Time.timestamp does not seem to satisfy the semantics implied in documentation

[jvoigtlaender]

See https://groups.google.com/d/msg/elm-discuss/AWx51F_WzUE/iKBdMM94JJoJ

[jvoigtlaender]

Also, I think the current implementation makes it possible that in:

test : Signal (Time,Time)
test = timestamp (every second)

pairs occur as value on the signal whose first and second component differ, contrary to what one would expect from the documentations of Time.every and Time.timestamp.

See this thread: https://groups.google.com/d/msg/elm-discuss/KzevSWc0gfU/UO5-BQozwEUJ

[jvoigtlaender]

Also, I think the current implementation makes it possible that in:

test : Signal (Time,Time)
test = timestamp (every second)

pairs occur as value on the signal whose first and second component differ, contrary to what one would expect from the documentations of Time.every and Time.timestamp.

See this thread: https://groups.google.com/d/msg/elm-discuss/KzevSWc0gfU/UO5-BQozwEUJ

[jvoigtlaender]

The PR #76 also fixes the above issue concerning timestamp (every second) as well as a similar issue concerning interaction between timestamp and fps/fpsWhen.

[jvoigtlaender]

The PR #76 also fixes the above issue concerning timestamp (every second) as well as a similar issue concerning interaction between timestamp and fps/fpsWhen.