Ptime.Span.pp with leading zeros. For example 1.036s would render as 1.36s. This is a rendering bug in a function for human display, not a bug in the computations or conversion functions of Ptime.Ptime.weekday type for naming the result of the Ptime.weekday function.ptime.clock.os library is deprecated. Use ptime.clock instead.ptime.clock export ptime.Ptime.of_rfc3339 timezone offset parsing. Be even more lenient in non-strict parsing mode: allow hhmm and hh timezone offsets. (strict is hh:mm). Allows to parse an even larger subset of ISO 8601 than RFC 3339 (#31).Ptime.{to,of}_year. Less costly than extracting the first component of Ptime.to_date_time. Useful for example to find out which DST rules a timestamp is subjected to for rendering.?tz_offset_s optional argument to Ptime.{of,to}_date (#32).Ptime.weekday_num. An integer is often more convenient than the enum value of Ptime.weekday (#30).Ptime.rfc3339_string_error convenience function.js_of_ocaml META ocamlfind standard to link JavaScript stubs (#28).ptime.clock package, this package is now empty.js_of_ocaml strategy for Ptime_clock's JavaScript implementation. Primitives of ptime.clock.os are now implemented in pure JavaScript and linked by js_of_ocaml. This means that the ptime.clock.jsoo library no longer exists, simply link against ptime.clock.os instead. Thanks to Hugo Heuzard for suggesting and implementing this.result compatibility package.ptime/{os,jsoo} are now installed in ptime/clock/{os,jsoo}. Also a ptime_clock.cm[t]i is now installed in ptime/clock/. The ocamlfind packages are unchanged except for ptime.clock.os.top which no longer exists.Pervasives deprecation.Ptime.truncate to always truncate down. Thanks to David Kaloper Meršinjak for the report & fix.js_of_ocaml 3.3.0's namespacingPtime_clock: Windows support. Thanks to IndiscriminateCoding and David Allsopp for the contribution.Ptime.frac_s on pre-epoch time stamps. The function computed a span of 1s - f instead of f on these. This function is not used internally so this only affects users of this function that apply it on pre-epoch time stamps (#12). Thanks to David Kaloper Meršinjak for the report.?tz_offset_s optional argument to Ptime.weekday. Thanks to Maxence Guesdon for suggesting.Ptime.v and Ptime.Span.v to safely deal with trusted inputs. Thanks to Matt Gray for suggesting.Ptime.weekday, to help conversions to denormalized timestamp formats. Thanks to Romain Calascibetta for suggesting.First release. Thanks to Raphaël Proust for lodging support.