9 ms·
I was the project lead on this, which involved converting the Feynman Lectures from LaTeX to HTML. I'd be happy to answer any questions. Update: Most of the qu
by mhartl 13y ago
I was the project lead on this, which involved converting the Feynman Lectures from LaTeX to HTML. I'd be happy to answer any questions.
Update: Most of the questions center on cost. I've answered in more detail below, but the short version is simple: no off-the-shelf converter was remotely sufficient for our needs, so we had to write lots of custom software, and writing custom software is hard.
- lvryc 13y agoTrevyn, above, mentioned, > The conversion from LaTeX to HTML was expensive: we raised considerable funds, but ran out before finishing Volumes II and III, so we are only posting Volume I initially. I don't know too much about digitalisation, but why does it cost so much?
- mhartl 13y agoThe LaTeX source of the Lectures is long and heavily formatted, so the conversion to HTML involved writing a large amount of custom software, including lots of tests to catch regressions. In other words, this was a speculative software project; as anyone who has experience with such projects can tell you, it is virtually impossible to anticipate all the problems you'll run into. (The costs would have been even higher had I not worked for a reduced rate, often pro bono, because I wanted to help make the Feynman Lectures accessible to a wider audience.)
- npsimons 13y agoI am very grateful for your work, and I hate to pick on details, but surely you reused much of what's already available for LaTeX? I hope you didn't spend much of your time reinventing the wheel. I freely admit that I am probably vastly underestimating the amount of work that went into this, reuse or no.
- mhartl 13y agoWe did make use of as many existing resources as possible. If you take a look at the [Ruby on Rails Tutorial book](http://railstutorial.org/ http://railstutorial.org/) and [The Tau Manifesto](http://tauday.com/tau-manifesto http://tauday.com/tau-manifesto), both of which are converted to HTML from LaTeX, you'll see that I have deep experience in such conversions. Combined with my previous role as Caltech's editor for the Definitive and Extended Edition of the Feynman Lectures (and my Caltech physics Ph.D.), this made me uniquely qualified to work on this project, which is the main reason I did it. (It certainly wasn't for the money; the opportunity cost of the time I spent on this project was high.) Unfortunately, there's a vast gulf between "kinda-sorta mostly converted"—which is the best you can get with off-the-shelf converters—and "a faithful representation of the Feynman Lectures". Closing this gap required a large amount of gruesome work.
- vermontdevil 13y agoWe should actually do a crowdfunding effort to pay you for your hard work for doing Volume I and for future volumes. Really really appreciate this effort. I've sent this to some of my deaf friends who are studying or trying to get into Physics and they are over the moon about this.
- edanm 13y agoI would absolutely pitch in if someone organised a Kickstarter campaign to raise money for mhartl, even just to pay him back for the current efforts. I'm ecstatic that the lectures now exist online!
- npsimons 13y agoThis was one of my first thoughts; I know it's oversimplifying it by a wide margin, but there is a latex2html program in existence.
- mhartl 13y agoUnfortunately, there's no general mapping from LaTeX to HTML, so for any realistic pre-written source you have to write many custom modifications to complete the conversion. This is a time-consuming and error-prone process, involving a large amount of custom software.
- lutusp 13y agoThere are various ways to do this, none of them perfect. Online searches suggest that "htlatex" (for HTML output)and "xhlatex" (for XHTML output) are pretty good: http://tug.org/applications/tex4ht/mn-commands.html http://tug.org/applications/tex4ht/mn-commands.html
- magottlieb 13y agoHello, all. Many people seem to have gotten the impression from my letter that because we ran out of funds paying for the conversion of FLP Volume I from LaTeX to HTML we need additional funds to complete the conversion. However, that is not the case because some chapters of Vols. II & III were also converted for us (though imperfectly), and after comparing the Vol. I HTML to the original LaTeX, I found I could do the conversion myself quite easily using parametrized regular expression search and replace operations and a minimum of hand-editing. I am able to convert one chapter of FLP from the LaTeX to HTML in about 2 hours. I have been working on completing the remaining Vols. II & III chapters in my spare time, and they are very nearly done. Only a few chapters remain to be converted, and a few others only need some HTML tables. We are planning to make a second release when our two PDF editions (formatted for the desktop and for tablets) are available for sale online. In fact, the PDF editions have been ready for a long time, but the publisher is having some difficulties getting them into their retailers' sales queues. Today we have the first evidence of the publisher's success in resolving the difficulties: https://www.cafescribe.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=10001&categoryId=152754&langId=-1&productId=4000000000003036262&storeId=216405&productStoreId=216405 https://www.cafescribe.com/webapp/wcs/stores/servlet/Product... So, hopefully it will not be long before we add the completed Vols. II and III chapters to the published HTML edition, with the few remaining chapters to follow shortly. Michael A. Gottlieb Editor, The Feynman Lectures on Physics www.feynmanlectures.info
- tzs 13y agoOne obvious question: why not use crowd funding to pay for finishing converting volume II and III?
