This is a very grey area, because while excel is fun to enter, and play around with, often times it's seen as a pseudo-programming language.
In Excel, one can drag-down formulas to automatically match cell entries, and other minor patterns. Usually, these formulas are dragged to the bottom of the input index, or (often times) all the way down to the bottom of the sheet.

One can start to see how this gets strange very quick, as these drag-downs are useful, but hard to score.

Note, I'm not talking about answers where dragging down is part of the input. I'm explicitly referring to answers that already have the formulas in them before any input.

How should we handle these answers? Should they be ignored? Scored?

| |

Scored answers must be keyboard-only

Given that the Range operator exists that can replace the mouse drag-down operation, and the arrow keys can move between cells to (for example) enter multiple snippets of code in separate areas, I can't think of a reason why the mouse needs to be involved.

This also has the benefit of being able to be scored similar to Vim, for example, based on keystrokes.

| |

Just count the bytes

This is a common question for oddball languages, from Scratch to Minecraft to Lego WeDo. The answer is usually (from what I've seen) to just count the bytes of the saved file.

It's a simple method, easily understood by anyone. Will this make the score go up? Probably, but if you cared about having a super low score you probably wouldn't be using one of these languages in the first place.

Note: If you can save it one of several Excel formats, choose the one with the lower byte count. Assuming that format has the features necessary to then open it back up and run it properly.

| |
  • \$\begingroup\$ what if the formula takes up a whole column (infinite formulas)? \$\endgroup\$ – user56309 Sep 22 '16 at 18:29
  • 4
    \$\begingroup\$ That's exactly the type of problem this should eliminate. Save the file and count the bytes. I'm fairly sure it won't take infinity bytes on your drive. \$\endgroup\$ – Geobits Sep 22 '16 at 18:30
  • \$\begingroup\$ I don't think this is much practical, because .xls/x files can be different, even if they are saved by different versions of Microsoft Office, say 2007 and 2010 for example (yes, that close). \$\endgroup\$ – Erik the Outgolfer Sep 30 '16 at 13:47
  • \$\begingroup\$ I don't really see the issue with that. Use whichever you can to save bytes if you want. Choosing between Excel 2007 and 2010 is like choosing between Python 2/3 or Java 7/8 for this. If there's a difference, just make sure you mention the version. \$\endgroup\$ – Geobits Sep 30 '16 at 13:50

Score as a polynomial

The way I wanted to suggest scoring these is to leave the score as a polynomial. for example, say you have some code in excel like the following:
enter image description here

because "=(some more code)" is at a single point in the sheet, this will be the 0th term in the polynomial. In this case it is 17 bytes.
The "=(Some code)" snippet occurs along one dimension in the array (the A column), so that would be the 1st term in the polynomial. In this case it would be 12n bytes.
And in the rare case that someone has a multidimensional formula, then it will be considered the 2nd term in the polynomial. In this case it would be 25n² bytes

so, if we were to score this sheet all together, it would be 25n² + 12n + 17 bytes.

While this method describes the nature of the entry well, it still lacks a proper method of scoring.

| |