Hacker News new | past | comments | ask | show | jobs | submit login
Free Textbooks from Springer, Categorised (hnarayanan.github.io)
542 points by hnarayanan on June 14, 2020 | hide | past | favorite | 115 comments

This sort of thing literally brings out the starving aggressive animal in me. I get so frantic trying to download all I can!

Very, very surprising to me, one of the free books turns out to be directly related to my new job (like, my boss's PhD advisor wrote the book).

Can I introduce you to library genesis? Or, if you don't like piracy, upenn's online books Page?

That would definitely be a form of Nerd Sniping[1]

[1] https://xkcd.com/356/

UPenn in particular wow great site thank you.

And one was written by a mentor of mine who passed away shortly after completing the book. :(

Ya. I'm embarrassed by my hoarding behavior. Took me a decade to let go of my CDs, even though I had stopped listening to them much earlier. I've only managed to cull ~1/2 of my physical books.

My mitigation strategy for ebooks is quickly adding to my bookmark's "Books" folder and close the tab.

I used to be much more of a packrat than I am now. I've slowly been winnowing out stuff liked used computer equipment and enough books so that they all fit in my existing bookshelves (more or less). And I've slowly been digitizing more and more things.

I'm probably still at least something of a digital packrat but a lot of that is it's just easier to hold onto stuff than expend the mental energy to figure out what to keep and what to delete.

> This sort of thing literally brings out the starving aggressive animal in me. I get so frantic trying to download all I can!

This feeling is not useful, unless you follow through with your initial impulse and go through all the books.

I feel like this feeling is partially what fuels our unsustainable consumerism and growth; we all have it, sadly.

I don't think it's necessarily bad -- I collect free ebooks when I can, even if I'm not interested in reading them now but think I might someday. Then, later if I'm thinking about learning Rust (for example), I can go to the free book collection of mine and see if I have anything, then read through them.

I've done this a few times. Even if I only get value out of 20% of that collection, I wouldn't have known in advance what 20% I'd eventually be interested in.

And I do still pay for physical books too (but I don't stockpile those, and buy them when I actually want to read them)

Problem is that you rarely find good books this way so you have pile of books which are not worth reading. Determining which books are good takes a lot of time. This is reason I have stopped hoarding books from cheap or free sources. It is way better to download/buy/pirate books recommended on the internet than trying to find them myself.

A reasonable strategy for me is: 1. Trawl through a resource like this when I come across them 2. If one looks interesting, do a quick google search of "[sub-category] recommended texts for [my-level] level" 3. Buy or Pirate the top recommended text

Resources like this serve as good inspiration to learn about something new.

I can see your point. From my perspective, the "let's collect it because what if I need it later" mindset is not wrong at all times, but I think we're selectively blind to those 95% in which case it's a miss (i.e. in our case, you'd download an ebook and forget you even have it in like a week).

In ebooks, this doesn't really create a bad outcome, but I think this mindset can manifest in other aspects of our lives as well and that's where it's very much less desirable.

It was just an offtopic remark uttered with no judgement, but it's something I've been sensitive in my own attitude.

I collect every book that I come across, and I can't count the number of times it's been helpful (one month, or five years later). It doesn't matter if you can't remember the book after a week, fzf is here to help you. If you got interested in a new topic (like the aforementioned Rust), press Ctrl+T and type "rust".

Or you can suppress the impulse and skip getting the book, then either forget the site altogether, or visit it five years later and find that it got shut down four years ago.

For every one site that shut down, there are a dozen replacements.

Not legal.

There have been plenty of instances where you could get lifelong legal ownership of something very valuable only if you got it at a certain time, and that time often ends abruptly without warnings.

Umberto Eco had quite a library of unread books.


eco walking in his personal library https://www.youtube.com/watch?v=bF9tG5Q6NTA

And I thought that my ~1000 e-books were a record...

1000 books is really not much. After culling, I have 1000 books on my read shelves and easily 200 or so unread books. And that's just what I have in print and doesn't count the music books or the rare book shelf or my wife's two bookcases of print books.

You're right indeed, maybe in the e-books side it's rarer...

Anyway, my wallet felt it, considering I've done most of it in about 4 years with a relatively modest salary

After I got my first job, I went on a bookstore crawl through Santa Monica back when the Third Street Promenade was full of independent shops and bookstores. I came home with a car full of books and a thousand-dollar credit card bill. It's easy to forget that your income is finite when you're young and foolish.

My personal story instead is that I had always been convinced that books were useless in software or web development, that everything you needed (and had to do) was read the specifications, standards and official manuals (I dropped out of college very early).

