The highly involved build process for my doctoral dissertation

https://spwhitton.name//blog/entry/dissertmake/

by pabs3 • 3 years ago

66 55 3 years ago

Does anyone know any library that can swap a page in the pdf with another without destroying the metadata?

analog312• 3 years ago

I use a Python library called pdfrw to take PDFs apart, notably to work with sheet music. I think when you ask it for a page, it returns the whole thing, and the PDF files that get written are still readable PDFs.

I assume that within the file structure are things that enable an internal integrity check, such as maybe checksums or CRC's, which would necessarily be different if the file contents are different.

cozzyd2• 3 years ago

qpdf?

clankyclanker• 3 years ago

pdftk?

agarsev• 3 years ago

You can dump the metadata and then restore it, maybe that'll work for you?

cozzyd2• 3 years ago

I too had a complicated Makefile for my dissertation, but since then I've learned to use latexmk which seems to more or less do everything for you...

(ok, I guess I still often have a Makefile that calls latexmk and sometimes produces some figures from scripts)

tkiolp4• 3 years ago

I’ve known PhDs in philosophy who can barely create slides in power point… while others (like OP) go and use latex, create open source software and use emacs. I’ve never seen such a drastic difference in any other field.

KeplerBoy2• 3 years ago

Latex is the norm in all natural sciences and engineering fields. Most people however are happy with putting their content in someone else's templates.

progbits2• 3 years ago

Honestly I wish it was even templates, usually it is just default Word document with random extra whitespace and hand-written section references that get out of sync.

something98• 3 years ago

As someone who has degrees in both, I can firmly say this is wildly incorrect.

jacurtis• 3 years ago

I'm not sure how its possible to get a PhD in the modern era without learning LaTex. It is the required markup and publishing tool for any respected research publication. Since PhDs require (at least mine did) multiple published works in respected journals and conferences, I feel like LaTex is table-stakes for doctoral research and degrees.

Furthermore, outside of my dissertation work, the classes I took often required you to produce other research works which also had to be created with LaTex. Overleaf (a SaaS platform for Latex collaboration and sharing) was essentially the operating system of my PhD. I spent 75% of my time in there.

I feel like it would be impossible to get a Doctoral degree without LaTex knowledge. Maybe it depends on the university?

rnadomvirlabe• 3 years ago

Microsoft Word still reigns supreme in many fields. Some journals still require it as the format for submission.

svara2• 3 years ago

This is all completely untrue in general. There are many fields in which you'll be the annoying weirdo who tries to force their tools on others if you insist on using latex.

For example, Nature is so kind as to accept final submissions in latex, but they'll convert it to Word. So it's completely pointless.

Frankly people using latex-beamer make me roll my eyes too. It's a sign you care about tools more than outcome. If you're making a presentation, your goal should be to produce the ideal design to convey your message to the audience. Latex always gets in the way of this. (Some latex wizards are going to disagree obviously, but for the vast majority of users it's true.)

shpongled• 3 years ago

Maybe in CS?

In the life sciences, practically no one knows how to use LaTeX, or has even heard of it.

bigbillheck• 3 years ago

How many graduate students in other disciplines did you talk to?

ghodith• 3 years ago

In my experience most grad students mostly just end up doing whatever their advisor is most familiar with.

jszymborski• 3 years ago

I'm not sure how its possible to get a PhD in the modern era without learning LaTex.

I have a fairly complex Pandoc pipeline so that I can send my supervisor Microsoft Word documents and create LaTeX documents all at the same time.

Which is to say that I am swimming upstream to get a PhD in the modern era using LaTeX. It would be far easier if I just used M$ Word.

2cynykyl• 3 years ago

I need to disagree with this strongly (I'm an engineering professor of 15 years). I am so "anti-latex" that I border on zealot. I'll save you my sermon, but I can't help but point out that saying "I spent 75% of my time in there" (ie Overleaf) is a pretty obvious "red flag". Did you ever consider that using Overleaf may have hindered your phd productivity? Shouldn't you have spent 75% of your time in the lab, or your IDE, or the field? Does this statement not make you pause and reflect that maybe using an archaic and arcane typesetting language as a "word processing tool" was a trap? The time people need to spend with the formatting and markup in latex rather than the actual creative process is at the top of my many complaints about latex.

I only hopped on this thread to say this so that I could add to the evidence against using latex in case any grad students are reading this and considering using latex. It is absolutely NOT necessary and is in NO WAY helpful. Please leave the type-setting to the publishers and focus on knowledge and content creation.

amelius• 3 years ago

Latex coding is the favorite bike shedding project of many CS students.

ezequiel-garzon• 3 years ago

On the technically advanced end of the spectrum you'll find John MacFarlane [1], professor of philosophy at Berkeley and creator of pandoc [2]. Some people are just amazing.

[1] https://johnmacfarlane.net/

[2] https://pandoc.org/

slow_typist• 3 years ago

Incredibly talented guy. Incredibly useful software.

nohuck13• 3 years ago

"goals: ... replace the second page with a scanned copy ...after it was signed by the examiners... reproducibly."

Oh man, purist side quests like this are exactly how I would procrastinate a philosophy dissertation. But building a complex process for _reproducibly_ handling the once-in-a-lifetime event of accepting your dissertation examiners' signatures is taking it to a whole other level.

anonymouskimmer• 3 years ago

For my M.S. I just looked up the relevant PDF merge utility for Linux online. I can't even remember what it was now, but it took the first x pages from the unsigned thesis, intercollated the signed signing page, and then took the remaining pages from the unsigned thesis, and output a single PDF.

