PlainBudget – Minimalist Plain Text Budgeting | Hacker News Reader

PlainBudget – Minimalist Plain Text Budgeting

plainbudget.com

96 pointsjgalvez31 comments

I love budgeting and I love plaintext, doing more with textfiles than most people. I'm going to give some quick feedback as someone that should be in the target audience, but doesn't find this particular application interesting.

* Setting the budget is one thing, but you never actually spend exactly the budget. There is no way import transactions which is crucial. It's a good habit to compare actual expenditures to projected budgets.

* One of the major problems that successful budgeters run across is how to track savings over time. Savings are shifting in nature. In bank accounts or bonds the interest the occur is predetermined. What if the savings is in stock, crypto, precious metals or collectables? It's a shame that my tool for tracking the value of my retro game collection doesn't communicate with my tool for tracking my savings accounts.

* I'm concerned about using your GUI tool to edit a text file instead of my own primary text editor. The nice thing about my text editor is that my font is set how I like, my keybindings are set how I like, and I know a ton of useful shortcuts. I'm not sure I'd want to spend the time to configure your tool, and even if they have the configuration options I'd want.

* Honestly there are a ton of budgeting apps with more features. Some good ones are even free.

Thank you so much for the feedback — this is a very early iteration.

I've been using it for several years and wanted to get it into an app. I've been getting a lot of suggestions and requests from early adopters. I will definitely take yours into consideration. There's a lot of room for improvements.

Re: GUI limitation, there's a CLI which is open source and free:

https://github.com/galvez/plainbudget

The parser and processor have a comprehensive test suite.

npm i pbudget -g

I talk about the DSL design and its capabilities in detail here:

https://hire.jonasgalvez.com.br/2025/may/8/plainbudget/

I'm sure you're aware of the plaintext accounting/budgeting options, but for others that might be interested, here's a list: https://plaintextaccounting.org/#pta-apps

I personally use hledger, but beancount is also popular.

This is interesting. I really like the DSL. It is a little limited though, just because it works on the month-level only.

One thing that I think is missing is having different dates for transactions to occur. So like some things happen just once ever, or transactions on the last day of the month which shifts.

This just means the stats that are given by this app are a bit rough. Looking at the source it seems to estimate the projections. Not a bad thing, just something to note.

I recently translated my own script-based cash flow simulator to a web app as well, which might be interesting: https://nicktrevino.com/tools.cashflow-simulator.html

One last thing, if you haven't heard of Wails, and you like Golang, I recommend it when thinking about making a desktop version of a web app: https://wails.io/

Pity this doesn't follow the PTA (Plain Text Accounting) "standard" but invented its own DSL.

https://plaintextaccounting.org/

Especially because there is a wide range of apps, ranging from CLI to web and even mobile, which can be used for PTA and due to the fact that they acknowledge the same standard are interoperable:

https://plaintextaccounting.org/#software

The most well known ones being

- hledger (https://hledger.org/),

- beancount (https://beancount.github.io/),

and

- ledger (https://ledger-cli.org/)

Important note: All free and open source

I was just checking it out.

         $2145.00  Assets:Checking
        $-1000.00  Equity:OpeningBalances
          $192.00  Expenses
           $65.00    Food:Groceries
          $127.00    Unknown
        $-1337.00  Income:Salary
  ---------------
               0

Nice, I missed that. I'll consider making it PTA-compatible!

Only issue I have is with the indentation-based subitems.

I'd rather keep it restricted to top-level groups for clarity.

PTA is more or less a collection of formats. Beancount, for example, uses a colon-separated account name per transaction line, such as "Income:Work:PrimaryJob:HolidayAllowance" (excluding the quotes). There are a couple of fixed main accounts (such as "Income"), and the user can arbitrarily define subaccounts (in this example "Work:PrimaryJob:HolidayAllowance").

Although I see your point in wanting clarity, if the format supports arbitrary subaccounts, you can use the tool itself to limiting the number of subaccounts to show (e.g. with 1 you would only see the accumulated amount in all accounts in the "Income:Work" hierarchy). This still fits your use case for clarity, but allows other users to expand the budgeting tool to their needs.

