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

The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

wsj.com

Post reply on HN
Print view View original

41–50 of 435 posts

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#41

Post by PascLeRasc »

rhodysurf wrote: post #14
css wrote: post #5

I do not understand the mindset of not wanting to help other people. Enabling people to learn more about the software they use every day is better for everyone involved. Yes, some problems are too complex to solve in a few minutes, but pointing someone in the right direction requires little to no effort. Excel lacks the community programmers have and also lacks the learning opportunities given by open-source software…

There is definitely value in teaching others, but first they have to want to learn. I work with a lot of mechanical engineers and I am the only real "software guy" in our office. Most of the time they don't want to learn how to do computer things themselves because (mostly the older engineers) consider it to not be their job or out of their scope.

At my previous job in the power industry we had to clean up our shared drive and move all the CAD files of one type into a certain new folder, and the CAD files of the newer type into another, and PDF outputs in another, GIS exports in another, etc. Three people were dragging-and-dropping by hand for hours until I had some free time and made a Python script in 10 minutes to do this all. Got a few free drinks that day.

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#42

Post by irrational »

throwanem wrote: post #10
crooked-v wrote: post #6

Earlier quoted context omitted.

Have you considered trying to get people to use something like Jupyter or R Notebook?

"Now you have two problems."

Can you help me with RegEx in Excel?

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#43

Post by lordnacho »

Having used Excel for many years on the trading floor, I have to say it's awful.

It does everything, badly. Which is why it's popular. People can get by without properly thinking things through, just jamming a few formulas on a spreadsheet, and patching the calculation when something comes up. You're never forced to think rigorously.

You can use it as a crappy database. Or a crappy UI. Or a crappy place to call external DLLs from. It's the perfect tool for that guy who "has a great idea, but just needs it coded up".

If a guy comes to you with an Excel question, he wants you to help him. Contrast this with a range of other technologies. If a guy comes to you with a question about c++, you tend to be helping each other. Same can be said for any number of things that non-techies would not think to ask. When was the last time someone asked you about Haskell or Erlang where you were treated merely as a means to someone else's end?

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#44

Post by css »

simonbyrne wrote: post #40
css wrote: post #5

I do not understand the mindset of not wanting to help other people. Enabling people to learn more about the software they use every day is better for everyone involved. Yes, some problems are too complex to solve in a few minutes, but pointing someone in the right direction requires little to no effort. Excel lacks the community programmers have and also lacks the learning opportunities given by open-source software…

There is also a lack of what is considered "Excel best practices", for want of a better word. The only resource I've seen along those lines is Joel Spolsky's "You Suck at Excel" video (which is excellent, and should be necessary viewing for anyone who use Excel on a regular basis): https://www.youtube.com/watch?v=0nbkaYsR94c

Yeah, great video. I often send this to colleagues. Another blog I recommend to others is https://itsnotaboutthecell.com/

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#45

Post by bachmeier »

ams6110 wrote: post #37
peatmoss wrote: post #33

Earlier quoted context omitted.

Past a low complexity bar, I feel like R is a much easier answer than Excel. Excel can get pretty hairy pretty fast. As you mention below, debugging even terrible R code is easier than debugging Excel.

Really? I took a stab at R about a year ago and found it utterly confusing.

I think the point is that formulas and macros are more utterly confusing.

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#46

Post by bachmeier »

lordnacho wrote: post #43

Having used Excel for many years on the trading floor, I have to say it's awful. It does everything, badly. Which is why it's popular. People can get by without properly thinking things through, just jamming a few formulas on a spreadsheet, and patching the calculation when something comes up. You're never forced to think rigorously. You can use it as a crappy database. Or a crappy UI. Or a crappy place to call exter…

> Having used Excel for many years on the trading floor, I have to say it's awful.

But the interface is not awful - it's absolutely wonderful as a quick way to enter and share data and do a fast analysis. The part that's awful is when you do more than that. And once you've got the data into Excel, why go through the trouble of changing to something else just to do one additional thing, even if that additional thing is complex?

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#47

Post by peatmoss »

ams6110 wrote: post #37
peatmoss wrote: post #33

Earlier quoted context omitted.

Past a low complexity bar, I feel like R is a much easier answer than Excel. Excel can get pretty hairy pretty fast. As you mention below, debugging even terrible R code is easier than debugging Excel.

Really? I took a stab at R about a year ago and found it utterly confusing.

I’ve taught R (and Python) to mostly non-programmer scientists as part of multi-day Software Carpentry workshops. Especially with R, I feel like most attendees after a couple of days were able to do things that would be decidedly non-trivial in Excel.

EDIT: emphasis on my original point of “past a low complexity bar.” If someone is looking to manually enter data and maybe sum a row or two, Excel is probably easier. I’d say at the point where someone needs to reshape data or do any kind of non-trivial formula, R becomes easier—certainly easier for me to imagine teaching.

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#48

Post by roel_v »

ams6110 wrote: post #37
peatmoss wrote: post #33

Earlier quoted context omitted.

Past a low complexity bar, I feel like R is a much easier answer than Excel. Excel can get pretty hairy pretty fast. As you mention below, debugging even terrible R code is easier than debugging Excel.

Really? I took a stab at R about a year ago and found it utterly confusing.

It is. It's convoluted and idiosyncratic for no reason and all together a puzzled together mess. R is not (much) better than Excel. It's better in the same way that getting stabbed in the eye with a hot poker twice is better than getting stabbed with it three times.

Re: The First Rule of Microsoft Excel: Don’t Tell Anyone You’re Good at It

#50

Post by lordnacho »

bachmeier wrote: post #46
lordnacho wrote: post #43

Having used Excel for many years on the trading floor, I have to say it's awful. It does everything, badly. Which is why it's popular. People can get by without properly thinking things through, just jamming a few formulas on a spreadsheet, and patching the calculation when something comes up. You're never forced to think rigorously. You can use it as a crappy database. Or a crappy UI. Or a crappy place to call exter…

> Having used Excel for many years on the trading floor, I have to say it's awful. But the interface is not awful - it's absolutely wonderful as a quick way to enter and share data and do a fast analysis. The part that's awful is when you do more than that. And once you've got the data into Excel, why go through the trouble of changing to something else just to do one additional thing, even if that additional thing i…

As technological scrap paper, I admit it can be useful.

But as soon as your "fast analysis" is done, you need to be hardening whatever process it is you are creating. After all, that is what your analysis is about, right? Building some sort of repeatable, often auditable, transparent process.

Post reply on HN