Does beancount teach me anything about my finances?

[Viren Bhanot]

I cannot understand how to actually *use* beancount to do *anything*. From what I see, bean-report doesn't exist anymore, and certainly not in my installation of beancount from AUR on Arch (btw).

If I load my ledger into fava, I see some poorly-designed plots not useful for analyses. I don't get the fuss. The plot area is too small, the colours don't contrast well, the tooltips often goes off-screen, or covers the plot so I cannot see where I am. Also, from what I can see, there is no way to limit the plots to just 2023 or 2024 etc. The most useful view is the Treemap of Expenses, but even their the text contrast is poor and I can't seem to generate it year-by-year. The documentation also seems non-existent.

The act of actually wrangling ten years of financial data into beancount was meditative, and a task I enjoyed thoroughly. But now what? How can I answer questions like:

* What is my net worth?

* What is my savings rate?

* What is my FIRE date at the current savings rate?

* How much have I been spending on Expenses:Restaurants year-over-year?

Must I generate bean-queries for everything? Because that rather defeats the purpose of a ledger file in my opinion. I could just as easily list my postings on CSV and write python queries to it. What's more, I don't see how it could be interactive and *fast*? In the absence of a quick feedback loop, what exactly is the tedium of writing out all of my postings going to get me?

As I said, I apologise if I come across a bit cranky but I think that. despite the trove of writing by Martin, I'm missing something fundamental about beancount.

[Adrian Utrilla]

On Mon, Aug 18, 2025 at 10:06 AM Viren Bhanot <goo...@virenbhanot.fastmail.com> wrote:

I cannot understand how to actually *use* beancount to do *anything*. From what I see, bean-report doesn't exist anymore, and certainly not in my installation of beancount from AUR on Arch (btw).

If I load my ledger into fava, I see some poorly-designed plots not useful for analyses. I don't get the fuss. The plot area is too small, the colours don't contrast well, the tooltips often goes off-screen, or covers the plot so I cannot see where I am. Also, from what I can see, there is no way to limit the plots to just 2023 or 2024 etc. The most useful view is the Treemap of Expenses, but even their the text contrast is poor and I can't seem to generate it year-by-year. The documentation also seems non-existent.

The act of actually wrangling ten years of financial data into beancount was meditative, and a task I enjoyed thoroughly. But now what? How can I answer questions like:

* What is my net worth?

 

In fava, go to Balance Sheet, and then instead of "At Cost" select "Converted to <CURRENCY>".

* What is my savings rate?

How do you define it? Before tax, after tax? I use fava-dashboards with a query I wrote myself for this.

 

* What is my FIRE date at the current savings rate?

Again, how do you define this? What are the expected returns of your different assets?

 

* How much have I been spending on Expenses:Restaurants year-over-year?

 You can either run a query, or fava-dashboards comes with a default dashboard that basically does this.

Must I generate bean-queries for everything? Because that rather defeats the purpose of a ledger file in my opinion. I could just as easily list my postings on CSV and write python queries to it. What's more, I don't see how it could be interactive and *fast*? In the absence of a quick feedback loop, what exactly is the tedium of writing out all of my postings going to get me?

Yes, you must generate queries for everything or use a plugin that does this for you. The plugin would have to necessarily be opinionated and make many assumptions. I don't really see why you think this defeats the purpose of a ledger file: a ledger is a book where transactions are recorded, not a crystal ball that predicts what questions you want answered.

I used a Google Sheet before beancount. Beancount gets you many things. You can add properties to transactions. There's many plugins (e.g. beancount-periodic) that do things that would be tedious in a spreadsheet. You can write automatic importers that make best guesses at transaction categories. It's easier to write queries than in a spreadsheet. The data is more structured. You have balance assertions, you can specify tolerances, and a million other features.

Can you do this with a CSV file and Python scripts? Probably, as long as it's computable, Python can compute it. But I think you'll find that you're rewriting chunks of beancount.

As I said, I apologise if I come across a bit cranky but I think that. despite the trove of writing by Martin, I'm missing something fundamental about beancount.

