G'day all.. checking in with my usual 2 (or 20..) cents.
I'm loving this article, and discussion, I always learn lots from these.
Eventual consistency is a nice phrase to highlight. I feel (https://fosstodon.org/@simonmic/113444957161144239) the same way.
For me it's about not feeling you're on a time-sensitive treadmill, where if (when) you fall behind, you might risk losing data or never catching up.
(Though book-keeping regularly is still advantageous because it's less work.)
I believe the "pure function" style of many PTA tools, where all inputs are known and the output is deterministic, combined with version control, are a big part of this.
For newcomers, I like to make clear there's multiple styles of doing plain text accounting, including:
1. Entering all your own data by hand (optionally assisted by data entry tools).
2. Downloading data periodically from banks (as CSV, OFX, PDF, or whatever you can get) and importing it once to your journal file.
3. Downloading data periodically from banks, keeping that as your master data, and frequently regenerating journal files from it.
Each of these has its pros and cons, and there are people successfully using each of them.
hledger's own docs focus on methods 1 and 2.
Several third-party workflows using method 3 have been built on top of hledger; Full-Fledged Hledger that OP is using is the best known.
1 requires nothing but hledger.
2 adds: either manual bank downloading, or setting up one or more bank download tools/APIs; plus maintenance of conversion rules/scripts.
The workflows implementing 3 add: a recommended file layout, a recommended process, and tools/scripts which depend on these. They bring some benefits (a well documented featureful workflow) and some costs (added complexity). This isn't justified for all users, so don't think you're required to use these workflows.
4. Downloading CSV (TSV, SSV, *SV) data periodically from banks, and using it directly as hledger's data files, without converting to journal format at all. (Optionally adding a journal file just for configuration directives.)
5. Generating one of hledger's input formats from somewhere and piping it into hledger's stdin, running it without saving any data files.
Comments
G'day all.. checking in with my usual 2 (or 20..) cents. I'm loving this article, and discussion, I always learn lots from these.
Eventual consistency is a nice phrase to highlight. I feel (https://fosstodon.org/@simonmic/113444957161144239) the same way. For me it's about not feeling you're on a time-sensitive treadmill, where if (when) you fall behind, you might risk losing data or never catching up. (Though book-keeping regularly is still advantageous because it's less work.) I believe the "pure function" style of many PTA tools, where all inputs are known and the output is deterministic, combined with version control, are a big part of this.
For newcomers, I like to make clear there's multiple styles of doing plain text accounting, including:
1. Entering all your own data by hand (optionally assisted by data entry tools).
2. Downloading data periodically from banks (as CSV, OFX, PDF, or whatever you can get) and importing it once to your journal file.
3. Downloading data periodically from banks, keeping that as your master data, and frequently regenerating journal files from it.
Each of these has its pros and cons, and there are people successfully using each of them. hledger's own docs focus on methods 1 and 2. Several third-party workflows using method 3 have been built on top of hledger; Full-Fledged Hledger that OP is using is the best known.
1 requires nothing but hledger.
2 adds: either manual bank downloading, or setting up one or more bank download tools/APIs; plus maintenance of conversion rules/scripts.
The workflows implementing 3 add: a recommended file layout, a recommended process, and tools/scripts which depend on these. They bring some benefits (a well documented featureful workflow) and some costs (added complexity). This isn't justified for all users, so don't think you're required to use these workflows.
More details: https://hledger.org/workflows.html
4. Downloading CSV (TSV, SSV, *SV) data periodically from banks, and using it directly as hledger's data files, without converting to journal format at all. (Optionally adding a journal file just for configuration directives.)
5. Generating one of hledger's input formats from somewhere and piping it into hledger's stdin, running it without saving any data files.
Thank you for laying these out. And thank you for all the work you've done to make my tiny niche workflow possible.