The #1 Reason I Don't Want to Work in Academia - Formatting Papers Is A Huge Waste of Time | /en/2017/12/formatting-papers/

yihui 2022-12-17 01:59:35

The #1 Reason I Don't Want to Work in Academia - Formatting Papers Is A Huge Waste of Time

https://yihui.org/en/2017/12/formatting-papers/

3 Comments

giscus-bot 2022-12-17 01:59:36

Guest *Conor Neilson* @ 2017-12-04 06:47:07 originally posted:

Yihui, this was a fantastic post. I agree that the idiosyncrasies of journal formatting are needlessly time consuming, and doing away with them would be hugely beneficial. What do you think is the ideal end goal here? You say you'd rather do away with journals, so what do you think should happen to the peer-review model? Would this still exist in a modified form?

yihui 2022-12-17 01:59:40

Hi Conor, thanks for the comment! I have a lot to say on journal publications, although I'm probably not very qualified (I don't have a lot of journal publications). Formatting is only one of the relatively small problems. The peer-review model is a bigger one and also needs a revolution. Perhaps next year I'll be able to do some substantial work. I wrote some ideas in 2012: https://yihui.name/en/2012/03/a-really-fast-statistics-journal/ Some ideas can be much better implemented now, given our long-time investment in R Markdown and other technological advances.

I gave a talk on this topic in Australia this summer and you can know a brief summary of my wishes from my slides: https://slides.yihui.name/2017-DSM-Journal-Yihui-Xie.html

Originally posted on 2017-12-04 15:59:11

giscus-bot 2022-12-17 01:59:37

Guest *yihan* @ 2017-12-04 13:31:52 originally posted:

Formatting paper is bitter enough, but the worse thing is that many professors have learnt to judge a paper by its format. One of my professor would refuse to read paper not properly formatted in APA style...

@yihanxu

yihui 2022-12-17 01:59:39

Sounds like a professional prisoner... The professor should have read the "Poor Charlie's Almanack" when he/she was young ( https://en.wikipedia.org/wiki/Poor_Charlie%27s_Almanack ), because Charlie Munger constantly reminds people of avoiding intense ideology: http://thereformedbroker.com/2015/02/16/qotd-charlie-munger-on-ideology/ I don't see any point of being so loyal to a particular citation style.

Originally posted on 2017-12-04 15:39:53

giscus-bot 2022-12-17 01:59:38

Guest *Jiena Gu* @ 2017-12-04 16:57:09 originally posted:

Hi Yihui,
Fantastic article. I agree that the academic paper is to communicate the ideas and results but not the style or the look. Unfortunately we spend most of time worrying about the format instead of the content.

I really like the html format research paper because I can embed some D3 animated and interactive plots into the research paper which is impossible in the printed paper.

@jienagu

yihui 2022-12-17 01:59:41

I absolutely agree with you. I also mentioned the same thing in my talk (HTML first, and PDF later): https://slides.yihui.name/2017-DSM-Journal-Yihui-Xie.html#29 I believe the future on HTML is much brighter than PDF.

Originally posted on 2017-12-04 19:50:30

giscus-bot 2022-12-17 01:59:42

Guest *Jumping Rivers* @ 2017-12-07 21:35:11 originally posted:

Although the problem with html/css/javascript, is how well will it age. If you write a paper with fancy D3 graphics, how likely is it that it will still work in 1, 5, 10 years time?

Of course this doesn't apply to markdown.

yihui 2022-12-17 01:59:43

That is a really good question. Here are a guess and a thought of mine:

D3 is less likely to survive longer than PDF. Many technologies, especially web technologies, come and go quickly. However, I believe HTML, CSS, and JS will survive longer than PDF.
I think it is fine if a particular technology or library fails to work in 5 years, but works exceptionally well before then, e.g., if it helps us make important scientific discoveries. When a popular technology/library stops working, I'm pretty sure there will be a new one replacing it, so I'm not too concerned about it.

Originally posted on 2017-12-08 05:03:02

giscus-bot 2022-12-17 01:59:44

Guest *Colin Gillespie* @ 2017-12-08 08:44:33 originally posted:

Not wanting this to descend into a massive thread, so don't feel the need to respond (although response is good).

Yep I think most people would agree with that.
5 years; that's terrible! Having publications break in 5 years would be a disaster. In Maths/Stats I regularly look at papers, 5, 10, 20, 40 years old.

Issue logging on with Disqus, so using guest log-in

yihui 2022-12-17 01:59:45

Software broken in 5 years does not necessarily mean publications or scientific discoveries will break, too. Results may still be valid. Software may only be auxiliary or for exploratory purposes (I think D3 is a good example). For example, I don't think many people can still run programs written on punch cards today, but that does not mean all results calculated from punch cards are no longer valid.

Originally posted on 2017-12-08 15:03:05