I'm interested but I have more questions than answers.

It might make sense to provide some system for figuring out how to create the budget, and then how to track your expenses against it.

In my mind the simplest form of budgeting is so-called "Envelope Accounting", where you have physical envelopes full of money where you pull money out when you spend it.

There are electronic envelope accounting systems which retain the simplicity.

I'm a bit unsure of how this system works, where it differs, etc.

In summary, I'd love to see:

1. An explanation of how to construct a budget using your system.

2. An explanation of how to compare your spending to your budget? (bonus points if I can use my existing plain text accounting system, or at least a csv file)

3. An explanation of how to track my budget over time.

The processor considers monthly income when calculating your projections. It'll sum all credit and debit operations (from all flows) and calculate from there.

> In my mind the simplest form of budgeting is so-called "Envelope Accounting", where you have physical envelopes full of money where you pull money out when you spend it.

I've done that, it helped me a lot!

I read a fair bit of criticism here in the comments. I think I understand where they are coming from. From my point of view, I like the simplicity of the DSL. I also love long-term projects. Great to see that you are working on it since 2018! And I think, from a western perspective, 10 bucks is a fair price to support you.

Some observations and bugs I encountered:

- A license added to Github: Currently, I don't know if I'm allowed to build an application using the package, say an integration to Obsidian or whatever.

- The example in your readme features the wrong group name ("Housing" vs "Main")

- Your DSL does not like comma-values. :)

- The DSL also does not like to see groups being used multiple times. So, what I wanted to try was:

  = Housing
  - 2000 Rent
  - 1000 Utilities
  - 500 Leisure
  
  # 2025-05-09
  = Groceries
  - 10 Coffee x 12
  - 10 Milk x 12
  - 20 Cereal x 6
  
  # 2025-05-10
  = Groceries
  - 2,60 Bread
  
  = Income
  - 5000 Salary
  - 1000 Side hustle

  + Income
  - Housing
  - Groceries

- The CLI will crash if it does not find a file. A better way to handle that would be to create an empty file.

- You are using Electron, so a Windows build would, of course, be nice.

I'll push a new release with the license — always forget this!

Thank you for all your observations. Two things: 1) floats were originally purposefully not allowed, for simplicity, but I might change that and 2) multiple groups with the same name are indeed not allowed, but you raise a good point: it should be possible to reference all groups with a common label if the only difference between them is an attached timestamp.

And yes, will definitely push Win/Linux releases soon!

Great to hear! Maybe, as someone pointed out in another comment, it might make sense to show how you use your tool yourself in a dedicated blog entry, and link that in the readme. People usually learn very well from examples. :)

I guess it's obvious that a plain spreadsheet does this job nicely. I use one row per day and one column for each account, e.g. rent, utilities, each credit card, savings accts etc. I make a new tab each year. Very easy to see in advance the low point in your bank balance each month/pay period.

Just wanted to give 2 cents on the rationale behind this:

Most people don't budget at all.

Most people live paycheck to paycheck. I have lived paycheck to paycheck for a long time. When your finances are in distress, and you have no planning, the least helpful thing you can do is trying a fully featured app, or even worse, a spreadsheet. It requires a lot of attention to details and they can get overwhelming and confusing to manage — not everyone is fluent in spreadsheet formulas, as basic as they may be, to the point of getting the setup right, and most template spreadsheets available pack a lot of unnecessary things and customizing them becomes a project of its own. There's also the feeling of despair realizing you're not even close to even having that much data or assets to put in. This is not only my opinion — this is what I've gathered asking friends and family on the topic.

So this is a way to keep things simple. Extremely simple. No integration to banks, no mental overhead, just a smart replacement for a piece of paper where you write down your income and your expenses. There is a lot of people that still use a physical notebook to keep track of their finances, in this day and age, out of sheer choice — it's a way to maintain focus on the big picture, and not miss any detail.

