1. 15 May, 2019 2 commits
  2. 19 Mar, 2019 1 commit
  3. 02 Feb, 2019 1 commit
  4. 10 Jan, 2019 1 commit
  5. 12 Dec, 2018 1 commit
  6. 22 Nov, 2018 1 commit
  7. 08 Nov, 2018 1 commit
  8. 17 Oct, 2018 1 commit
  9. 09 Aug, 2018 1 commit
  10. 08 Aug, 2018 1 commit
  11. 24 Jul, 2018 1 commit
  12. 14 Jul, 2018 1 commit
  13. 25 Jun, 2018 1 commit
  14. 13 Jun, 2018 1 commit
  15. 27 Mar, 2018 1 commit
  16. 09 Feb, 2018 1 commit
  17. 07 Feb, 2018 1 commit
  18. 31 Jan, 2018 1 commit
  19. 07 Dec, 2017 1 commit
  20. 02 Dec, 2017 1 commit
  21. 30 Nov, 2017 3 commits
  22. 21 Nov, 2017 3 commits
  23. 17 Nov, 2017 1 commit
  24. 06 Nov, 2017 1 commit
  25. 01 Nov, 2017 1 commit
    • Nathan Scott's avatar
      Addition of a redis-doc subpackage for Redis v4 onward · 0500a44e
      Nathan Scott authored
      This adds a redis-doc subpackage, containing the upstream
      docs (https://github.com/antirez/redis-doc/), as well as
      the existing man page content we've added.  Care is taken
      to handle the movement of man page files between packages
      and the naming convention of the redis-doc source is also
      careful to follow the prefered Fedora conventions.
      Optional conversion of markdown to html is done by pandoc
      if available.
      In the future, it would be to modify the redis build to
      generate the redis src/help.h file using the content from
      redis-doc now that we have it.  It's often out of date, I
      am noticing - seems to be updated in an ad-hoc fashion in
      upstream releases.
  26. 02 Oct, 2017 1 commit
    • Nathan Scott's avatar
      Redis v4 modules packaging preparation · da65cc3c
      Nathan Scott authored
      Add spec file macros to create a default location for
      Redis modules, and to allow module packages to depend
      on a specific redis module ABI.  A build time test is
      added to ensure the upstream sources API version does
      not differ to one the redis package Provides:.
      The devel package now provides an rpm macros file for
      the %{redis_modules_abi} and %{redis_modules_dir}.
      Much of this is ideas from Haikel and Remi - thanks!
      Small cleanup - the _hardened_build macro is dropped
      as this has been the default since f22.
  27. 27 Sep, 2017 2 commits
  28. 26 Sep, 2017 3 commits
    • Nathan Scott's avatar
      Drop duplicate LDFLAGS setting via LUA_LDFLAGS · e35b514e
      Nathan Scott authored
      Remi noticed an inconsistency in setting these
      flags - auditing the makefiles reveals there is
      no need to use the LUA-specific settings as the
      supplied LDFLAGS and CFLAGS are honoured.
    • Nathan Scott's avatar
      Incorporate recent feedback from Remi, build once · 59378174
      Nathan Scott authored
      Drop modification of the configuration file to add
      systemd supervision - Remi correctly points out this
      is redundant (due to unit files already overriding)
      and will just cause issues.
      Noticed the build runs twice also, once during the
      %build phase and once again during %install (full
      rebuild) - this was because of differing options on
      the make lines (CFLAGS= etc vs PREFIX= etc) and the
      Redis src/Makefile causes full rebuild as a result
      (including of deps/* which it "make distclean"s).
      Also fixed a rpmbuild "warning: Macro expanded in
      comment" message in a couple of places, and changed
      the way we squash hidden build messages to use the
      "V" macro as the Makefiles intended.
    • Nathan Scott's avatar
  29. 12 Sep, 2017 1 commit
  30. 07 Sep, 2017 2 commits
    • Nathan Scott's avatar
      Incorporate recent feedback from Haikel and Remi · c374de36
      Nathan Scott authored
      Add "--supervised systemd" to the command line, and
      drop the "--daemonize" option as Redis already does
      this internally in supervised mode.
      Add Provides:bundled spec file lines for linenoise,
      lua-libs and hiredis.
    • Nathan Scott's avatar
      Refactor some patches out of redis.spec · 9f0877ed
      Nathan Scott authored
      For cases where we have two ways of doing one thing (using
      a patch or sed-based modification of files), rework these
      such that we're only using one method.  Allows several of
      the patches to be dropped.
      Described in https://github.com/antirez/redis/issues/1417
      is a method for allowing integration/replication-psync to
      be reliably run - start using it, and drop the patch that
      was disabling that test.
      Other small side-effects of this commit:
      - the build warning about "jemalloc not found" is fixed;
      - CFLAGS additions are now consistently handled (for deps
        as well) - using the upstream-provided mechanism for
        compiler flag propogation.
  31. 28 Aug, 2017 1 commit