Annual Review of Attachment Preview Distrust
A year-end assessment of the gap between seeing a file and believing it has been received.
ANNUAL REVIEW: ATTACHMENT PREVIEW DISTRUST Review Period: January through December Subject: Downloading files already previewed in the same session
Q1: Origin and First Occurrence
Attachment preview distrust was documented in its earliest form when email interfaces began rendering document contents inline, eliminating the technical need to save a local copy before reading. The behavior persisted across this transition. People downloaded files after previewing them. The download produced a local copy. The local copy was sometimes opened once and never consulted again.
Q1 status: present and stable.
Q2: Contributing Factors
The preview pane renders the document. The document is visible. The text can be read. The images are displayed. The data is accessible. And yet.
The preview exists in a location that is not the device. It is contingent on connectivity, on the continued existence of the attachment in the mail server, on factors that are not controlled by the person reviewing it. A local copy is permanent in a way the preview is not. The local copy does not require a network.
This logic is sound. It applies in roughly two percent of cases. In the remaining ninety-eight percent, the preview is sufficient, the attachment is available indefinitely, and the local copy occupies storage for months before being discovered and deleted.
Q3: Consolidation
Q3 established the baseline: a file rendered in the preview pane is not treated as a file in possession. The preview existed. The download happened anyway. This is the behavior under review.
The local copy provides no additional information but produces a feeling of finality that the preview does not. Finality is the product. The file is not the product.
Q4: Forecast
Attachment preview distrust has acquired standing. It will continue to produce local copies at approximately the rate observed this year. Storage will accumulate. Periodic cleanup events will remove the downloads. New downloads will replace them.
The behavior costs approximately three seconds and an unquantified amount of disk space per instance. It is not a major expense. It is the kind of expense that repeats so many times it becomes structural, and structural expenses are rarely reviewed as expenses.
This is the review.