"reader@example.com","message":"Export stopped at row 18."}
In your own preparation, review every field and all free text, including copied notes, URLs and filenames. Check for identifiers repeated in different places; a search for one known address is useful but cannot certify that every private detail has been found. A small example written from scratch is easier to review than a large log with selected fields replaced.
If an address itself is necessary to reproduce the issue, use a fictional test address and verify that the example still demonstrates the relevant behavior in your test environment. Our encoding checks do not reproduce an application export failure.
Verify the final decoded copy against your public example
Encode the rebuilt record, then decode that new output. Compare the entire decoded record with the intended public example, rather than inspecting only its email field. In our checks of the current tool code, the three round trips gave these results:
| Version | Decoded email field | Original fictional address still in message? |
|---|---|---|
| Original | maya@workshop.example |
Yes |
| Email field changed only | reader@example.com |
Yes |
| Rebuilt public example | reader@example.com |
No |
All three decoded strings matched their respective inputs exactly. That proves the conversions preserved these examples; it does not establish that an arbitrary record is anonymous or safe to publish. The content review happens before encoding, and the final decode checks that the intended copy was used.
Include the readable fictional example alongside its encoded form when it helps the recipient understand the issue. If accented text or symbols change during conversion, the separate UTF-8 Base64 round-trip guide covers that encoding problem. If the decoded JSON needs a clearer presentation, see preparing a fictional API example for review.