3 ms·

The same way that engineers saying you shouldn't use dung to build houses are guilty of architectural elitism. Excel is the wrong tool because it mixes present

by konjin 6y ago

The same way that engineers saying you shouldn't use dung to build houses are guilty of architectural elitism. Excel is the wrong tool because it mixes presentation and data, to the point that geneticists had to change the names of genes so their spreadsheets would stop turning them silently into dates and then destructively changing the original data[0]. >Excel works great - I've personally built some very complicated models that work just fine. I've personally cleaned up after people like you, which cost the business millions of dollars because they didn't realize how unqualified they were for doing their jobs. >Goes through the same process of QA and line-by-line checking of requirements. Somewhat difficult as excel is not line based. "Oh yes, these constants are on a different sheet, if you just click here, here, here and here you can see where we get the original data, we multiply it by 1.0 to make sure it's not a date over there and then we feed it back ..." if excel was sanely serializable you could easily version control it and see meaningful diffs between check ins. Double points if you could run tests on sheets and run them in batch mode easily (I said easily). [0] https://interestingengineering.com/genes-renamed-to-stop-microsoft-excel-from-mistaking-them-for-dates https://interestingengineering.com/genes-renamed-to-stop-mic...

syshum 6y ago
>>>I've personally cleaned up after people like you, Yep... The critical question that always gets blank stares is "How is this data and calculations validated" There is never an answer to that question from the ExcelMaster, it is always a variation of "it looks correct to me" Wonderful....
- gxs 6y ago
  Except you can do it if you know what you're doing. No one is saying to manage millions of records etc. But you can get more mileage out of excel than people think - and people who half understand it are the worse because they know enough to know that it doesn't work. Custom solutions are great - but aren't always the answer and with proper process Excel will work just fine in many use cases.
  - konjin 6y ago
    >Except you can do it if you know what you're doing. Ye gads, why didn't we think of that in any other programming language? If we know what we're doing we can do whatever we feel like, and if we get the wrong answers, it was obvious that we didn't know what we were doing which can be fixed by just knowing what we were doing! Excel's only use case is for data that fits on one screen, or an exploratory poke at the data to see what's in which column and if there are any patterns you can eyeball. Then you put those hunches in a script and start doing the work for real.
  - lmm 6y ago
    How can you achieve anything equivalent to unit testing? How can you do any kind of change control (unless you treat every edit to the sheet as a change)? Excel does some things well but for anything that needs to work robustly as part of a process (as opposed to one-time exploratory data analysis) it lacks things that were basic in 1996. It's an "unsafe at any speed" situation: it may work as long as you do everything right, but it's missing basic safeguards as soon as you make a single mistake; I'd put it in the same category as using a memory-unsafe language.
rvba 6y ago
This geneticist thing is a complete misunderstanding. Excel is tuned to autoconvert things that look like dates to dates. Best would be to be able to disable this, but Microsoft research shows that more users need it. Meanwhile scientists use Excel for lots of stuff.. since it works. It has a not nice "feature" of those conversions. So option 1) throw away the tool you use every day 2) change the gene name Option 2 makes sense, although everyone would be more happy if Microsoft gave an option to toggle that auto-conversion off [there is option to properly load CSV files but people dont use it, since it takes more time...]
- konjin 6y ago
  So I'm wrong because I'm right? Have you looked at Stockholm syndrome?
  - rvba 6y ago
    The generic tool for offices is so good at everything that geneticists also use it; they change a name of a gene to make own life easier?
gxs 6y ago
And I've cleaned up after many, many shitty engineers, one example being engineers who want to use a sledge hammer to drive a finishing nail and again, that's just one example. And if you want to get personal to that level they all the same know it all personality who think they write MUCH better code than they actually do. Pretty sure they cost companies even more money. Luckily for me, I have the practical sense to know when to use what tool and so far it's worked out just fine. I wouldn’t use excel where not appropriate, so kindly don’t make assumptions. But more power to you, keep shitting on tools that are more sophisticated than you probably think. People who half know excel are the worse, they are the ones who know it just well enough to think it doesn't work at all. Custom solutions are great - but aren't always the answer and with proper process Excel will work just fine in many use cases that you might not expect. I could blab on about this forever, because, you know, there is a lot of gray area or whatever.
- konjin 6y ago
  Imagine asking people to use the custom solution sledge hammer of python, git, unit tests and sqlite. Not one off the shelf piece of software there. There comes a point where using a spork to drill holes in walls is not the right solution. That you're shitting on people pointing out the obvious tells me you're still stuck in the 90s windows developer mindset.
  - gxs 6y ago
    The problem is you only hear about botched excel jobs - you never hear about all the successful projects, so when you hear a horror story you just jump to the conclusion excel sucks. Ok, ask a geneticist to build what they need with their “off the shelf” parts you mentioned vs 90% of the time being self sufficient with excel and allowing them to, you know, do genetics work. (The other 10 percent being where it makes sense to build something more custom). Reminds me of the classic HN post when Dropbox had just launched where the user exclaimed “this is just a mounted blah blah using subversion blah anyone can do it” (paraphrased) That old Dropbox post summarizes this mindset quite well. And if being practical with your head not in the clouds making sane business AND technical decisions is what the 90s were like, man I missed out. I guess at the end of the day we’re arguing over where you draw the line regarding when to stop using excel and graduate to something different, a blurry line that at the end of the day is a judgement call. Oh and for the record I wasn’t even arguing against or for this particular use case in this post - speaking more in general terms.

