Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Accounting for computer scientists (2011)

martin.kleppmann.com

Post reply on HN
Print view View original

11–20 of 51 posts

Re: Accounting for computer scientists (2011)

#11

Post by gavinhoward »

I learned the basics of accounting from the GnuCash guide. [1]

One of the best things I ever did.

If the featured article helps you understand accounting, good! If it's something else, that's fine too.

I agree with an HN comment [2] from long ago: learning the basics of accounting is a superpower. And not just when talking to finance people.

I do my budget in double-entry. (Can you guess my program of choice?) It allows me to budget across long periods of time and to more easily focus where my wife and I are spending money.

Learn accounting. It may suck, but you'll be glad you did.

[1]: https://www.gnucash.org/viewdoc.phtml?rev=5&lang=C&doc=guide

[2]: https://news.ycombinator.com/item?id=32497570

Re: Accounting for computer scientists (2011)

#13

Post by seltzered_ »

Outside of the concepts, one of the biggest life organization lessons I learned was from the user guide of beancount (a plaintext accounting software), particularly for managing finances:

- Create folders for Assets, Liabilities, Expenses, and if applicable Equity

- Organize files starting by ISO date YYYY-MM-DD.mortgageStatement.pdf

https://docs.google.com/document/d/1wAMVrKIA2qtRGmoVDSUBJGmY... talks about this.

You can use a file renaming tool (like `qmv`, built into GNU's fileutils) to help with getting started. This has helped myself get organized and help in organizational settings too.

Re: Accounting for computer scientists (2011)

#14

Post by EvanAnderson »

gavinhoward wrote: post #11

I learned the basics of accounting from the GnuCash guide. [1] One of the best things I ever did. If the featured article helps you understand accounting, good! If it's something else, that's fine too. I agree with an HN comment [2] from long ago: learning the basics of accounting is a superpower. And not just when talking to finance people. I do my budget in double-entry. (Can you guess my program of choice?) It all…

I appreciate the call-out.

I agree w/ you re: the benefits for personal finance, too.

For everybody else: If your work at all touches accounting of finance learn basic bookkeeping. It'll pay tremendous dividends.

Re: Accounting for computer scientists (2011)

#16

Post by rendaw »

I'm eternally interested in learning more about accounting, but I think this is missing crucial information to actually explain what "accounting" is while laying on this patronizing anti-intellectual "accountants use hard words" writing style.

1. What are you trying to solve by "doing accounting"? After reading this article, it seems like it's some sort of arbitrary aesthetic, like people who think the point of organizing is to have no more papers on your desk.

2. How do the actions in this guide solve those questions? Why make a graph? Why make extra buckets? Why color them? Relatedly, I couldn't find an answer to "why is double entry bookkeeping revolutionary" despite spending a fair amount of time searching. I think it's either related to manual accounting (i.e. not relevant to computers) or is directly related to this question.

I appreciate the mapping of accounting terms to graph features, but the title is way too grand for the contents. I just noticed another commenter who seems to know a lot more than me says roughly the same (although much more negatively).

In addition, here are some things I'd like to know (in addition to the above, although I have guesses to the answers to some):

1. How to deal with accounting over time? When received payments and made payments are at different dates, amortization, etc. If you're trying to identify missed payments, time disconnects create permanent gaps which may or may not be an issue.

2. How to deal with multiple currencies. Related to variance over time: exchange rates, balance calculation, etc.

I feel like "make a graph" is the easy 5m answer and the first thing you arrive at if you look up double entry bookkeeping.

Lastly, I agree that accounting is a fundamental skill, but before you can do accounting you need _data_. All transactions, categorized, all accounts, all balances. This is the hard part. If you're an accountant and this is your job, okay. If I need to manage N accounts in my time after work, well that's a significant chunk of my non-work time.

Mint doesn't work well outside major western spheres. For a while I maintained my own scrapers, but I've fallen back to "just spend as little as possible and you'll avoid going negative" which is not where I want to be but until there's open banking here I don't see many options.

Re: Accounting for computer scientists (2011)

#17

Post by daft_pink »

I feel this doesn't explain double entry bookkeeping well enough.

The connectedness of double entry bookkeeping and the way that it links the balance sheet to the PNL is what makes accounting powerful and accurate.

The fact that you can look at an accounting balance sheet and if it matches what exists in reality, you can somewhat trust the profit and loss statement is a very powerful concept.

Anyone who's going to write accounting programs really needs to focus on this concept.

Re: Accounting for computer scientists (2011)

#18

Post by hobs »

xyzzy_plugh wrote: post #8

Kleppmann knocks it out of the park here. Most accounting is actually insane. What should be an immutable ledger is actually the farthest thing from. A few years ago I had a gig writing some custom reporting for a business that was struggling to get reproducible reports (i.e. I ran this report yesterday and now I can never produce the same report again) and it turned out their accounting practices were the root of th…

Soooo many people do this. Temporal tables can help to see dumb stuff happening.

Re: Accounting for computer scientists (2011)

#19

Post by krackers »

ghewgill wrote: post #3
dave333 wrote: post #2

I've always thought that accounting is arithmetic made difficult, but this helps a lot.

It's worth remembering that accounting was developed before negative numbers.

>I've always thought that accounting is arithmetic made difficult

It basically is. The fact that operations like "debit" and "credit" behave differently based on the sign-convention of the account is legacy baggage that should have been done away with a literal millennium ago. The whole situation is as absurd as if physicists decided that the concept of a "negative" voltage was a bad thing so KVL was formulated as sum{delta_sources} = sum{delta_loads} instead of just sum{delta_voltage} = 0, and then people had to memorize under which conditions circuit elements were treated as sources instead of loads, and when anyone questions why it's all so confusing they just get told that's the way things have always been done.

Re: Accounting for computer scientists (2011)

#20

Post by AlbertCory »

I was briefly in a night MBA program at San Jose State (until I decided there was no point to that).

However, the Accounting and Finance courses were totally worth it. We're playing the game of Business (even in a non-profit), and knowing how they keep score is immensely valuable.

Post reply on HN