The project is inspired by excellent work of John Wiegley, ledger-cli: https://github.com/ledger/ Ledger.pl is written in Perl, which probably facilitates development. The major reason for the project is however to get rid of ubiquotous (re)valuation of commodities, which is major feature of ledger-cli. While being very handy in many circumstances, it is not suitable for situations, where historical cost shall be maintaind.
https://groups.google.com/forum/?fromgroups=#!topic/ledger-cli/f9CAsKtiyp8
The following ideas/principles are borrowed from ledger-cli:
- transaction data text file format
- application concentrates on reading data and producing reports. Amendment/addition to the data file is outside the scope of program
- 'commodity' concept: everything is commodity - currencies, benzin etc. It appears to me as the most valuable idea in ledger-cli. This implies multi-currency accounting, but embraces much wider scenarios, such as management accounting with transactions of non-monetary items.
- base currency (or probably, more properly, reporting currency) is not predefined in data file. Provided that appropriate rates are available, this makes it possible to prepare reports in any currency.
- traditional basic accounting data attributes, such as 'company', 'client', 'contract', 'project', 'period', 'journal' etc. wich is common to most accounting / erp applications, are not predetermined.
- 'complex' transactions with more than one Dt and Ct
- double entry system
- autobalancing mechanism: last transaction value, if empty, shall be auto-calculated
- tags at both transaction and single record level.
Changes to the concept of ledger-cli:
- lesser flexibility of the data file. This not only simplifies parser mainenance, improves performance, but also facilitate readability of the data file.
- separation of data from algorithms: no encolsures, prescribing data handling in data file.
- explicit data modification / revaluation
I hope to go as far as to write decent vim plugin.