aboutsummaryrefslogtreecommitdiff
path: root/module/entry-points/ical.scm (unfollow)
Commit message (Collapse)Author
2020-07-07Merge (datetime util) into (datetime).Hugo Hörnquist
2020-05-23Remove (vcomponent load) module.Hugo Hörnquist
2020-05-04Add ability to download ics in entirety.Hugo Hörnquist
2020-05-04Add --help to ical.Hugo Hörnquist
2020-04-30Update ical parts to use app context.Hugo Hörnquist
2020-04-01Update remaining code to use new config scheme.Hugo Hörnquist
2020-03-31Attempt at decentralized configuration system.Hugo Hörnquist
Placing all possible configuration items in a central (parameters) module scales really badly. This idea that any module can register configuration parameters is better. The current implementation however has the drawback that it requires that the module exposing the parameter is loaded before the value can be sat, but that scales even worse. A probable solution would be to abandon binding everything to guile's module system, and instead let (util config) provide a `conf-ref' and a `conf-set!' procedures. A `define-configuration' similar to emacs `defcustom' could be of use, mainly for retroactively type checking parameters.
2020-02-18Freed datetime from its srfi-19 prison.Hugo Hörnquist
2020-01-24Start of new date structures.Hugo Hörnquist
2019-12-30Add file flag to ical entry point.Hugo Hörnquist
2019-12-24ICAL handling of events different from display handling.Hugo Hörnquist
Previously repeating events where always instantiated to a stream of all events to come (possibly infinite), and then zipped with the list of regular events to create a stream of all events in the world. This commit allows access to the raw lists of parsed regular and repeating events before they are extrapolated and merged.
2019-12-24Rework program initialization.Hugo Hörnquist
Old init setup had the fancy idea to parse all files before anything could be done with them. This however led to problems when a part of the program which didn't care for the calendar files (such as text formatting). It also made testing performance almost impossible since to much code was run before I had a chance to init statprof.