After a few years, when trying to figure out a new thing, I conceded to turn to some book and a world opened.

Along with the discovery of O'Reilly's, and then other publishers', late shop of drm-free PDFs (which also solved the serious space problems I would have had with paper books), I went on a e-books binge, from software craft to every conceivable CS field.

I am scared to estimate how much I spent, but I'm reasonably glad of what I learned and what I have at hand.

I actually agree it isn't that useful. At least for me. My attitude is that high quality sources are worth more than low quality ones.

Thank you! Here's your list as a text file, which may be helpful for people who like text.


Thanks for posting this. I imported all these books into a collection at LearnAwesome.org: https://learnawesome.org/users/8a16a2e4-dcb7-4167-a2a2-51d3a...

In the math books, a gem is "Proofs from THE BOOK" [1].

[1] https://link.springer.com/book/10.1007%2F978-3-662-57265-8

I just lost my morning skimming through this book. It looks like there are a lot of other good books among this collection.

It won't be possible to go through all, but are there are any recommendations from HN? Personally, I have come across Elements of Statistical Learning[1], Recommender Systems[2], The Algorithm Design Manual[3] in many recommended lists.

1. http://link.springer.com/openurl?genre=book&isbn=978-0-387-8...

2. http://link.springer.com/openurl?genre=book&isbn=978-3-319-2...

3. http://link.springer.com/openurl?genre=book&isbn=978-1-84800...

If I'm not mistaken Elements of Statistical Learning and all other books by the authors are already available for free on their website


I can second a recommendation of the Elements of Statistical Learning. It's well presented, self contained and features good exercises. It may also be slightly different in presentation from most books on ML because the authors are all professors of statistics, instead of CS. (But, it is NOT a classical statistics textbook.)

My cousin's husband who is a math professor asked for an earlier edition of this as a Christmas gift in our family's secret Santa exchange while I was in grad school. Since I had drawn him that year, when I saw this on the list I ordered two copies since it sounded cool to me too. Parts of it are still over my head, but it's great reading.

Somewhat off topic, but I've never had much luck with digital textbooks. For some reason dead tree makes it much easier for me to actually pick up a book and start working through it. Does anyone else have this problem? :-/

I had that too until I installed Zotero, its zotfile extension, and a PDF reader that lets you highlight selections and add notes.

Zotero then has "right click -> extract notes", and you get the stuff you highlighted in a new document, with links back into the PDF.

That has completely turned around the experience of reading PDF textbooks for me, it's such a neat feature.

Which PDF reader do you use for that?

I occasionally scribble on PDFs using my reMarkable. But it's a bit too slow (eink) for my needs.

I use Okular for that. It's open-source and present on three major platforms. I really got surprised to find out how feature-rich it is. It has a lot of basic/important features - like highlighting, underlining, strike through, inline notes, pop-up notes, freehand line drawing, inserting shapes (like arrow, rectangle, polygon) etc.

Tip: you actually don't need to install kde toolkit to be able to run Okular on Windows and macOS. Search for nightly builds - I'm on a nightly build and it's pretty solid.

Worth mentioning that Okular (alongside other KDE software, Kate and KStars) is on Microsoft Store.

Unrelated, but do you have the first or second gen. reMarkable? I've been lusting over the 2nd gen since it was announced, but $500 for something I'll use fairly infrequently seems hard to justify

First gen. And to be fair, I use it so infrequently that I've repeatedly considered selling it. It's pretty decent, it's sturdy, the writing experience is very pleasant. But the software is clunky. It doesn't support USB file transfer, instead it implements a webserver over USB so you can drag files to it using a browser ... it's a pain.

I use the free version of "PDF-XChange Editor", under Windows. No idea if that is the best choice, but it works fine for me.

Yes. There's something about ergonomy of a digital textbook. It doesn't matter whether it's on an iPad, computer, or my Kindle. They all make it difficult.

I realized though that it might be due to the fact that we do so many things on the computer. I know how many websites are just a click away. It takes a certain mindset, but typically, after 10-15 minutes, I can focus enough on the materials that even the computer is fine.

It just takes practice and focus, the latter being increasingly scarce in today's world of notifications.

Same here ... for technical textbooks that I really need to focus on, dead tree is better. Esp if the content is math-y, scribbling notes, underlining etc seem to amplify learning.I bought a printer just for printing out such content so I can read and scribble away.

I wonder if this is something people of all age-groups universally feel, or do kids today have no problem grokking hardcore technical details from ebooks.

> I bought a printer just for printing out such content so I can read and scribble away.

You might want to buy a cheap hot glue binding machine as well, or use the one at your local library or university.