Edit: .zsh_history had it:

pdftk A=thesis.pdf B=approval.pdf cat A1-5 B1 A7-end output signed_thesis.pdf

heyoni• 3 years ago

How have you never cleared out your history? Or do you rarely use your terminal?

/edit not cleared but filled up so that it gets rotated out

cinntaile• 3 years ago

This doesn't really work as intended when I tried to do something similar. Pdftk produces pdfs that were much bigger than the ones generated by LaTeX and you might end up outside of the allowed size range.

earthboundkid2• 3 years ago

I spent my whole grad school procrastinating by working on a system for publishing my papers, which turned out to be the only marketable skill I got from my PhD, so I wouldn’t knock procrastinating.

earthboundkid2• 3 years ago

For anyone who cares: https://github.com/earthboundkid/Watsuji-and-Aesthetics/blob...

lapcat3• 3 years ago

True story: While writing my philosophy dissertation, I dropped out of grad school to become a Mac developer, and a key factor was the open source Mac bibliography app BibDesk. https://bibdesk.sourceforge.io/

dmd2• 3 years ago

Five years seems pretty normal? Shorter than that would be extremely unusual. What are you saying here?

lapcat3• 3 years ago

I think you're confusing time spent in the doctoral program vs. time spent writing the doctoral dissertation?

This appears to be your résumé, according to which you spent five years total in the doctoral program. I'm pretty sure you didn't start writing your dissertation on day one, since there are years of coursework and other requirements before you start writing your dissertation. https://www.linkedin.com/in/danielmdrucker

joosters• 3 years ago

IMO, LaTeX is just a way for computer scientists to keep on programming when they've finished their research project and are writing it up.

rubicks• 3 years ago

It's also a great way to keep on programming when you need a CV and cover letter.

1letterunixname• 3 years ago

It's a clever way to foist effort on all other fields and enable research paper procrastination by playing with a complicated publication toolchain. Or is it an intelligence and quality barrier-to-entry filter to reduce the number of papers that need to go through review? ;@]

ribit• 3 years ago

My dissertation building process was similar. Only it also involved a bunch of custom pandoc filters and the markdown had runnable R code in it. Fun times. I didn't bother with replacing the signed page though, because enough is enough.

progbits2• 3 years ago

So, why not use \includepdf{the printed, signed and scanned page} in the .tex file? You can do \ifdefined to have it generate both the initial draft for signing and then second iteration with the scanned page. This should preserve document metadata as far as I can tell.

hnarayanan• 3 years ago

This prompted me to pull the code of my 15+ year old dissertation[1] and see if it still builds.

Still does!

1: https://github.com/hnarayanan/phd-dissertation

pimlottc• 3 years ago

PDF metadata is interesting, there’s a lot of fun stuff you can do like custom page numbers, sections, page sizes, tables of contents, internal links, OCR overlays…

But the format and syntax for these can be tricky, and tools and documentation can be hard to find.

BrandoElFollito2• 3 years ago

PhD thesis are overrated.

People create some kind of life achievement monument where they seek to summarize the world knowledge and yet nobody is going to read the document. Same for MSc thesis.

What matters (to a point) are your papers and it should be enough to just gather the in a file, add a conclusion and call it a day. We have this in France and some people use the opportunity.

My PhD thesis (physics) was 45 pages long. The expected number of pages was about 200 or 300, a friend of mine wrote two tomes.

The introduction was basically "if you do not know about the topic you should probably to read this thesis anyway but if you really want to, here are some books you can use". 2 lines, introduction done.

The main part was mostly copy/paste from my articles.

The conclusion was that the results were great and groundbreaking.

I added a special section called (I think) "Is this thesis really useful, a discussion with the author" - where I pointed out everything wrong with what I did and where to go further. 75% of the reviewers loved it (less work for them) and 25% were wrong :)

Finally, the acknowledgments were thanking my thesis director profusely (he did not help me much in the research part but eased considerably (considerably) my life and helped me to navigate the muddy waters of Academia), a fried who during a dinner gave me an idea that changed the whole thesis, my wife, my parents and my dog.

The defense was epic (3 hours instead o the expected 45 minutes) because some people decided to manage their political battles there. I was switching between grabbing the popcorn and mopping my forehead when some questions out of the end of the universe were asked (one of them was asked by the dept asshole in residence (AIR) and I understood all the words, but not put together in one sentence (nothing to do with my topic). The president of the jury waited for me to answer, I mumbled something, he said that this was disappointing and asked the AIR to answer. He said that he has no idea but thought I may.)

I will go one day to my dept (I left academia in the meantime) and ask if someone has ever borrowed the thesis (it exists only in one copy).

yummypaint• 3 years ago

Im an experimentalist and I'm constantly looking at the online copy of my thesis to grab measurements of the facility, setup info, formulas, etc. I also look at theses of other people for the same reason. It's generally more convenient than going through accelerator documentation.

BrandoElFollito2• 3 years ago

Good for you - it means that your thesis is one of these that get referred to.

I did not need that because most of what I was writing about was hardly documented (neural networks and genetic algorithms in the '90s) but I can imagine (having done my thesis at CERN) that having all accelerator and detector information in one place is useful.

The highly involved build process for my doctoral dissertation

Related Stories