Skip to content


Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Commits on Apr 16, 2014
  1. @russelldb

    Refactor: rename entry for dot (cos that's what it is)

    russelldb authored
    De-opaqueise the dot. Add a `pure_dot` for non timestamped comparing
Commits on Feb 28, 2014
  1. @seancribbs

    Fix types in vclock.

    seancribbs authored
    * vc_entry is spurious
    * Only one opaque type was needed, dot()
    * Empty lists are vclocks too! No need to widen the type.
Commits on Jan 23, 2014
  1. @russelldb
  2. @reiddraper

    Remove redundant test

    reiddraper authored
    This test duplicates the behavior of `vclock_qc`
  3. @russelldb @reiddraper

    Add explanatory comment to `dominates` for posterity

    russelldb authored reiddraper committed
  4. @cmeiklejohn @reiddraper

    Dominates should not take timestamp into account.

    cmeiklejohn authored reiddraper committed
    Ensure that dominates does not take the timestamp into account; two
    concurrent puts happening one second apart should not be considered
Commits on Dec 19, 2013
  1. @russelldb
Commits on Dec 18, 2013
  1. @russelldb
  2. @russelldb

    Address review comments

    russelldb authored
Commits on Dec 9, 2013
  1. @russelldb

    Make get_entry return more information for callers

    russelldb authored
    Add test for get_entry
    Add valid_entry function to test if a dot is a valid format
Commits on Dec 4, 2013
  1. @russelldb

    Add get dot and dot type for dvv style causality

    russelldb authored
Commits on Dec 3, 2013
  1. @seancribbs
  2. @seancribbs
  3. @seancribbs
Commits on Oct 22, 2013
  1. @jtuple
Commits on Sep 10, 2012
  1. @argv0
Commits on Nov 15, 2011
  1. @matehat
Commits on Sep 4, 2011
  1. @jonmeredith

    Changed get_counter to return 0 on missing rather than undefined.

    jonmeredith authored
    Only consumer I could find was in riak_kv_vnode - and that was
    wrapping vclock:get_counter with a function to do the conversion.
Commits on Sep 1, 2011
  1. @jonmeredith

    Incorporated vclock:merge improvement suggested by Scott.

    jonmeredith authored
    Saves re-createing tuples.
Commits on Aug 30, 2011
  1. @jonmeredith
  2. @jonmeredith
Commits on Aug 1, 2011
  1. @Vagabond
Commits on Jul 8, 2011
  1. @eriksoe

    vclock: faster timestamp().

    eriksoe authored
  2. @eriksoe
  3. @eriksoe
Commits on Jun 9, 2011
  1. @Vagabond
Commits on Jun 6, 2011
  1. @eriksoe @Vagabond

    vclock: simplifying merge/3 using lists:reverse/2.

    eriksoe authored Vagabond committed
  2. @eriksoe @Vagabond

    vclock: simplification of descends/2

    eriksoe authored Vagabond committed
  3. @eriksoe @Vagabond

    vclock: Replacing proplists:get_value() calls with far faster lists:k…

    eriksoe authored Vagabond committed
Commits on Apr 4, 2011
  1. @argv0

    dialyzer: riak_core fixes

    argv0 authored
Commits on Mar 17, 2011
  1. Add vclock:increment variant that accepts a timestamp.

    Jon Meredith authored
    For bz://977 value duplication issue.
Commits on Mar 3, 2011
  1. @jonmeredith
  2. @eriksoe

    vclock:prune(): In case of oldish vclock entries with identical times…

    eriksoe authored
    …tamps, the entries to be deleted was arbitrary. This might lead to merge conflicts later on. Making the sort order deterministic to avoid this issue.
Commits on Jul 21, 2010
  1. @argv0

    initial import

    argv0 authored
Something went wrong with that request. Please try again.