Hacker News Reader

Comment by ancarda | Hacker News Reader

I'm sure this new website is a big improvement, but to me (not an Emacs or Org Mode user), it seems Org Mode does... everything? Kinda like Emacs?

Like, that spreadsheet example seems clunky to me, why would I want to use this?

View on HN

In the first appoximation you can use org-mode the same way you'd use markdown -- merely for the syntax. But indeed, it shines when you use it within Emacs because you can extend it. Even the 'vanilla' setup is very powerful -- you get the markup, datetimes, todos, scheduling, agenda, clocking, literate programming, etc.

Spreadsheets.. personally, never had the need to use formulas, but they are very nice when you need.. well, to organize information in tables. For example, I'm using it with org-capture to track sleep data straight into the table, which results in a table row for each day (with an auto-generated timestamp).

> For example, I'm using it with org-capture to track sleep data straight into the table, which results in a table row for each day (with an auto-generated timestamp).

I'm curious how one would do that. Would you mind sharing your capture template?

My capture templates are here https://github.com/karlicoss/dotemacs/blob/d3d26d4315868298b...

, but sleep logging is in a 'private' config for some reason, so I'll just paste it here:

  (setq --my/extra-capture-templates
      '(("s"
         "sleep"
         table-line
         (file+headline "/path/to/capture.org" "sleep table")
         "| %U | %? | | | |")))

"sleep table" is the heading, under which the table is contained. So:

- I press the capture hotkey (F4 for me)

- then 's' which prompts me for a table line

- %U automatically substitutes the timestamp

- the cursor is placed at %? (second column), so I can just log whatever I need to, tab for the next column, and then finish capture

What’s the advantage to this over a csv in excel?

I track my blood pressure similarly. The entry is simple and I can read the csv into R/python whenever I want to visualize it.

You can keep it all in one file, along with the code and results of its execution.

That org capture template will feed new data into a table; below it, in the same .org file, you can keep your R code and your Python code and whatever other code you prototype (I often write Elisp code), and you can execute all that code inside the .org file.

This approach is not giving you any new features, it's just optimizing for convenience. Think of .org files as Jupyter notebooks on steroids :).

I was going to say, sounds like Jupyter. Is there much of an advantage I'm missing out on as someone who prefers GUIs over the terminal and vim/emacs?

Org lets you do that all in one place :)

I do not think the intention is at all to replace spreadsheets, if that is the concern. More, this is a great way to move some data through a series of exploratory questions without having to push the data through many other tools.

In large, it is Emacs taken to an extreme. With the major feature point that all cards are on the table. Meaning the file you get at the end, is just plain text. There is no need to ever worry about export of data being locked out from you. Yes, using it is currently best exemplified with Emacs. No, nothing is forcing that.

That is correct and it is a pity other editors are so far behind in all things org. It means that for people, who have not tried it and gotten used to it, there is a double obstacle to take and most will miss out on it for that reason.

I use spreadsheets in org when:

- teaching classes and grading students - working on a quote / invoice for a customer (I add numeric properties to an org item, org gather them in a table, and I add columns to that to get sums, taxes, discounts, etc) - designing my woodworking and other building stuff, for cutlist and so on

I am not the kind of guy that lives in emacs, I only use it for org mode, in a way that resembles a scientific notebook, mixing text, data and code.

I would not force it onto anyone, but using anything else feels backwards.

Spreadsheets can be surprisingly useful in a text document. For example, I regularly need to grade programming assignments. I do this in.a structured text document, with sub-headings per student and category, where I take notes about their work, and assign points.

So far, this could be easily done in markdown, or even plain text. But then I tally up the points, and convert them to a grade based on some simple equation in a table. My colleagues need to do this in a separate Excel sheet. I can do it right in my org document.

I do the same, and I email the student my notes, by calling the org mime function.

Org mode seems to be like Markdown in spirit - sometimes you are just writing stuff and realize that you need to do something slightly more complicated with a table than HTML and pure ASCII-art allow. Doing it in spreadsheet software entails either a workflow break or having to do over the rest of the document in a text processor.

My take: You're not going to be working with some 500 row monster etc. in Org, but if you just want to quickly do something like get the row-wise averages of a table, you can just do that in-place easily. The underlying math engine in Emacs is also pretty fancy, it can do stuff like symbolic algebra which I have yet to see in Excel :)

I think part of it at least is being able to operate on spreadsheet-style data in a contiguous environment, building it into the document in a way that you can’t if there’s a file type divide

To add on this - you can use it as inputs (and outputs) for org babel. Eg. A tabulated list of ENV vars to be passed to a bash script, which in turn generates a list of URLs that you can pass to a RESTclient script or python.

When editing, it updates the width of columns automatically and supports tabbing through, along with re-arranging rows using common shortcuts within org mode.

Org-mode tables are super convenient for organising information. You just press | and tab and you have a new table in your text file! You can insert lines with |- tab and the columns auto expand to fill your text.

The formula functionality is pretty useful too, although I rarely use it and working it out is a bit of a puzzle still.

I have a bookmarklet in my browser for recommendations. I have org-protocol set up so when I highlight a book title and click the bookmarklet, org mode opens and prompts me for several other things (author, genre, notes, format, etc.) and then it saves all that to a table with the website I found the recommendation. The table allows me to sort different genres and go back to see why I found that title compelling in the first place.

I also use org's ability to show properties as a table. When I read something I write a little review for myself and have some relevant info in the properties drawer. I can toggle a table view to see and sort all the properties.

Not specifically spreadsheets, but the org table code is also used in the feature-mode for writing gherkin tests. One of the powerful aspects of cucumber tests is the tables, and having the org tables code in the editor means that the feature mode didn't have to do anything to support tables that auto size and align and add columns etc.

I use org-mode mainly just for note taking, and I've only been using it for a little over a year. In addition to notes, I do software planning and cost estimating. I prefer this to a "real" spreadsheet application, because they are mostly text, with a few important numbers. Those numbers can still easily be added to computations.

Also, stubbornly, I prefer my comfortable emacs environment to the clicky unfamiliar interface of something like Google Sheets / Excel. Finally, org-export is fairly infinite in potential. I actually sync these notes to GitHub, which displays the org file as Github-Flavored Markdown -> HTML without even needing to export. The tables export very nicely, and are easily consumable in a read-only fashion by those who asked for them.

Imagine having small patches of spreadsheet along your document without the clutter and slowness of a GUI. Also the equation evaluator is a symbolic engine. I manage all my nutrition with it, from macronutrient count to ingredient prices.

Any chance you could share some of that? Sounds cool!

Something like this. There are other parts calculating the requirements.

  | -1 | req. | 100 | 100 |  50 |               
  |  3 | milk | 2.92| 4.37|  3.5|
  |  # |      |-91.2|-86.8|-39.5|             
  #+TBLFM: @>$3..$5=(@<$1..@>>$1)*(@<..@>>)

For the same reason it is sometimes useful to put tables in Word.

Its plain text one note/ evernote, plus a bunch of macros for formatting text.

Because when freely and quickly writing notes, being able to freely and quickly jot down a table is enormously helpful.

Source: I use org-mode for project notes, and they do end up with a lot of tables.

I'm with you. If I want to subtotal and subtract one subtotal from another, I'm popping over to excel. But... columnar text is a big win.

Reply on news.ycombinator.com