That is a good suggestion, thanks, but usually I am very particular about printing specific pages: the part of me that doesn't like reading such texts on a device is typically at loggerheads with the part of me that doesn't like print paper ending up in recycle bins :-) ... its a very real struggle ...

Anyone here with experience with Remarkable? Maybe thats the answer?

remarkable v1 user here. The pdf/epub reading experience is not perfect, turning page is a bit distracting as the screen refreshes with all sort of artifact for nearly 1 second,but the can't beat the ink factor. And can annotate. I prefer it to the kindle to read because it's larger.

Thank you for the helpful review!

Yeah I don't mind reading "normal" books on my ereader, but for technical books and textbooks, I much prefer dead tree. Probably because it's a huge pain to browse and flip pages in ebooks.

You just need to find the right digital device for you. I much prefer reading on my iPad Pro 12.9 inch to dead tree form. The only disadvantage is that I cannot have 2 books open at the same time easily :-)

For me it’s mostly the commitment I feel when physically buying something that gets me started.

Most of the time I still end up reading the digital copy on my Laptop, as I personally find this more comfortable for textbooks.

I got a $100 black and white laser printer, with the ability to print double-sided, in 2011. One of the best purchases I ever made.

I did the same thing. So awesome. Also wrote a script to print 4 pages on each side. Saves a ton of paper, and makes it easier to absorb.

I used to have it, but it gets easier with time.

Yes, definitely (apart from the weight of a big one).

We still await a killer e-ink device to read them on - even the appropriately big-screen ones are still far too slow ux-wise - but even that would not quite be as good as having it to flip back and forth in your hands.

I think that's one of the important differences between textbooks and fiction or articles. The latter are normally a sequential read so a bog standard Kindle or equivalent is ideal for them; textbooks you need to be more interactive.

Tablets (ipads, etc) obviously have some of that, but have an inferior pure reading experience to e-ink/paper.

Still paper for me for textbooks as well.

I'm totally with you on this. I think it has to do with how I remember and retain information as I learn. I subconsciously use contextual, visual cues to remember where some piece of info was to refer back to ("that one figure in the middle of a page in the section with a giant blue block of text"). For some reason, I completely lose this when learning via digital textbooks. I think it may have to do with the need to zoom in to really read anything properly.

Perhaps it's an idea to tilt your monitor and put it into portrait mode.

My guess is that it is more than just visual clues on the page and facing page currently visible. I suspect it is also the 3D layout of the pages in space, and how that changes as you go though the book.

With a book, it feels different when reading material at different places in the book. Even if I'm not going through the book linearly but instead looking up things in the index and jumping to them, I get a sense of where things are in space. Later, if I'm trying to find something again I can usually go to close to the right place without bothering with the index.

I suspect that the reason for this is is related to why/how the method of loci [1] (AKA memory palace) works. We are really good at remembering spatial relationships, and the book lets us tap into that.

[1] https://en.wikipedia.org/wiki/Method_of_loci

Good point, but wouldn't an always-visible scrollbar have the same effect?

It might. I'd expect that the effect would be stronger with a physical book because more senses are involved. With the physical book you not only have the visual difference in how the book looks depending on how far into it you, you also can feel a different weight distribution.

Digital books sure don't smell as good as a brand new physical book. Don't underestimate the animal part of your brain. For anything.

But nor do the web pages people tend to read more these days than they read books.

> Does anyone else have this problem? :-/

This is not a problem! I also prefer dead trees, I just like the experience of holding and reading them. For technical books, a digital copy is useful also to find all occurrences of a word (especially on older books with incomplete indices).

It's a problem because dead tree textbooks cost hundreds of dollars.

I would say that they are well spent. Never regretted buying too many books.

For many students it's not a choice, especially outside rich countries like the US.

Sure. As much as I love physical books, library genesis is a godsend for me and for millions of students all over the world!

I have a cheap Surface that I bought at a promotion, with a pen. I can scribble/underline and it has a 3:2 screen ratio. With some discipline it is a dedicated device for a very specific kind of work (reading Math books and using LaTeX in WSL).

Yes, me too. Give me something I can hold in my hand and read with a cup of tea in a completely distraction-free environment any day.

(The UX of e-book readers may get there one day, but it's not there yet.)

Kudos for the author for making the list more accessible!

> and expects you to download some Excel sheet to figure out what they have on offer

TBH, Excel isn't all that bad UX; the list is simultaneously more complete, more dense and faster to read than the web equivalent. I'd love it if there was an element in HTML for displaying modern data tables directly in the browser (with the ability to sort and pivot them).