--
You received this message because you are subscribed to the Google Groups "Beancount" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beancount+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/beancount/dcd8ed97-0fca-47a2-965d-dcb30b41a981n%40googlegroups.com.

[Timothy Jesionowski]

You absolutely can filter by time, btw. It's one of the two search boxes on top (I forgot which). You can look at your balance sheet and income statement which is exactly what bean-report would've given you. And when you're looking at either statement, you can click on individual accounts to see their stock (or flow) over time.

The other search box on top lets you filter accounts in or out. So you could get your after tax savings rate by pulling up your income statement, filtering out income taxes, and looking at the expenses. But the better way to do that is which a plugin that renames accounts, because then you can put the income taxes under the Income account and both your income and expense statements will magically be after tax.

Dashboards are where you get fancy, though. And I haven't gotten into dashboards myself because I don't need them. 

Sincerely,

Timothy Jesionowski

To view this discussion visit https://groups.google.com/d/msgid/beancount/CADMOc8Voa9WrCfADrubxrO-CTxhmuM0CxzTvrao65%2Bdiufq5xQ%40mail.gmail.com.

[Chary Ev2geny]

If Fava does not give you all the answers you are looking for, then yes, you need to write queries.

The issue with default reports is that everybody wants to have a slightly different report, hence generic reports will never satisfy everybody. For this reason even in a big multinational company I am working for, the IT team has largely given up producing reports (as they were always blamed for not being able to provide all the needed reports) and instead has given people access to the data warehouse and Tableau to allow everybody to generate own reports.  By dropping bean reports and by promoting beanquery Martin is doing pretty much the same with beancount.

E.g.

* What is my net worth?

I use the following query to get a Net Worth at a specific date, converted to a target currency and grouped by account

    query = f"""

    SELECT account, convert(SUM(position),'{currency}',{date_iso}) as amount

    where date <= {date_iso} AND account ~ 'Assets|Liabilities'

    """

I personally created a Jupiter notebook, where I run different queries in beanquery and then display and visualize the output.

So, I have my Jupyter notebook -based Fava equivalent (I am planning to release it, once it has reached the proper status, but here is an example on how beancount can be use from Jupyter). 

And in this case beancount is extremely powerful. I just don't think there is any other tool, which can provide as much power. 

Surely you can read postings from CSV in pandas dataframe and analyze it there. Also in the same way you can import CSV data in Excel and create pivot tables and reports there. And it will work. But in my experience it only works if you have one bank account in one currency. If you have  anything more complex then that, then it becomes quite difficult. Even if you add one credit card or you lend money to someone, even in this case it becomes not obvious how to do it.

But Imagine you have different bank accounts in different currencies, credit cards, investment accounts. Plus you have mortgage, car, which yearly depreciates in price etc. And then you move founds from one account to another,  you lend money, pay taxes, spend money in different currencies, get paid in different currencies, at the same time exchange rate between currencies constantly changes.  Can you manage this complexity with some simple python script of Excel pivot table? I do not think so.

And here beancount comes into play. As any complexity can be logged into the simple beancount format, checked that it follows double entry accounting rules and then analyzed with beanquery

[Red S]

You’ve received a lot of excellent responses and perspectives already. To add to those, it seems to me like most of the issues you presented either are already solved and are trivial, and you might need just a bit of exploration of the Fava UI to get what you want; or are actually very deep questions requiring custom solutions that you might not realize as such. Specifically:

If I load my ledger into fava, I see some poorly-designed plots not useful for analyses. I don’t get the fuss.
The plot area is too small, the colours don’t contrast well, the tooltips often goes off-screen, or covers the plot so I cannot see where I am.

These seem to be UI issues that I don’t have, and I get annoyed rather easily with such issues. Platform/OS/version info might help. Recording a screen video and showing what you’re seeing would help a lot.

Also, from what I can see, there is no way to limit the plots to just 2023 or 2024 etc.

Type in 2023 or 2024 or ‘2023 - 2024’ into the ‘Time Box’ on the top. It’s very versatile. See here.

