Installing beanount on mac

[Jason Crews (the1gofer)]

Hello Everyone, 

I've tried installing bean count on a m1 macbook, and a m2 mac mini, and haven't had much luck.  The documentation seems pretty old, so I'm not sure what to try. 

I tried installing with 

sudo -H python3 -m pip install beancount  

and 

sudo -H python3 -m pip install git+https://github.com/beancount/beancount#egg=beancount

which didn't work both giving me the error message /Library/Frameworks/Python.framework/Versions/3.8/bin/python3: No module named pip

I then tried 

sudo -H python3.9 -m pip install beancount  

which seemed to finsih without errors.  I'm even able to use bean-check.  However, using bean-report --help or web-bean both give the follwoing

zsh: command not found: bean-report

and

zsh: command not found: web-bean

Googling either of those has no results. 

I also tried brew install beancount, even though that's not in the documentation, and nothing changed. 

Any suggestions?   I've been experimenting with ledger and like it, but would like the better features found in bean count.

[Petr Beránek]

Hi,

I wouldn't recommend to install python packages via sudo, as it may break system libraries. For installing applications like Beancount, I recommend pipx: https://pipx.pypa.io/stable/. Pipx installs every application under your regular user (no need for sudo) into a separate virtual environment. This is especially useful, if you decide to install Fava, which depends on an older version of Beancount. Follow installation instructions from the pipx docs, then run:

pipx install beancount

pipx list  # this should list Beancount and its tools like bean-check etc.

Best regards,

Petr

--
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 on the web visit https://groups.google.com/d/msgid/beancount/2f077dec-2064-4d31-bf90-a3aec6b94deen%40googlegroups.com.

[Martin Blais]

bean-report and bean-web are gone.

Deprecated, removed from the 3.x line.

(Search the mailing-list)

--

[Jason Crews (the1gofer)]

Is there a more recent tutorial?

Jason Crews

You received this message because you are subscribed to a topic in the Google Groups "Beancount" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beancount/qWXjk2vb5UI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beancount+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/beancount/CAK21%2BhNiLog7w-vjJvBaXP3FXWPZ%3DLy%3DbHWSz-%3D_oCKCQ4ygrA%40mail.gmail.com.

[Daniele Nicolodi]

On 13/10/24 01:47, Jason Crews (the1gofer) wrote:
> Is there a more recent tutorial?

No one contributed one so far.

Cheers,
Dan

[Jason Crews (the1gofer)]

How do I generate reports now? All documentation points to bean report. at least that I can find.

[Mariano Draghi]

The successor to bean-report is beanquery.

At one point (with v2) it was part of Beancount, but now it's an independent tool. It's much more powerful and flexible than bean-report.

You can install beanquery from here:

https://pypi.org/project/beanquery/

The problem is, the only documentation that exists is from when Martin was experimenting with the idea and publishing the first versions/drafts:

https://beancount.github.io/docs/beancount_query_language.html

Most of what is written there is still valid, but beanquery has grown quite a bit since then, and the documentation for all the new features is basically non-existent, unless you start digging in this mailing list and in GitHub issues.

The release of Beancount v3 in its current form somehow made things significantly more difficult for new users like you, because there are lots of things that are not documented, or have incomplete documentation. The maintainers of each subproject don't have much time these days, and we as a community have not stepped in to help with that either.

So IMO you have two options:

a) keep using beancount v3, install beanquery, and learn a bit of BQL (Beancount Qurey Language) from the doc I linked above, and also searching here in the mailing this. This is the officially recommended path. Beancount v3 and Beanquery are the future, anyway. But right now there are several rough edges, and the learning curve is a bit steeper than it should, mainly because the lack of documentation and examples

b) install beancount v2, which includes bean-report (and the first version of beanquery build-in), use that, and revisit the v3 ecosystem in a few months. If you go down this path, please be aware that beancount v2 is EOL, so to speak, and it won't receive any new features or improvements at this point. But it still works perfectly fine.

I'm still using Beancount v2 because I can't use Beancount without Fava. And yes, I know it's technically possible (and not that difficult) to install Beancount v3 and Fava in separate virtual environments sharing the same files, but it's not that Beancount v3 is receiving lots and lots of features and is in active development either, so there's really no incentive for me to change my perfectly working workflow with Beancount v2 + Fava. 

Hope that helps!

[Red S]

Great summary, very valuable for new users, thank you Mariano. Perhaps you could paste your post just as it is (or even simply link to it) here. The source is here which you could ask Martin for edit access to.

​

[Jason Crews (the1gofer)]

I made some minor updates to include pipx. Hope that helps a little. 

On Saturday, October 12, 2024 at 2:34:35 PM UTC-7 bl...@furius.ca wrote: