Changes in version 0.4.1                        

  - Enhanced slim_setup() function with improved reliability and user
    experience:
      - Added automatic conda Terms of Service acceptance to prevent
        installation failures
      - Implemented download retry logic with exponential backoff for
        more reliable binary downloads
      - Added structured error messages with platform-specific guidance
        and troubleshooting steps
      - Added post-installation verification that automatically tests
        the SLiM installation
      - Added informative progress messages throughout the installation
        process
      - slim_version parameter now accepts specific version numbers
        (e.g., "4.2.2", "5.1") in addition to "stable" (default:
        SLiM 4.2.2)
      - Warns users about potential compatibility issues when installing
        SLiM 5.x

                        Changes in version 0.4.0                        

  - Updated slimr to be fully compatible with SLiM v4.1 updates
  - Changed SLIM_HOME environmental variable to SLIMR_PATH, which now
    should be set with the full path to the SLiM executable, not just
    the directory containing it
  - Updated slim_setup() function which now provides cross-platform
    automatic installation using conda via the reticulate package. Now
    it should not be necessary for users to manually install SLiM any
    more
  - Simplified slim_run() API and temporarily removed ability to use R
    callbacks (the plan is to eventually replace this functionality with
    a new 'slimr verb' function)
  - Added new 'slimr verb' r_message(). r_message() can be called in a
    slim_block() to print a message to the R console while the SLiM
    simulation is running. This is useful for debugging and monitoring
    progress of a simulation.
  - slim_run() with progress = TRUE now produces a progress bar for any
    simulation (previously it only worked if you created output with
    r_output()) and is much more efficient (produces less overhead)
  - slim_script_render_dots() function added. It works just like
    slim_script_render(), but accepts values to be filled in a templated
    slimr_script as named arguments. For example
    slim_script_render_dots(x, mut_rate = c(1e-6, 1e-7)) will render the
    slimr_script called x with mut_rate set to 1e-6 and 1e-7. This can
    sometimes be less verbose than using a list or data frame to fill in
    a templated slimr_script, especially useful when you simply want to
    test different values quickly.

                        Changes in version 0.3.0                        

  - Added slim_auto() function to automatically insert r_template() and
    r_inline() statement in useful places in a slimr_script, based on
    computational graph analysis
  - Added as_slimr_code() function, which is similar to
    as_slimr_script(), which converts SLiM code as a character vector
    into a slimr_script object, but as_slimr_code() returns the code
    necessary to create the slimr_script instead. this allows copy and
    pasting the text into the R console or IDE so that it can be
    subsequently edited and modified.
  - Added vignette showing how to create a multispecies model in slimr,
    and documented the multispecies feature in the documentation for
    slim_block()

                        Changes in version 0.2.1                        

  - allow user to manually set SLiM executable directory with SLIM_HOME
    environmental variable

                        Changes in version 0.2.0                        

  - all SLiM callbacks now supported
  - slimr now fully supports SLiM versions 4.0 and greater, including
    support for multispecies models (though these have been extensively
    tested yet) and are currently undocumented
  - slimr verbs are now prefixed with r_ (e.g. r_output, r_inline,
    r_template). The old version slimr_ prefixed version are still
    available for backwards compatibility, but should be considered
    soft-deprecated

                        Changes in version 0.1.0                        

  - SLiM Classes available for help with autocomplete now have nested
    class objects (e.g. sim$subpopulations references a Subpopulation
    object, so you can then autocomplete sim$subpopulations$individuals,
    for example)
  - $ accessor can now be used in place of SLiM's . operator.
    (previously user's always had to use %.% to represent .)
  - slimr now replaces any <- assignment operators in SLiM code with =.

                     Changes in version 0.0.0.9001                      

  - slim_run() function runs a slimr_script in SLiM and collects results
  - slimr 'verbs' implemented (slimr_template(), slimr_inline(), and
    slimr_output()). These can be used inside SLiM coding block to
    powerful effect!
  - slim_block() creates blocks of SLiM code, and slim_script() puts
    them together into scripts.
  - Package functional

                     Changes in version 0.0.0.9000                      

  - Added a NEWS.md file to track changes to the package.