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

I was wrong about spreadsheets (2017)

reifyworks.com

Post reply on HN
Print view View original

131–140 of 378 posts

Re: I was wrong about spreadsheets (2017)

#131

Post by TrackerFF »

My only real problem with spreadsheets is that once they get too big, there's almost no turning back. There's no easy fix if someone screwed up at the beginning.

Even though conventional programming languages are less visual, it's so, so much easier to modify models.

But with that said, I do use spreadsheets for 90% of my daily needs, as far as calculations go. Just type "sheets" into the search bar, and I can start working (google sheets) in 2 seconds. It's even faster than firing up notepad.

Re: I was wrong about spreadsheets (2017)

#132

Post by Lukas1994 »

thrwayxyz wrote: post #80

I worked as a spreadsheet DevOps in a financial company. We used to have a python script to spin up windows vms with the exact version of windows that was known to be good, with the exact version of excel and then automatically kickoff the computation. The reson for this was someone once got the order of execution wrong and screwed up a couple billion dollars worth of trades. Excel is great until you have to maintain…

Many more stories like this: http://www.eusprig.org/horror-stories.htm

Re: I was wrong about spreadsheets (2017)

#133

Post by Xelbair »

mdorazio wrote: post #7

The sheer number of everyday business use cases that can be solved quickly and efficiently with spreadsheets is pretty astounding. Of course, there comes a point when heavily-used spreadsheets become intricate behemoths on which core business functions are run, and then things can get bad in a hurry. The challenge many programmers face is thinking that every business problem is best solved with code, when in reality…

one of my first 'real' programming tasks involved replacing excel sheet with simple application.

Bcause excel sheet was doing a lookup in HUGE(i am talking about few million rows of raw data) dataset - basically a fulltext search in a database with unstructured address data.

The idea was to fuzzy join two datasets, and to make manual joining a bit easier with search functionality.

It worked fine on dev's machine, but on user's toasters it took an hour to load.

Re: I was wrong about spreadsheets (2017)

#134

Post by weinzierl »

jedberg wrote: post #45

> and transmit not only the results of those calculations, but the actual environment itself, to anyone in the world, and expect that if they have a computer, they can replicate your results. I think this is the real secret. It's really the only kind of "model" where the environment travels with it. Everyone has the same Excel setup. If it works for you it will work for them.

> Everyone has the same Excel setup.

I wish this was true. The biggest problem is different language versions that cause all sorts of problems. Did you know that Excel formulae names are localized? For example AVERAGE is MITTELWERT if you have a German Excel and there are some with umlauts as well.

But there are more subtler problems. I'm an engineer and I used to work a lot with radians and degree values. There is this nice little trick in excel where you can enter radians and have it display as degrees. It works using the date format but only if you have 1900 dates enabled. I used to to use that for a short while until I noticed that all dates in spreadsheets from other people are off by 30 years.

Re: I was wrong about spreadsheets (2017)

#135

Post by qrybam »

Excel is powerful. It does so many things really well. Sharing data between sheets in real-time isn’t one of them though (maybe I just haven’t found the right tool, so we went ahead and built one).

I worked at a large fintech company a few years ago. During my interview Excel popped up as a topic, the interviewer quipped “Excel is terrible!”, he was referencing the heavy reliance of his customers on spreadsheets rather than the “better” functionality they offered via their platform. A few years before I would have agreed, but Excel really is amazing, it allows almost anyone to just-get-things-done.

There are plenty of cases where Excel projects had grown to the point where specialised software would be better for the business and the users... and if you’re in an industry with heavy and advanced usage of excel (like the fintech space), it’s a great place to mine solid ideas for a startup. Just don’t try to recreate everything Excel does! Focus on areas it doesn’t excel in.

Re: I was wrong about spreadsheets (2017)

#136

Post by etatoby »

viraptor wrote: post #94
skrebbel wrote: post #67

Earlier quoted context omitted.

> the fact it's too easy to I'm not sure you fully appreciate why Excel is so successful.

No, I fully understand that this is great for many users and that it cannot be made harder in any way to make whatever mess you want on a single sheet - or excel would die. I only want it easier to not do that. (There's nothing guiding people to better design right now)

That's by design. People don't want to be "guided to better design," they want to get their math done and over with, leaving some trace (the Excel sheet) for the next time they or somebody else needs to revise that math.

Re: I was wrong about spreadsheets (2017)

#137

Post by kinj28 »

viraptor wrote: post #21

My issue with spreadsheets is that they could be improved a lot with minimal changes, not nobody seems to do it in popular office packages. Making them more database-like and making table data first-class (at least you can make named tables in excel on windows) could be used to push people a bit more towards organised data, without changing how anything works. Half of the mess that makes excel hell comes from the fac…

In my opinion - "No Code" tools will fix this gap to build real world apps yet being as simple as excel.

While airtable is just a DB as a service, a real no code tool would have ideally all the basic building blocks:

a. DB as a service (like Airtable) b. BPM capability -> to add workflows to the data c. Ui Building to abstract data d. API integration (may not advanced learning for power users)

Re: I was wrong about spreadsheets (2017)

#138

Post by thom »

I always like these threads where programmers talk about tools like Excel and say they’re fine for small stuff but grow out of hand, as if 90% of codebases built by professional programmers aren’t exactly the same.

Re: I was wrong about spreadsheets (2017)

#139

Post by davedx »

viraptor wrote: post #29
dgaudet wrote: post #24

Earlier quoted context omitted.

named ranges and DGET "sql-like" queries are in google sheets as well. (or perhaps what you meant is that few people know about these things, and don't use them, so you end up with unreadable messy code...)

Yeah, I meant that they're not right in your face and ready to be used. I know it's hard to force people to use new methods, but if anyone could do it, MS / Excel could. It's all possible though, if you know where to look (I think excel even has table sheets now?)

I see what you're saying, but this is actually a deliberate usability principle: make the basic features easy to access (visible without clicking into a menu), and put advanced features inside menus. Then basic users who don't need more advanced features can do what they need without getting lost in a super complex UI, and power users can still get to the advanced stuff.

If people want to do advanced stuff with software, they really do need to invest the time to learn where the advanced features of the software are. I think this is reasonable.

Re: I was wrong about spreadsheets (2017)

#140

Post by pbowyer »

viraptor wrote: post #21

My issue with spreadsheets is that they could be improved a lot with minimal changes, not nobody seems to do it in popular office packages. Making them more database-like and making table data first-class (at least you can make named tables in excel on windows) could be used to push people a bit more towards organised data, without changing how anything works. Half of the mess that makes excel hell comes from the fac…

> at least you can make named tables in excel on windows

I built a spreadsheet that loaded external data sources (filtered log files from a production system) into Excel tables. I then combined two log files into another table, and created a pivot table from this, which I then filtered (date ranges etc) and analysed.

That worked great at the start - until I came to update the external data and found the pivot table didn't update. Then when I force an update it lost my groupings, filtering etc.

Any idea where I went wrong?

Post reply on HN