- mhartl 13y agoThat's worth thinking about. One issue is that it's hard to tell how more more money is needed to finish it. Alas, custom software is rife with uncertainty. (Practically speaking, this means that, e.g., custom hack #37, needed to complete Chapter 17 in Volume I, might not work for Chapter 38 in Volume II, etc., etc. It's a long, hard slog.)
- WesternStar 13y agoAsk for a million dollars. Its worth it to humanity.
- zellyn 13y agoHave you thought about putting the source and scripts into a github repo, and taking community contributions? I'm sure people would pitch in to get later volumes converted.
- mhartl 13y agoI'll look into it, but unfortunately this isn't up to me. Convincing the publisher to let us put up an online version was hard enough.
- ColinHayhurst 13y agoNo questions but a BIG thank you
- mhartl 13y agoYou're welcome! I appreciate the kind words.
- baiitsu 13y agoFirst of all a big thank you for the outstanding work! Does Caltech have any plans to release the LaTeX source? And would a PDF or EPUB be made available (free or otherwise)?
- mhartl 13y agoThanks! I don't know what their plans are. There is PDF version, but I can't find it anywhere (well, not a legal one). I'll look into it.
- mhartl 13y agoIt turns out the publisher still has authorized PDF sales. I'm not in line to get royalties, so it's not my problem, but I still hate to see a missed opportunity.
- mikevm 13y ago> It turns out the publisher still has authorized PDF sales. Did you mean "has not authorized"? Otherwise, I would gladly pay for PDF editions of the 3 volumes. If they already went with producing a free online copy of the 1st volume, I don't see why they should be against selling official eBook PDF editions for people like me who prefer to read this offline. Anyway, great job on the conversion, it looks marvelous!
- mhartl 13y agoSorry, I meant to say "still hasn't authorized".
- pseut 13y agoFrom browsing around, it looks like an exceptionally well done conversion. Thanks for doing it.
- michael_nielsen 13y agoThanks for doing this, Michael, it's awesome that this is now available! I notice the material is copyright Caltech. Any chance that they'd consider releasing it under an open license? A CC-BY license would enable this material to be repurposed and improved in all kinds of interesting ways.
- mhartl 13y agoThanks. The Feynman Lectures IP is a treacherous minefield, and it took a lot of effort even to get this far. I'd love to see a Creative Commons license, but I'd be surprised if it happens. I try to emphasize to the interested parties that, by doing things like giving away online content and using CC licenses, they stand to make more money (if they execute right), but it's a hard sell to people who don't live on the bleeding edge like we do. And since I don't have equity in the project, my efforts in this direction are strictly pro bono.
- ma2rten 13y agoMaybe we could crowd-fund this?
- ISL 13y agoWow. CC-BY-ing the Feynman Lectures would have a meaningful impact on undergraduate education worldwide. Where I'm sitting in our lab, I'm probably within 50m of at least ten copies. CC-BY-ing just a few of the old stalwarts (I like the 1970 Sears and Zemansky, for example, and old editions of Thomas' Calculus are fine (probably better than present editions) ) would stop a lot of expensive silliness with the generation of open-source textbooks. Buying out the rights to the best textbooks is in our national interest.
- sillysaurus2 13y agoHow were you able to convince the publisher to let you put up an online version for free? Also, please collect data regarding whether this harms book sales. If it turns out not to have any significant effect, then this could be cited as evidence in any future circumstance where someone is trying to convince a book publisher to approve of releasing an online version of a book.
- adamgravitis 13y agoThis is outstanding. Thank you for the effort. Has there been any consideration of making a semantic version of the equations involved? The great thing about having this on the web, is I could easily see myself maintaining links in long calculations to various formulas as the authority. (Also: what's the font in the SVG images? It's great!)
- zodiac 13y agoIf you're asking about the equations etc, the site uses MathJax to render LaTeX.
- yiyus 13y agoYou've done a great job capturing the exquisite design of the paper edition. When I saw the title I was expecting jpg scans of the equations, or something worse, and I was surprised of finding something Tufte would be proud of. Thank you very much.
- radicalbyte 13y agoAwesome work, I bought the book versions with my first paycheck, only to sell them on years later. I never read them :'( Now I've just read the first two chapters online, and I have a question. Have the lectures been updated to 2013 physics, or should I read comments like this: "We seem gradually to be groping toward an understanding of the world of subatomic particles, but we really do not know how far we have yet to go in this task." (last sentence in chapter 2) ..as being the state in the 1970s? If so, what has changed then in the "big picture"? String theory?