The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

abba 0.2.0

Breaking changes

This release removes implicit file writes to the user’s filespace in order to comply with CRAN policy. Several functions now require arguments that previously defaulted to a location derived from other inputs. Update call sites to pass these arguments explicitly.

create_batch_api() and create_batch_job()

• path is now required. The previous default of "." (the working directory) has been removed.
• Before: create_batch_api() wrote plumber.R to getwd().
• After: callers must supply an explicit path, e.g. create_batch_api("path/to/api_directory").
• Also fixes a bug where a path containing . (for example "~/api_directory/plumber.R") was silently rewritten to "./plumber.R". The user-supplied path is now respected.

abba_slurm_submit_job(), abba_rslauncher_submit_job_local(), abba_rslauncher_submit_logrx_job_local()

• log_path is now required. There is no longer a default that writes the log next to the submitted program.
• Before: omitting log_path caused the log to be written to dirname(program_path).
• After: callers must pass log_path explicitly, e.g. abba_slurm_submit_job("/path/to/script.R", log_path = "/path/to/logs").
• When using abba_submit_batch() with a Workbench or SLURM submit function, pass log_path through the ... arguments, e.g. abba_submit_batch(progs, submit_func = abba_rslauncher_submit_job_local, log_path = "/path/to/logs").

abba_submit_batch() and abba_submit_batch_and_get_results()

• cache_folder is now required whenever update_cache = TRUE or rerun_unchanged_programs = FALSE. The previous fallback that wrote a .abba_cache directory next to each program has been removed.
• Before: caching silently wrote to dirname(program_path)/.abba_cache.
• After: callers must supply cache_folder explicitly or set it via the abba.default_cache_folder option. A clear error is raised if caching is requested without a cache_folder.
• When update_cache = FALSE and rerun_unchanged_programs = TRUE (the defaults), no cache writes occur and no cache_folder is required.

Internal

• Cache helper functions (cache_match(), abba_save_file_cache()) now require cache_folder and error clearly when it is missing. This also stops an existing code path from writing a cache into the program’s directory when the top-level call did not request caching.

Documentation

• Package Title and Description now quote external software names ('Kubernetes', 'SLURM', 'Posit Workbench') per CRAN conventions.
• Examples and vignettes updated to show explicit paths for create_batch_*, log_path, and cache_folder rather than defaults.
• The standalone LICENSE file has been removed; License in DESCRIPTION is now Apache License (>= 2) without the file LICENSE pointer, which is the correct form when there are no additional restrictions.

abba 0.1.0

• Initial CRAN release
• Support for Kubernetes job submission and monitoring via kubectl
• Support for SLURM job submission and monitoring via sbatch/scontrol/sacct
• Support for Posit Workbench job submission via RSLauncher
• Batch execution with parallel and sequential modes
• File-based caching to skip unchanged programs in batch reruns
• logrx integration for detailed execution logging
• Plumber API template for secure identity segregation on Posit Connect

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.