Surprising number of scientific papers on genetics contain serious errors due to Microsoft Excell number errors

archive.is/RUYwA

TL;DR: MS Excell's shitty type system has garbled countless gene names by assuming that they're dates. Most of the time, the scientist can fix the errors manually, but with n way to turn the autocorrect off, nearly 20% of genetic papers have been left with uncorrected errors in gene names in the research.

Other urls found in this thread:

nature.com/nature/journal/v482/n7386/full/nature10836.html
twitter.com/SFWRedditVideos

you'd think scientists of all people would care enough not to use windows, but that couldn't be farther from the truth.

The smartest of people often make the dumbest mistakes.

Most/all commercial scientific instruments like you would find in a bio lab are locked into proprietary wincuck environments straight out of the factory so it's the manufacturers who are to blame. There's not much of an incentive for them to switch. A lot of the software is truly awful but it does just enough to work so it gets used. Most scientists have the same computer literacy as your local plumber and just don't give a shit.

I hope you people are just pretending to be retarded. This isn't an Excel bug. It's users not being competent at using Excel and not bothering to check their work despite blatant errors.


Neither Libre/Open Office is anywhere close to as powerful as Excel anyway.

Why would they even bother with excel? My only guess is to make readable tables? Otherwise they should be manipulating this data with scientific python, octave, or at least MATLAB.

Pic unrelated.

Microcucks, everybody.

I'm still trying to comprehend why Microsoft is still the standard in business.

as usual, carry on

It's the "if you pay for it it's good" argument.

It's the easiest way to make scientific graphs for normalfag scientists.

Probably because the stack is too complex. If we were not stuck with a software paradigm from the late 60s and early 70s things might be better. I am thinking a modern day LISP machine. A LISP machine has a feature that people like from Excel, interactivity and quick results.

LISPcucks gonna LISPcuck.

LISP pretty much invented it all right out of the gate. It lost because Moore's Law. Technical limitations forced people to write Assembly for home computers and backwards compatibility is a thing. Since PCs drove the market, that is what we are stuck with.

Fucking Australians.

Automatic datatype guessing was a mistake.

This. Intelligence != understanding.

What pisses me off about computer-illeterate people is that they don't realize that they are computer-illiterate. I for example know that I have no idea about plumbing, so when I need some plumbing done I will call an actual plumber. When I need something to be done with my car I will find an actual car mechanic, I won't fiddle with the engine myself. But when it comes to computers people will either be ignorant of their illiteracy or downright proud.


How would Lisp have solved anything, even assuming that it was an Excel bug and not the scientists being too stupid to use it?

F-f-fuck you and your facts, w-wincuck!

The LibreOffice version can be turned off. The Excell version is what is literally holding back aggregate human knowlege of genetics. All you are trying to do is distract from this fact. You can't escape the truth, wincuck.

Why blame Windows when Microsoft sells Excel and their whole Office suit for other platforms. It seems really silly to blame an OS for a piece of software that runs on top of it, do we even know if these people use Windows? Windows (as well as other platforms) have the option to run a different set of software, they don't have to use Microsoft's Excel.

You're kidding, right?

wew

The article acknowledges that formatting can be turned-off, just that it's a per-column setting rather than for the whole program. I'll take them at their word that it's impossible to turn off the automatic type conversions for the whole program, but even if it were, it wouldn't have helped those scientists because they're too negligent to even bother checking their work to see such an obvious error in the first place. It's hard to call this anything other than user negligence.

Fucking Australians that invented key parts of WiFi, cheers burgerfag.

Don't get me wrong, I've done some light statistics in Excel, like calculating expectation values for various wagers/gambling, but anything serious I used to do in MATLAB back in college. I'm currently switching to scientific python.

Why does everyone default to this?

Because we are the most populous English speaking country.

It's not a bad assumption to make.

Fair enough.

...

this paper addresses the problems with proprietary software for science more specifically:
nature.com/nature/journal/v482/n7386/full/nature10836.html