It's a way to express your finances in a portable, human-readable format that is essentially computable plain text. You can express your finances at a 10,000-feet level, know what your savings will look like, for motivation, and know where your money is going. This is it. Surely it will be too simple for many, but perhaps just about right for some.

You're also not vendor locked, your data is plain text and you can use the CLI (free and open source) to process it. I myself use the CLI and manage my sheets from Sublime Text. Yep :)

Fair enough and I can see how this will help people a lot. I was just pushing what has worked well for me.

the first level of budgeting is to track your expenses against your income. you look at whether your bank account is growing or not.

then you categorize expenses to get an understanding of your spending.

only after you have those categories, you can start to budget.

you can look at the categories and consider whether you are spending to much on a category or not.

if your cash flow is negative you need to find categories that you can reduce. likewise if you want to save money for future events.

i'd like an expense tracker that helps me categorize my expenses. your app seems useful only after i have done that.

the workflow i am looking for is:

i enter expenses as i make them, one line at a time. including breaking out groceries into individual items. then it should allow me to group items into categories. tracking which items don't have a group yet. finally it should tell me how much i am spending per group each month.

I'm not sure that level of detail (specifying receipts to the article level) helps you get a grasp of your overall spending, nor do I expect it giving the "right advice". Suppose you buy a lot of healthy fresh fruit and vegetables, would you want that tool to suggest you reduce that spending?

the point is to group articles by category. a single visit to a supermarket may contain important and vanity items mixed together. i want to break that out and then categorize accordingly. i do want to know how much i am spending on candy or sugar drinks, or, if you are so inclined, cigarettes or alcohol. toys, gifts, school supplies are all different categories that are worth tracking. if you go shopping at places like target or walmart you'll get all of that mixed up on a single receipt.

fruits and vegetables don't need their own category. they could be grouped into food but candy bars or soda are not food.

i don't want the tool to suggest anything, but i want the numbers so i can draw my own conclusions.

entering data by article is easier than categorizing on the spot. i could just scan the receipt and let the tool work out the categories (once i assigned a category to an item, the tool can do it for any repeat appearance)

I suppose a tool like Beancount (or (h)ledger, but I only have experience with beancount) and bean-web would fulfill your needs. The latter offers HTML reports per arbitrary subaccount, based on the data entered in the beancount file (which is plain text).

I use Soulver with a file for each month’s expenses as well as a file for overall assets and liabilities. If you already have that app, it’s a great solution. It supports (global) variables and has natural language support.

https://soulver.app/

It’s not plain text, but it feels like plain text.

This was my thought instantly. It looks like a worse Soulver with fewer features and not many benefits.

I am confused why you would want to bother with plaintext and go out of your way to use a limiting editor to do that.

Ah well the macOS utility app market is weird.

This is great. Having been in and out of paycheck-to-paycheck living several times, I find that during transitions to stability individual transaction tracking is useful for understanding what’s even happening to you financially. But once you understand that do you really still want to micromanage spreadsheets every day of your life? Would much rather leave my accounting rules on autopilot and base my spending decisions off of a tool like this instead.

Tried out the Mac app snd didn’t encounter any obvious bugs. looking forward seeing some functionality around annual expenses and potentially dating entries to create some sort of anticipated distribution across a date range

I also wrote a small napkin buget calc script with my Python `reaktiv` library so I can use REPL and commit it to GIT. Here is a small example: https://github.com/buiapp/reaktiv/blob/main/examples/napkin_...

This is great. I’m gonna try it out but just looking at screenshots, this was really close to something I was going to try building not that long ago. I’m glad someone else did because I didn’t/don’t have the time right now (and don’t need another project for my ADD brain to hyper focus on )

The wording on the site/blog was confusing. The homepage says "Download Beta" and the blog says "purchase". Sure enough, when you click "Download Beta" you're asked to pay $10.

I've updated the button to reflect reality.

I love this, could you add a licence for the cli code?

Done.

I expected ncurses :)

It’s wild to me that people would rather build their own apps than use excel.

I did a budget using Google Sheets and it’s not a bad way to go. The other hand is, it’s nice to have a starting point.

Reply on news.ycombinator.com