5 ms·

Watch a Ph.D. thesis dissertation being written in (somewhat) realtime

56 points by hds 12y ago

pagekicker 12y ago
I was expecting to see the ten-year process involving teaching assistantship, employment, marriage, and children, all illumninated by the light of a candle at 2 am in a basement office next to the crib.
ygra 12y ago
I wrote something similar while I was writing my diploma thesis, albeit not that visual. For me it was handy to see whether I was making any progress, so I regularly updated a CSV file with the current number of pages, lines, words, figures, code lines, and code comments. The idea was that on every day I did something I could at least look at the data and see how much (also the script automatically adjusted page, line and word count for the appendix which I didn't want to count).
sanderjd 12y ago
That's really neat! I'm not sure if this is intentional because your work is private, but I would love it if I could click on the images or SHAs to go actually see the thing.
- hds 12y ago
  At the moment I'm a little hesitant to make it public. Partly because there's a large chunk of unpublished research in there, but also because some of the early drafts are a little embarrassing. Perhaps when the thesis is deposited (and hence public) I could add that in for "historical" interest.
  - sanderjd 12y ago
    Yeah, totally understandable.
  - j2kun 12y ago
    Definitely don't make it public. I know people who have had their thesis topics scooped.
- Braunbart 12y ago
  Yep, I instinctively clicked on one of the pages to see it bigger.
burritofanatic 12y ago
Excellent! People writing important papers really should be using Git. When documents are extremely complex, previously removed portions can have pertinence later.
- Fomite 12y ago
  This doesn't work particularly well for fields not using LaTeX as their document creation tool.
  - dovel 12y ago
    Why is that?
    - jaredsohn 12y ago
      I would think the GP means that git works much better if you are creating a document in a non-binary format (of which LaTeX is one example.)
- irremediable 12y ago
  Seconded -- at least for anyone using LaTeX or another git-friendly format. I used git for my masters dissertation recently, and it was very useful to be able to checkout old versions and compare them with its then-current state.
toppy 12y ago
That's an excellent solution for big problem in my country (not at PhD level though) namely plagiarism. Make students to use such system, turn autocommit on and you can see their progress and all deviations (big chunks of text appearing from nowhere etc.).
- doctorKrieger 12y ago
  once they learn about it nothing will change.
archgrove 12y ago
Whilst not as slick as the JS version, I did much the same thing with mine: https://vimeo.com/76969179 https://vimeo.com/76969179 when bored one day. I couldn't find any tools to create exactly what I wanted, so knocked up https://gist.github.com/archgrove/6766129 https://gist.github.com/archgrove/6766129, https://github.com/archgrove/PDFProofSheet https://github.com/archgrove/PDFProofSheet, and a bunch of shell scripts to generate a folder of images. Fed this lot into iMovie, and voila.
clarkmoody 12y ago
While you mention that you are hesitant to release the content of the dissertation prematurely, could you instead release the build tools you used to make your "flatplans"? It would be a huge help to those of us about to write our dissertations :-)
- hds 12y ago
  The code is horrible, but it's here: https://github.com/hds/repo-flatplan-animation https://github.com/hds/repo-flatplan-animation You'll find the JS bits in the repository for the website which is there too.
mcguire 12y ago
Forget the dissertation. Nobody really reads those (really, not even my committee). What we want to watch in real-time is the defense. Especially the spanking tunnel at the end.
apepe 12y ago
You could do it in Authorea (Git+LaTeX) and actually show us what you are writing too!
muhfuhkuh 12y ago
This is really cool stuff. Reminds me of a stripped down version of notch's live coding sessions for Ludum Dare. Any reason you chose LaTeX as your content dev language and publishing over something a little "easier" like DocBook XML and Publican (which uses LaTeX as an intermediate build language to create PDFs)? Is it required, what you're used to, or purity of notation?
- SuitAndThai 12y ago
  It's probably because the author needs to describe complex mathematics and DocBook is unable to describe mathematics[1]. I can't speak much about Publican however. 1. http://www.sagehill.net/docbookxsl/Math.html http://www.sagehill.net/docbookxsl/Math.html
- hds 12y ago
  Partly it's what I'm used to, but I've yet to find anything "easier" for the large amount of mathematics I typeset. There is a lot of notation and very little graphics or difficult formatting.
- irremediable 12y ago
  I think LaTeX remains the go-to language for this kind of stuff. Certainly, every STEM academic I've worked with has used it.
- j2kun 12y ago
  There's nothing more suspicious than a math paper written in something besides TeX.
- jisterdev 12y ago
  'content dev'. yuck.
CaRDiaK 12y ago
Why am I obsessing over the arrow not quite lining up on the points of time? Until the fourth tick, then following some pattern..
Monkeyget 12y ago
Related, I loved watching pg writing an essay live : https://code.stypi.com/hacks/13sentences https://code.stypi.com/hacks/13sentences

