No description
Find a file
Folling c4b06c09d1
filter ENOENT from boost::error_code check
Signed-off-by: Folling <mail@folling.io>
2025-01-01 14:07:47 +01:00
docs finalise interface & documentation 2025-01-01 14:07:46 +01:00
include fixup missing ikarus_ prefix for error_info_name 2025-01-01 14:07:47 +01:00
src filter ENOENT from boost::error_code check 2025-01-01 14:07:47 +01:00
vendor update sqlitecpp 2025-01-01 14:07:47 +01:00
.clang-format change error system & function signatures 2025-01-01 14:07:47 +01:00
.clang-tidy add src/ikarus subdir and make names unique for objects per scope 2025-01-01 14:07:47 +01:00
.gitignore a new beginning 2025-01-01 14:07:46 +01:00
.gitmodules a new beginning 2025-01-01 14:07:46 +01:00
CMakeLists.txt fixup Boost_INCLUDE_DIRS -= S 2025-01-01 14:07:47 +01:00
LICENSE.md finalise interface & documentation 2025-01-01 14:07:46 +01:00
README.md change error system & function signatures 2025-01-01 14:07:47 +01:00

Data Longevity

All data returned by libikarus is ephemeral and only represents the state of the project at the time of the request. A snapshot if you will. One must not rely on it representing the actual state of the project at any given time. The data is simply copied from the underlying data sources and returned to the caller.

No mechanisms are provided to avoid race conditions. libikarus itself should only be used in a single-threaded context. However, nothing breaks if you do use it in a multithreaded context, that is, libikarus is threadsafe. You just cannot rely on the data being consistent. This goes especially for inter-process access to the same project.