The most useful view is the Treemap of Expenses, but even their the text contrast is poor and I can’t seem to generate it year-by-year. The documentation also seems non-existent.

The text contrast is 100% fine for me. I imagine this is an OS or browser setting at
your end. Does it match the screenshot here?

But now what? How can I answer questions like: - What is my net worth?

Ease: trivial.

That’s what the “Balance Sheet” tab already shows in its default view. Pick “Converted to USD”. Or use this link and bookmark it. Replace USD with your currency. Replace ‘main-accounts’ with your title in your ledger:

option "title" "Main accounts"

• What is my savings rate?

Ease: trivial.

Income Statements -> Net Profit (default). Use “Converted to “ and “Yearly”.

• What is my FIRE date at the current savings rate?

Ease: moderate - complex, but there are solutions that make this reasonably easy.

This is a much deeper question than you might realize, involving dozens of assumptions in your mind that no good software can figure out for you. What do you consider your threshold for retirement? Will you earn at the same rate between now and then? What’s your expense profile between now, retirement, and at various milestones? Eg: are there loans/mortgages to be paid off? Major expenses? Unforeseen expenses and how you’ll mitigate for those? Withdrawal pattern in retirement and how it might be based on how well the market is doing? Will you downsize? How will your tax rate and profile change? Assumptions about the market in the next few decades. Etc.

I personally use some scripts with Beancount to feed all my ledger data into a calculator like ficalc.app that includes things I said above

• How much have I been spending on Expenses:Restaurants year-over-year?

Ease: trivial.

Income Statment -> Click on Restaurants in the treemap -> “Changes” -> Pick “Yearly” on the top right -> “Balances (Yearly)”

This is really nice: it gives you stacked bars that instantly let you zero in on problem areas while also giving you an expandable tabular treemap.

Must I generate bean-queries for everything? Because that rather defeats the purpose of a ledger file in my opinion. I could just as easily list my postings on CSV and write python queries to it.

I’d imagine 85-95% of your needs would not need bean-queries or other advanced solutions when you’re starting out. That’s the problem Fava solves. As your needs get more sophisticated, you’ll find yourself solving those. Beancount makes it very easy to write these custom solutions as fundamentally, Beancount is a versatile python library into your double-entry ledger.

See a sample article for example, or Fava-Miler, which are among many things I use Beancount for.

What’s more, I don’t see how it could be interactive and fast? In the absence of a quick feedback loop, what exactly is the tedium of writing out all of my postings going to get me?

Generally, answers everyone wants to see like some that you asked above are trivial. Advanced queries require custom solutions.

Hope that helps! Happy accounting!

​

[Viren Bhanot]

Thank you very much to all of you for taking the time to reply. I guess my first mistake was to think that 'creating the ledger' was 90% of the work done. It appears that that is just the start! From your responses, it is clear that I need to spend more time learning bean-query, and then thinking through all of the answers that I want.

@Chary, the Jupyter notebook might be a great addition as it 'keeps state' aka I can go back up the notebook to see older plots. I'd be interested in when you release it. I agree with you and Adrian that with df + csv I probably would end up recreating beancount + fava anyway.

@Red, thanks for the detailed reply.

Year by Year: Ahh! I'm stupid. It says in big letters "Time". Don't know why my eyes skipped that.

Text contrast: My Fava instance in firefox doesn't match that. The grey is much too light. Screenshot below.

Net Worth: I think I have some errors with how I created my ledger, because the numbers it spits out seem three times too high. Perhaps it has to do with how I handle transfer of money between my different accounts with different currencies. At least I know where to head for this. My net worth should appear as a negative number due to the accounting sign conventions right?

FIRE: I accept. It is indeed a bunch of assumptions that I must sort out. My idea was to get a simplified answer assuming the trinity study, 4% SWR and current savings rate+expenses continuing indefinitely.

Restaurants: Yes! Thank you. Exactly what I was seeking. The Balances (yearly) table + the plot are perfect.

