Loading…
Guide
Put the dates, titles, notes and groups from a plan or an event record onto one time axis. A vertical layout suits reading item by item and a horizontal layout suits a slide. The date-proportional mode keeps the real gaps between dates, while the evenly spaced narrative mode highlights the sequence without expressing real duration. The diagram and the underlying draft can be saved together, so you are not left with a screenshot that cannot be edited again.
Updated 2026-09-193 sources6 min read
Timelines are a long-evolving form of temporal data visualisation, and not every modern style can be attributed to a single proposer. As primary research on modern interactive time narratives, Brehmer, Lee and colleagues' Timeline Storyteller shows how representation, scale and layout each affect the narrative of events. This tool only adopts the horizontal and vertical directions and the two spacings that suit lightweight generation; it does not claim to reproduce that system fully. Research and author description
This implementation maps valid calendar dates to a UTC day number, sorts by that number, then computes the interval between the first and last. In the proportional mode, the position of a date corresponds linearly to the number of days and note cards may be nudged aside; a card's position should not be read as another date. Events sharing a date keep their original input order. A single date cannot form a span, but the diagram still shows that event.
A full timestamp in RFC 3339 also carries a time and an offset; this tool uses only the complete date form and performs no time-zone conversion. CSV import follows the common quoting and escaping rules but does not claim to cover every spreadsheet dialect. The date specification and the CSV format description are the basis for data interchange, not project management theory.
Compared with a Gantt chart, this tool mainly shows event points and computes no durations, dependencies or resource conflicts. Compared with a cause-and-effect diagram, one thing following another does not prove that the first caused the second. A tidy diagram does not mean the plan is executable either; a review still has to ask who is responsible, how much effort goes in and when completion will be confirmed.
The “Service launch milestones” example starts with scope confirmation on 2026-10-01 and ends with the pilot review on 2026-10-22, 21 days apart. 2026-10-15 carries two events, end-to-end verification and support handover; they share a date and should not be drawn a day apart.
A second “Thesis and study progress” example runs from 2026-11-02 to 2026-12-10 and contains the research question, the source matrix, the outline, the first draft and the submission check. Importing the short table below shows the difference between the proportional and evenly spaced modes:
date,title,description,group
2026-10-01,Confirm scope,Record acceptance criteria,Product
2026-10-07,Prototype review,"Record problems, not preferences",Design
2026-10-15,End-to-end verification,Verify recovery and export,Engineering
2026-10-15,Support handover,Confirm the incident contact,Operations
2026-10-22,Review,Decide the next step from the observations,ProductUp to 80 events, with dates from 1000 to 9999 in four-digit years.
No BC years, no uncertain dates that give only a month, no cross-time-zone instants and no recurrence rules.
A blank date is not treated as today automatically; an invalid event is not drawn on the timeline, but its fields can still be kept in the complete draft.
Importing a CSV that contains an invalid date is rejected outright rather than guessed at.
Extremely long notes enlarge the canvas, and PNG may be scaled down to stay within the pixel budget.
Export SVG when the result has to be zoomed, and use JSON to restore editing.
Fonts in an SVG depend on the viewing device, so CJK or unusual characters may look different.
One centimetre in the evenly spaced mode is not one week. Do not substitute date order for causal analysis, and do not write an unconfirmed plan as a fact. Changing “launched” to “pilot users completed the specified task and confirmed receipt” makes a milestone more verifiable.
CSVs from different teams should first have the meaning of their dates aligned: start, completion, approval or first observation. Otherwise the same date column mixes different event semantics, and even a correctly sorted result will mislead the reader.
They are different events, and sharing a date should not cause one note to be overwritten. The input order is used for stable display.
Choose the “ordered events” table and it can be re-imported; the general raw-field table and CSVs from other tools are not guaranteed to be compatible. Use JSON for a complete restore.
On a narrow screen the editor and the canvas are switched with tabs while the export bar stays available; this changes neither the data nor the exported result.
Date sorting, drawing and exporting happen in the local browser with no account sync. A share link is an unencrypted snapshot and should not contain sensitive schedules or personal information; above 8 KB, use a JSON file. Browser archiving can be cleared, so JSON should be saved separately. Reference links are shown only as sources and are not visited automatically because of an import.
Updated 2026-09-19
Import dated events from CSV, select horizontal or vertical and proportional or equal spacing, and export SVG or PNG
Checking local recovery and shared data…
Local storage is not a backup and can be cleared by the browser. JSON restores all fields. Share links are not encrypted; do not include sensitive data.