> I got Emacs to work with the beancount.el file. It nicely color-codes
> everything. However, if I activate org mode the colors disappear. I suspect
> this is incorrect behavior and there must be something I'm doing wrong.
No, it's expected behavior.
> Any suggestions of what might not be
>> I finally got around to merging the patches I used to use with the old
>> (minor-mode) version of beancount.el into the new version.
> Thank you very much for the patch. Do you mind if I split it in a few
> commits and I submit it with a few tweaks and your attribution as a
> merge request on
I finally got around to merging the patches I used to use with the old
(minor-mode) version of beancount.el into the new version.
Most of the tweaks I was using aren't needed any more so all that's left
is the patch below:
- Remove the redundant `:group`s (since they default to the last
>> > BTW, as a quick note, purely textual include should be doable with
>> > a simple wrapper script that invokes the c++ preprocessor frontend.
>> Wouldn't that mess up line numbers and error messages?
> Ahh yes. Beancount would have to honor the #line directives.
There can be other surprises
> I would like to upstream my changes, however, now that I know how I
> would like the code to look like at the end, I would need to break up my
> changes in a patchset and submit if for inclusion. This would require
> some time investment. I haven't seen any interested from Martin about
> those
> mode into a major mode and use outline-minor-mode to preserve the
> section folding functionality used by some. I adapted some code from
> outshine-mode to allow headlines visibility cycling ala org-mode.
Looks great, thank you.
Could you submit `outline-cycle` as a patch to outline.el
(via M-x
> I renamed the two variables that control alignment. Now they are
> beancount-transaction-indent and beancount-number-alignment-column.
I recommend you preserve backward compatibility with
`define-obsolete-variable-alias` (make sure you call it *before* you
declare the new var name).
> problem you were seeing. What I would like to do now is to get rid of
> the need for beancount-init-accounts
Yes, please.
> but I am afraid completion would slow down considerably in very
> large buffers.
If needed, judicious use of caching should do the trick.
Stefan
--
You
cchiroli" wrote:
>
> Hi Stefan,
>
>
> On Sun, Jan 06, 2019 at 12:37:14PM -0500, Stefan Monnier wrote:
>> AFAIK, this is because your major mode is org-mode which does its own
>> thing with TAB.
>
> Right.
>
> Do you have any advice (and/or related
> - completion: I'm not sure how that is supposed to work. beancount-tab
> as a function works well for accounts and tags (but not metadata
> key/values, apparently), but if I bind it to TAB I lose indentation,
> whereas in ledger-mode the two seems to work well together. Maybe I'm
> not
> Yes, it's v2. No strong opinions about moving it to v3, I just don't
> have time to review the implications.
FWIW, long earlier discussions and analysis of the problem of
"GPLv2-only for Linux" concluded that the GPL is one of the very rare
licenses which doesn't *automatically/implicitly* say
> I've heard about the GNU prescription but in practice I've witnessed other
> packages using the C-c key and it seems like it depends a bit on how
> ominously package implementors have taken the word of GNU to the letter :-)
Note that the issue is not with "the C-c key". It's with "C-c
", where
latest update, version 8.6, was released on September
17th, 2016.
so it seems 3.5 didn't even exist when the last `stable` was released
(updates to a release focus mainly on security bugs). So "huge lag"
sounds a bit harsh.
Stefan
> On Wed, Oct 26, 2016 at 9:4
> However, this would coerce everyone up to Python 3.5.
That's annoying on Debian stable (about half my machines, including the
one which I use most often with Beancount) which is still at 3.4.
Stefan
--
You received this message because you are subscribed to the Google Groups
14 matches
Mail list logo