As a general aside, where did you guys learn how to use Fava? I have a preference for linearly-written documentation vs discord ask-around or just aimlessly clicking around playing with things to figure it out. The latter options always make me think I'm not yet getting how the tool behaves.

Thank you all!

[Adrian Utrilla]

I don't know if it's 90%, but I'd say the initial "import" is indeed what takes the most work, especially if as you said you imported 10 years of financial data. Myself, when I got started, I just imported my main asset accounts going back 2 years, and only started categorizing transactions from the initial import onwards, because it's way too much work otherwise.

Your net worth should be positive assuming you have more assets than liabilities, I think. It's your income that should be negative. For "FIRE" (but not really), I use https://github.com/beancount/beanlabs/blob/master/beanlabs/compensation/net-worth-over-time.py, modified with my own targets. It extrapolates from past data and assumes linear growth, so it's not perfect, but it might scratch the itch. You can go further and use an actual FIRE calculator, or projectionlab, or write your own thing.

I learned Fava mainly by using it, and every now and then when I had questions (e.g. time filtering), a Google Search sometimes gave me an answer, but sometimes it didn't. For example, I kind of forgot that the income sheet can give you year-over-year data, and instead I rely on fava-dashboards for it, when in fact fava can just do it natively. Most of the questions I want answered are either very simple (how much did I spend on X this year) or very complicated. For the things that seem simple, I look for a way to do it in fava. For the complicated questions, I don't bother and just write queries myself or look for a plugin.

--

You received this message because you are subscribed to the Google Groups "Beancount" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beancount+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/beancount/a9992310-c8fd-4981-9629-e0c55eec1ccfn%40googlegroups.com.

[Red S]

I learned Fava mainly by using it, and every now and then when I had questions (e.g. time filtering), a Google Search sometimes gave me an answer, but sometimes it didn't. For example, I kind of forgot that the income sheet can give you year-over-year data, and instead I rely on fava-dashboards for it, when in fact fava can just do it natively. Most of the questions I want answered are either very simple (how much did I spend on X this year) or very complicated. For the things that seem simple, I look for a way to do it in fava. For the complicated questions, I don't bother and just write queries myself or look for a plugin.

+1. Same here!

[Red S]

Net Worth: I think I have some errors with how I created my ledger, because the numbers it spits out seem three times too high. Perhaps it has to do with how I handle transfer of money between my different accounts with different currencies. At least I know where to head for this. My net worth should appear as a negative number due to the accounting sign conventions right?

It should be a positive number assuming your net worth is positive. The Balance Sheet page in Fava should help you track down the issue easily: since Net Worth is Assets - Liabilities, simply examine your Assets and Liabilities to figure out where in the tree they don’t match the real world.

​

[Stefano Zacchiroli]

On Sun, Aug 17, 2025 at 09:32:18PM -0700, Viren Bhanot wrote:
> Also, from what I can see, there is no way to limit the plots to just
> 2023 or 2024 etc. The most useful view is the Treemap of Expenses, but
> even their the text contrast is poor and I can't seem to generate it
> year-by-year.

For this, there are two ways:

- From the Income Statement view, select an expense category first (or
not, to see all expenses at once), and then change the frequency to
"Yearly" on the top right. You know have year-by-year view.

- If you are interested in a *specific* year instead, just enter it in
the "Time" filter on the top. This also works for months, for ranges,
etc.

> The documentation also seems non-existent.

I wouldn't say this, but the result is indeed the same: most of the
documentation in the Beancount ecosystem (which is *awesome* in terms of
quality) is nowadays out of date, and hence often useless. The Fava
documentation is pretty good though.

> * What is my net worth?

This is under "Balance sheet" > "Net worth", ready to use.

> * How much have I been spending on Expenses:Restaurants year-over-year?

See above, this is trivial with Fava.

> * What is my FIRE date at the current savings rate?
> * How much have I been spending on Expenses:Restaurants year-over-year?

> Must I generate *bean-queries* for everything? Because that rather
> defeats the purpose.

For everything else, yes, you either need to write queries or use
existing plugins. But I think you are coming to the Beancount project
with the wrong expectations. Beancount is not a full-fledged personal
finance system. It's a basic storage system/"calculator" for your
transactions. What you want should be built *on top of it*, and in part
it has been, in part not.

> As I said, I apologise if I come across a bit cranky but I think that.

Well, I think you could have asked the same questions in a less "cranky"
way indeed. It would have helped :-)

Hope this helps,
Cheers
--
Stefano Zacchiroli - https://upsilon.cc/zack
Full professor of Computer Science, Polytechnic Institute of Paris
Co-founder & CSO Software Heritage

[Martin Blais]

Your initial question does make a good point: in general there isn't as much easy tooling to extract information out of the ledger because many of us have custom scripts and otherwise bean-query is the way to get numbers out of it. I have a script for NW over time for example but it's in the beanlabs repo (see list archive I've shared it before). I should make API documentation (and improvements to the API) so that writing scripts feels more of a first class experience.

--

[Chary Ev2geny]

On Tuesday, August 19, 2025 at 12:53:20 PM UTC+2 bl...@furius.ca wrote:

 many of us have custom scripts and otherwise bean-query is the way to get numbers out of it. 

Martin, just out of curiosity, how do you analyze your ledger?

Do you just run ad-hoc queries in a terminal whenever you need to get some information?

Or do you have some  dashboard - like custom solution?

Or do you have something else?

[Martin Blais]

On Fri, Aug 22, 2025 at 6:38 AM Chary Ev2geny <char...@gmail.com> wrote:

On Tuesday, August 19, 2025 at 12:53:20 PM UTC+2 bl...@furius.ca wrote:

 many of us have custom scripts and otherwise bean-query is the way to get numbers out of it. 

Martin, just out of curiosity, how do you analyze your ledger?

- Individual queries using bean-query

- A bunch of custom scripts, some of which are shared on beanlabs (I use the API a lot, I really think I should document it better)

- Since I got rid of bean-web I sometimes use fava but mostly not; it's great I just never got used to it

TBH I'm really not making the most of the data that's there. 

I feel like there's so, so much more that can be done to pull useful data from it.

But I'm severely time-constrained, trying to grow a small family and on the work front the deluge of machine learning related input is impossible to keep up with (+ the time to put in work itself). I prioritize heavily but Beancount-related improvements rarely make it to the schedule, there's just too much else to do.

 

Do you just run ad-hoc queries in a terminal whenever you need to get some information?

Or do you have some  dashboard - like custom solution?

Or do you have something else?

--
You received this message because you are subscribed to the Google Groups "Beancount" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beancount+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/beancount/2f27499d-7025-4cab-8908-c51d2d0903fen%40googlegroups.com.

[fin]

Viren Bhanot wrote:
> ------=_Part_465578_1855231031.1755491538787
> Content-Type: multipart/alternative;
> boundary="----=_Part_465579_1569482801.1755491538787"
>
> ------=_Part_465579_1569482801.1755491538787
> Content-Type: text/plain; charset="UTF-8"

>
>
> I cannot understand how to actually *use* beancount to do *anything*. From

> what I see, *bean-report* doesn't exist anymore, and certainly not in my

> installation of beancount from AUR on Arch (btw).
>

> If I load my ledger into *fava*, I see some poorly-designed plots not

> useful for analyses. I don't get the fuss. The plot area is too small, the
> colours don't contrast well, the tooltips often goes off-screen, or covers
> the plot so I cannot see where I am. Also, from what I can see, there is no
> way to limit the plots to just 2023 or 2024 etc. The most useful view is
> the Treemap of Expenses, but even their the text contrast is poor and I
> can't seem to generate it year-by-year. The documentation also seems
> non-existent.
>
> The act of actually wrangling ten years of financial data into beancount
> was meditative, and a task I enjoyed thoroughly. But now what? How can I
> answer questions like:
>
> * What is my net worth?