56 points by hds 12y ago

I was expecting to see the ten-year process involving teaching assistantship, employment, marriage, and children, all illumninated by the light of a candle at 2 am in a basement office next to the crib.

ygra 12y ago

I wrote something similar while I was writing my diploma thesis, albeit not that visual. For me it was handy to see whether I was making any progress, so I regularly updated a CSV file with the current number of pages, lines, words, figures, code lines, and code comments. The idea was that on every day I did something I could at least look at the data and see how much (also the script automatically adjusted page, line and word count for the appendix which I didn't want to count).

sanderjd 12y ago

That's really neat! I'm not sure if this is intentional because your work is private, but I would love it if I could click on the images or SHAs to go actually see the thing.

hds 12y ago
At the moment I'm a little hesitant to make it public. Partly because there's a large chunk of unpublished research in there, but also because some of the early drafts are a little embarrassing. Perhaps when the thesis is deposited (and hence public) I could add that in for "historical" interest.
- sanderjd 12y ago
  Yeah, totally understandable.
- j2kun 12y ago
  Definitely don't make it public. I know people who have had their thesis topics scooped.
Braunbart 12y ago
Yep, I instinctively clicked on one of the pages to see it bigger.

burritofanatic 12y ago

Excellent! People writing important papers really should be using Git. When documents are extremely complex, previously removed portions can have pertinence later.

Fomite 12y ago
This doesn't work particularly well for fields not using LaTeX as their document creation tool.
- dovel 12y ago
  Why is that?
  - jaredsohn 12y ago
    I would think the GP means that git works much better if you are creating a document in a non-binary format (of which LaTeX is one example.)
irremediable 12y ago
Seconded -- at least for anyone using LaTeX or another git-friendly format. I used git for my masters dissertation recently, and it was very useful to be able to checkout old versions and compare them with its then-current state.

toppy 12y ago

That's an excellent solution for big problem in my country (not at PhD level though) namely plagiarism. Make students to use such system, turn autocommit on and you can see their progress and all deviations (big chunks of text appearing from nowhere etc.).

doctorKrieger 12y ago
once they learn about it nothing will change.

archgrove 12y ago

Whilst not as slick as the JS version, I did much the same thing with mine: https://vimeo.com/76969179 https://vimeo.com/76969179 when bored one day. I couldn't find any tools to create exactly what I wanted, so knocked up https://gist.github.com/archgrove/6766129 https://gist.github.com/archgrove/6766129, https://github.com/archgrove/PDFProofSheet https://github.com/archgrove/PDFProofSheet, and a bunch of shell scripts to generate a folder of images. Fed this lot into iMovie, and voila.

clarkmoody 12y ago

While you mention that you are hesitant to release the content of the dissertation prematurely, could you instead release the build tools you used to make your "flatplans"? It would be a huge help to those of us about to write our dissertations :-)

hds 12y ago
The code is horrible, but it's here: https://github.com/hds/repo-flatplan-animation https://github.com/hds/repo-flatplan-animation You'll find the JS bits in the repository for the website which is there too.

mcguire 12y ago

Forget the dissertation. Nobody really reads those (really, not even my committee). What we want to watch in real-time is the defense. Especially the spanking tunnel at the end.

apepe 12y ago

You could do it in Authorea (Git+LaTeX) and actually show us what you are writing too!

muhfuhkuh 12y ago

This is really cool stuff. Reminds me of a stripped down version of notch's live coding sessions for Ludum Dare. Any reason you chose LaTeX as your content dev language and publishing over something a little "easier" like DocBook XML and Publican (which uses LaTeX as an intermediate build language to create PDFs)? Is it required, what you're used to, or purity of notation?

SuitAndThai 12y ago
It's probably because the author needs to describe complex mathematics and DocBook is unable to describe mathematics[1]. I can't speak much about Publican however. 1. http://www.sagehill.net/docbookxsl/Math.html http://www.sagehill.net/docbookxsl/Math.html
hds 12y ago
Partly it's what I'm used to, but I've yet to find anything "easier" for the large amount of mathematics I typeset. There is a lot of notation and very little graphics or difficult formatting.
irremediable 12y ago
I think LaTeX remains the go-to language for this kind of stuff. Certainly, every STEM academic I've worked with has used it.
j2kun 12y ago
There's nothing more suspicious than a math paper written in something besides TeX.
jisterdev 12y ago
'content dev'. yuck.

CaRDiaK 12y ago

Why am I obsessing over the arrow not quite lining up on the points of time? Until the fourth tick, then following some pattern..

Monkeyget 12y ago

Related, I loved watching pg writing an essay live : https://code.stypi.com/hacks/13sentences https://code.stypi.com/hacks/13sentences