by konjin 6y ago

syshum 6y ago

>>>I've personally cleaned up after people like you, Yep... The critical question that always gets blank stares is "How is this data and calculations validated" There is never an answer to that question from the ExcelMaster, it is always a variation of "it looks correct to me" Wonderful....

gxs 6y ago
Except you can do it if you know what you're doing. No one is saying to manage millions of records etc. But you can get more mileage out of excel than people think - and people who half understand it are the worse because they know enough to know that it doesn't work. Custom solutions are great - but aren't always the answer and with proper process Excel will work just fine in many use cases.
- konjin 6y ago
  >Except you can do it if you know what you're doing. Ye gads, why didn't we think of that in any other programming language? If we know what we're doing we can do whatever we feel like, and if we get the wrong answers, it was obvious that we didn't know what we were doing which can be fixed by just knowing what we were doing! Excel's only use case is for data that fits on one screen, or an exploratory poke at the data to see what's in which column and if there are any patterns you can eyeball. Then you put those hunches in a script and start doing the work for real.
- lmm 6y ago
  How can you achieve anything equivalent to unit testing? How can you do any kind of change control (unless you treat every edit to the sheet as a change)? Excel does some things well but for anything that needs to work robustly as part of a process (as opposed to one-time exploratory data analysis) it lacks things that were basic in 1996. It's an "unsafe at any speed" situation: it may work as long as you do everything right, but it's missing basic safeguards as soon as you make a single mistake; I'd put it in the same category as using a memory-unsafe language.

rvba 6y ago

This geneticist thing is a complete misunderstanding. Excel is tuned to autoconvert things that look like dates to dates. Best would be to be able to disable this, but Microsoft research shows that more users need it. Meanwhile scientists use Excel for lots of stuff.. since it works. It has a not nice "feature" of those conversions. So option 1) throw away the tool you use every day 2) change the gene name Option 2 makes sense, although everyone would be more happy if Microsoft gave an option to toggle that auto-conversion off [there is option to properly load CSV files but people dont use it, since it takes more time...]

konjin 6y ago
So I'm wrong because I'm right? Have you looked at Stockholm syndrome?
- rvba 6y ago
  The generic tool for offices is so good at everything that geneticists also use it; they change a name of a gene to make own life easier?

gxs 6y ago

And I've cleaned up after many, many shitty engineers, one example being engineers who want to use a sledge hammer to drive a finishing nail and again, that's just one example. And if you want to get personal to that level they all the same know it all personality who think they write MUCH better code than they actually do. Pretty sure they cost companies even more money. Luckily for me, I have the practical sense to know when to use what tool and so far it's worked out just fine. I wouldn’t use excel where not appropriate, so kindly don’t make assumptions. But more power to you, keep shitting on tools that are more sophisticated than you probably think. People who half know excel are the worse, they are the ones who know it just well enough to think it doesn't work at all. Custom solutions are great - but aren't always the answer and with proper process Excel will work just fine in many use cases that you might not expect. I could blab on about this forever, because, you know, there is a lot of gray area or whatever.

konjin 6y ago
Imagine asking people to use the custom solution sledge hammer of python, git, unit tests and sqlite. Not one off the shelf piece of software there. There comes a point where using a spork to drill holes in walls is not the right solution. That you're shitting on people pointing out the obvious tells me you're still stuck in the 90s windows developer mindset.
- gxs 6y ago
  The problem is you only hear about botched excel jobs - you never hear about all the successful projects, so when you hear a horror story you just jump to the conclusion excel sucks. Ok, ask a geneticist to build what they need with their “off the shelf” parts you mentioned vs 90% of the time being self sufficient with excel and allowing them to, you know, do genetics work. (The other 10 percent being where it makes sense to build something more custom). Reminds me of the classic HN post when Dropbox had just launched where the user exclaimed “this is just a mounted blah blah using subversion blah anyone can do it” (paraphrased) That old Dropbox post summarizes this mindset quite well. And if being practical with your head not in the clouds making sane business AND technical decisions is what the 90s were like, man I missed out. I guess at the end of the day we’re arguing over where you draw the line regarding when to stop using excel and graduate to something different, a blurry line that at the end of the day is a judgement call. Oh and for the record I wasn’t even arguing against or for this particular use case in this post - speaking more in general terms.