It amazes and upsets me to see so much knowledge in non-STEM areas, because it’s being completely ignored. Seriously, does anybody have even the slightest expectation that your CEO has studied and mastered leadership theory at the depth that is expected for introductory level of practice in mathematics? No. We accept the idea of their wisdom deriving “Straight from the gut” and that such judgement is valued at thousands of times as much as their formally-educated workers. Even in academia it seems, non-stem phds and professorships are granted on the basis of emotion-level reasoning about political belief and activism rather than candid rational inquiry.

Being a CEO is more about assuming risks, that's where authority comes from. As a formally educated employee you get your check every month and if you're not happy you can hop to another job easily, but a business runs the risk of wasting its money on a grand scale.

Risk is assumed by founders. I think that's different. Through years of working at a bunch of 'big' places, the selection of leaderships seems more and more arbitrary. The mechanical aspects of the CEO job are not dissimilar from any other job in the company, and one would expect an expert to occupy that position. I didn't intend to make that comment about pay, except that we have collectively accepted a system where an ambiguous set of emotional reactions are massively better-compensated than formal systems of knowledge and analysis.

Nice! I've been meaning to check out "Linear algebra done right" for a while

Sadly this is the latest (3rd) edition. IMHO they ruined the beautiful typesetting typical of Springer math books with tons of super distracting boxes.

I guess the book is a victim of its own popularity, as other books such as Abbott still keep their old look.

My 2 favorite linear algebra books now have broken typesetting. Halmos is too antique, and would benefit greatly from modern LaTeX. Axler is now too flashy.

Don't know if I'm the only one with that pet peeve, but a textbook with ugly superfluous boxes is almost reason enough for me to drop it right away. Reminds me too much of all my terrible high school textbooks.

I agree with this! I prefer reading Axler's 2nd edition because it is a lot less distracting.

I did not know that the 2nd edition of the book is typeset in a simple black and white style. Thank you for sharing!

Skiena's Algorithm Design Manual and Axler's Linear Algebra Done Right stood out for me when skimming the list.

Both are regularly recommended in HN book threads.

The texts Philosophical and Mathematical Logic and Mathematical Logic should probably fall under Mathematics and Statistics rather than Religion and Philosophy, but seems a nice resource otherwise!

OP here. Thank you for pointing this out.

I didn’t manually categorise it, just extracted the Springer categories. Will manually review this tomorrow.

IMO it's quite perfect the way that is now and I would suggest keeping it categorised the way that Springer has categorised it.

But if you still decide to create your own categorisation of the books, I think you should have two different views; the Springer categorisation as the default view and your own categorisation as an alternate view.

Huh interesting, I'd guess they're basing that on the series the book is published in, which can often be misleading.

"Philosophical and Mathematical Logic" is by Harrie de Swart affiliated with the Faculty of Philosophy, Erasmus University Rotterdam, and the Department of Philosophy, Tilburg University.

> Throughout the text, the author provides some impressions of the historical development of logic: Stoic and Aristotelian logic, logic in the Middle Ages and Frege's Begriffsschrift, together with the works of George Boole (1815-1864) and August De Morgan (1806-1871), the origin of modern logic.

Those topics feel pretty typical of logic as studied in philosophy depts. I think traditionally, a lot of logicians working with formal or mathematical logic worked out of philosophy departments. I've been told that more recently, a lot of them now work out of Comp Sci departments. There just isn't enough of them in one place to have a Dept of Logic I guess! :)

"Mathematical Logic" by Roman Kossak is probably mis-categorized. Topics look like typical math topics. Affiliation is Math Dept at CUNY, Bronx Community College campus. Yeah, probably should be under math section.

Mathematical Logic and Philosophy cross over quite a bit, but I think any text which features proof sketches of Löwenheim–Skolem (present in Philosophical and Mathematical Logic) qualifies as a mathematics text. At any rate the category Religion and Philosophy is wrong, Analytic Philosophy maybe.

Totally agree that formal and mathematical logic crosses over lots between math and philosophy folks (well, analytic philosophers and logicians). Categorizing books is pretty subjective. Looking at the table of contents, it feels like a work from someone working in the philosophy dept. There's an entire chapter for each of Modal Logic, Philosophy of Language, Intuitionism and Intuitionistic Logic, and Fallacies and Unfair Discussion Methods --- when I learned those things, I learned them from philosophy department courses. Certainly it was a very analytic philosophy department, so I'm totally on board with putting it under an analytic philosophy section. I seem to notice publishers and bookstores like to group religion and philosophy together though, so maybe that's why Springer did that. And yeah, I agree that it'd be a better fit for an analytic philosophy section.

