Eyevinn / go-608 / 30906319596

85%

Build Type

push

github

Committed by tobbee

Commit Message

docs(generate): state roll-up window depth as rows against lines

Roll-up's window depth was described as if a second occupied one row, which is only true for a single-line caption. Every configured line is its own scroll step, so a cue of L lines consumes L of the rows rows, and every useful relation is rows against L rather than rows against seconds:

  rows == L      no history at all -- each cue scrolls the previous one clean off
  L < rows < 2L  the previous cue's bottom line survives
  rows >= 2L     the previous cue survives whole
  fill time      ceil(rows/L) cues, not rows-1

The consequence that most needed saying: with DefaultConfig's two lines, rows 2 is the no-history case -- and that is exactly what WithRollUp's zero value and go608-clock's plain -mode roll-up select. WithRollUp, generate/doc.go, BuildUnitRollUpCues, the -mode flag help, README and W7-W8 all claimed the previous seconds stay visible there. They do not; it looks like pop-on that types itself out. roll-up4 is the setting that keeps a previous second whole.

Two further claims were wrong for the same reason. A window fills after ceil(rows/L) cues, not rows-1, which happened to agree only at 2 and 3 rows: decoded at 30 fps, a 2-row window completes at 0.400 s but 3- and 4-row windows both complete during the *second* second. And 'a 2 s segment cannot fill a 4-row window at all' is false for a two-line caption -- two one-second cues of two lines fill 4 rows exactly, as a decode of -mode roll-up4 -unit-mode default -unit-seconds 2 shows. Under reset the depth is L*N rows for N = NumCues cues.

TestGeneratorRollUpHistoryDepth pins the settled window and the fill second for rows 2, 3 and 4 so the rule is enforced rather than merely written down. No emitted data changes -- this is the arithmetic stated correctly. Recorded as W10.

Also corrects two figures the previous commit missed: WithRollUp's own doc comment and TestGeneratorRollUpBudget's comment both still said the default ... (continued)

Coverage Stats

2 of 2 new or added lines in 1 file covered. (100.0%)

8 existing lines in 1 file now uncovered.

3819 of 4472 relevant lines covered (85.4%)

0.99 hits per line