...


hello,

i ran across this one and repeately was cranky about it but
eventually found the solution i needed:

$ alias netw='bean-query ledger.bc "SELECT convert(SUM(position),'\''USD'\'') as amount where account ~ '\''Assets|Liabilities'\''" | tail -1 | cut -d '\'' '\'' -f 2'

which is my bashrc alias for netw, i use this several times
a day when i'm actively doing any beancount updates or i'm
trading.


it can be simplified to:

$ bean-query ledger.bc "SELECT convert(SUM(position),'\''USD'\'') as amount where account ~ '\''Assets|Liabilities'\''"


the other one i use frequently are the current balances query:

$ alias bal='bean-query \-f text ledger.bc "SELECT account, units(sum(position)) GROUP BY 1 ORDER BY 1" '

from that i can grep for the specific item like:

$ bal | grep PnL | grep <symbol>


i don't use fava as i don't like point and click types of
interfaces unless i really have to.

as you've found out you can do range of date queries there
but i don't look at those kinds of things very often.

and for FIRE i got fired long before that became popular.
Sept 1, 1996. almost 30yrs ago already! eek! time sure goes
by quickly.


fin

[Martin Blais]

  and for FIRE i got fired long before that became popular.
Sept 1, 1996.  almost 30yrs ago already!  eek!  time sure goes
by quickly.

Congratulations for getting off the treadmill!

[fin]

Martin Blais wrote:
> --00000000000003efcc063d07e868
> Content-Type: text/plain; charset="UTF-8"

it was the best thing i ever did. i enjoyed my 15 years
of computer related work, but when i was done i was very
well done and rather crispy around the edges. :)

since then i've mostly been able to do what i want and
hit retirement age this past year so could start drawing
SS and that has helped a great deal until i get to 65 and
will start drawing on other retirement accounts.

i'm not super wealthy and it's not been easy but i'm
glad and lucky to be living where i am and in the
circumstances i have compared to so many around the world.

now i mostly spend my time gardening, reading and
fiddling around with computer things - nothing
professional. i was working part-time at a library and
i really enjoyed that until management changed...

my priority now is taking care of Mom and keeping things
up here as best i can. i'm not physically 100% but i can
get by. i'd probably be dead already if i weren't doing
some gardening to keep me moving and that is what makes
my back happy.

the website which describes the place and the gardens
and various things is at www.anthive.com...


fin

[Chary Ev2geny]

On Saturday, August 23, 2025 at 2:38:32 AM UTC+2 bl...@furius.ca wrote:

TBH I'm really not making the most of the data that's there. 

I feel like there's so, so much more that can be done to pull useful data from it.

Can you may be share your idea of what can be pulled, but is not being done? May be others can develop something?

 

But I'm severely time-constrained, trying to grow a small family and on the work front the deluge of machine learning related input is impossible to keep up with (+ the time to put in work itself). I prioritize heavily but Beancount-related improvements rarely make it to the schedule, there's just too much else to do.

It normally gets a bit better with the time availability when kids grow. But for now you have build a very solid foundation with the quality of  your code and, which seems to be much more rear on the open source hobby projects, with the quality of the documentation,  so, the community can carry on.

 

[Martin Blais]

On Tue, Aug 26, 2025 at 4:30 AM Chary Ev2geny <char...@gmail.com> wrote:

On Saturday, August 23, 2025 at 2:38:32 AM UTC+2 bl...@furius.ca wrote:

TBH I'm really not making the most of the data that's there. 

I feel like there's so, so much more that can be done to pull useful data from it.

Can you may be share your idea of what can be pulled, but is not being done? May be others can develop something?

Well there are many things, making pivot tables tables (e.g., expenses per period), parallel tracking of mirror accounts (like IRA$ allocated being used up), make currency conversions easier, make it possible to combine multiple ledgers into one, etc. all previous ideas we've discussed on the lsit.

But I think the area that stands to have the most impact is to come up with a really simple, generic and clean way to convert between tables (e.g., spreadsheets) and Beancount data with as little specification as possible. I do have a few places where I'm doing this, but I think this can be made more generic and commonly used. It provides a natural channel for people who aren't users to participate.

