Users want to be able to gauge relative time between log events. The absolute timestamps are useful for detail analysis, but they’re difficult for (humans) to parse to quickly compare for an overview of the timeline. An initial attempt adds an additional column in the table to show the duration (in seconds) since the last event. A potentially better implementation would use a slopegraph that maps the a timeline to discrete rows of unit height in the existing table. There would be no lines crossing since both axes are monotonically increasing.
jmakeig
commented
9 years ago 
Users want to be able to gauge relative time between log events. The absolute timestamps are useful for detail analysis, but they’re difficult for (humans) to parse to quickly compare for an overview of the timeline. An initial attempt adds an additional column in the table to show the duration (in seconds) since the last event. A potentially better implementation would use a slopegraph that maps the a timeline to discrete rows of unit height in the existing table. There would be no lines crossing since both axes are monotonically increasing.