But I'm severely time-constrained, trying to grow a small family and on the work front the deluge of machine learning related input is impossible to keep up with (+ the time to put in work itself). I prioritize heavily but Beancount-related improvements rarely make it to the schedule, there's just too much else to do.

It normally gets a bit better with the time availability when kids grow. But for now you have build a very solid foundation with the quality of  your code and, which seems to be much more rear on the open source hobby projects, with the quality of the documentation,  so, the community can carry on.

That's nice of you to say, thank you; but consider this: at the current pace of change we'll soon be able to say something like "Hey Gemini, please convert this entire project as close as possible to its original form but translated to Rust" and the output will be near identical but faster and type-checked. We're almost there... this *will* happen in the next ten years I think. So I suspect once humans start to leverage the models a lot more than they are now (most people are still operating on what they know), the nature of software itself will start to shift a bit. I predict a faster pace and more chaos. It's unclear as of yet whether the faster pace will win over the extra chaos (I've learned never to underestimate complexity so I really don't know).

[Chary Ev2geny]

On Thursday, August 28, 2025 at 1:47:30 AM UTC+2 bl...@furius.ca wrote:

 consider this: at the current pace of change we'll soon be able to say something like "Hey Gemini, please convert this entire project as close as possible to its original form but translated to Rust" and the output will be near identical but faster and type-checked. We're almost there... this *will* happen in the next ten years I think. So I suspect once humans start to leverage the models a lot more than they are now (most people are still operating on what they know), the nature of software itself will start to shift a bit. I predict a faster pace and more chaos. It's unclear as of yet whether the faster pace will win over the extra chaos (I've learned never to underestimate complexity so I really don't know).

We all will be leaving in a totally different world then, with 20% chance of AI eliminating humans, in accordance to Elon Musk. So may be migrating beancount to rust will not be the most pressing issue even for this small community.    

On a more close future,  I am waiting for the sate of AI development when we will able to ask AI to produce an updated documentation for beanquery, using original source code original documentation and updated source code as an input

 

[Simon Michael]

On 2025-08-19 07:46, Viren Bhanot wrote:
> Thank you very much to all of you for taking the time to reply. I guess
> my first mistake was to think that 'creating the ledger' was 90% of the
> work done. It appears that that is just the start! From your responses,
> it is clear that I need to spend more time learning bean-query, and then
> thinking through all of the answers that I want.

Hi Viren, just for the record I enjoyed your very questions and the
"cranky" level was just right - not offensive, very pertinent. It's
important to ask ourselves critical questions sometimes.

You're right that recording the journal is just the start. Accounting
and financial analysis even for an individual is an ongoing process.

I think for Beancount and all other PTA tools (maybe even all accounting
apps ?) it's a slightly unfortunate truth that most of our energy gets
used up just learning the tools, entering and managing and refining the
data, running a few reports, and there isn't much left over to progress
to more advanced analysis - or, that comes slowly, only after years of
steady use - and the solutions are custom to each person.

[Simon Michael]

On 2025-08-19 07:46, Viren Bhanot wrote:

> Thank you very much to all of you for taking the time to reply. I guess
> my first mistake was to think that 'creating the ledger' was 90% of the
> work done. It appears that that is just the start! From your responses,
> it is clear that I need to spend more time learning bean-query, and then
> thinking through all of the answers that I want.

Hi Viren,

Just for the record, I appreciated your post and the "cranky" level was

just right - not offensive, very pertinent. It's important to ask

ourselves critical questions.

You're right that recording the journal is just the start. Accounting
and financial analysis even for an individual is an ongoing process.

I think for Beancount and all other PTA tools (maybe even all accounting

apps ?) it's a slightly unfortunate truth that most of our time and
energy is used up in learning the tools, entering and managing and
improving the data, and running a few reports. Usually there isn't much
left over for more advanced analysis - or, that comes slowly, built up
over years. And then the solutions tend to be custom for each person.

Collecting concrete reporting questions like yours, and gradually
developing best practices and tools for answering them, is a good way to
improve things.

-Simon