I found it funny that the book "Robotics" (by Siciliano et al.) isn't in the category "Intelligent Technologies and Robotics". There's another book titled "Robotics" (by Mihelj et al.) in that category though.

There are so many and only so much time to read. It takes a lot of time to finish any book properly. And I have already much books in my backlog.

Same here. With so much high-quality information and entertainment available the choice of what to invest time in becomes harder and harder, eventually consuming a significant portion of that limited time budget to begin with.

In a lot cases skimming and reading some parts is already very beneficial

Here's a script to automatically download the entire set: https://github.com/tempname1024/springer-dl

Edit: I reported the issue below, and the author has fixed it.

Note: that will fail on Unix/Linux to get one book, "Systems Programming in Unix/Linux" [1].

The script wants to put the book in a directory named with the book's title, and wants to include the title in the name of the PDF and/or EPUB files. The title includes a "/", which is the directory separator on Unix/Linux, so this does not work out well.

You'll end up with a "Systems Programming in Unix" directory, containing a "Linux" subdirectory, and no PDF or EPUB files.

As far as I can tell, it gets everything else fine.

If you only want PDF files, make this change before running it:

  <         links = [x for x in p.links if 'content/pdf' in x or 'download/epub' in x]
  >         links = [x for x in p.links if 'content/pdf' in x]
A PDF only download of everything is 4.5 GB. There are 383 books.

[1] https://link.springer.com/book/10.1007%2F978-3-319-92429-8

I added filtering the list by category/title and sent you a PR. Thanks for your work!

Not sure if it's still up to date, but I appreciated the Quantum Physics text by R. Shankar when I was studying physics as a supplement to the courses I was taking for 4th/5th year physics.

Shankar is a great textbook. Up to date in the sense that the core material hasn't changed a huge amount in the last 50 years. I didn't see it recommended in many reading list at university, but I found it... somehow? It takes a totally different approach to teaching QM by introducing vector spaces up front and sneaking in the notation for bra/kets, commutation and other goodies. Before you know if you're happy with things like <A|B> and it comes as much less of a surprise later on when used in a QM sense. The book also introduces Hamiltonian mechanics so that by the time you get to the QM postulates, you get them in a much more rigorous form than the sort of hand-waving you'd get in other introductory textbooks.

I think unfortunately it does end up as a supplement as actually working through the first few chapters would take some time. However I think for getting a handle on the mathematics underneath QM, it's great. It covers a lot of stuff that's either optional or not covered on most undergraduate physics courses.

Any recommendations (in whatever area), apart from Elements of Statistical Learning and Linear Algebra Done Right? And are their Probability and All of Statistics books any good?

Can vouch for, Wasserman. Is good, imho.

Also lectures on YouTube and course info @ cmu.

There some pretty interesting, weirdly specific subjects in there:


Also some cool stuff related to some projects I'm working on. Unfortunately I find most textbooks involve a ton of mathematics and notation I'm not familiar with.

Now if only we had a good dependency graph for these...


I've been reading Types and Programming Languages by Benjamin Pierce. It's a fantastic book and I plan on eventually going through the whole thing. For now, I've been skipping around a bit to get to the stuff I'm interested in first. Pierce has included a lovely dependency graph between all of the chapters so you know what to read and in what order. It's super nice.

I can't figure out the copyright aspect of this; yes, it's probably fine for everyone to download them while they're available, but will you be theoretically allowed to keep them after they end the promotion?

I'd say yes, but maybe that's not what they meant... or maybe it's such a small revenue loss that they don't care

These are download copies, not streaming. What happens vis a via copyright after Springer stops offering these downloads?

What are you saying??

That we don't know if it will be legal to keep the files after they stop offering them (as it seems will happen).

Of course it's unlikely that some cop will come and verify that you deleted them, but someone prefers to stay legal/honest

This search on Springer also yields good results https://link.springer.com/search/page/3?facet-content-type=%...

When I first saw the books, they presented the number of citations. It was great to infer the book quality. I don't find this info anymore.

Citations, mentions, and downloads are all on the actual Springer site and not the GitHub link it appears.

I didn't see it. Maybe it is the mobile version.

Off topic, but last time I posted a link to upenn's online books page, it went unnoticed. Great resource IMO.

After some sed and wget, the total comes to 11GB.

You must be the reason I get a captcha to read a single book.

If they were smart, they would throw up some torrents themselves. One for each category, one for the whole library. Make use of the people who would leave up their p2p client to spread the load. Everybody wins!

Could you share your scripts, for those or us with lesser scripting skills

Putting out a script runnable by everybody might put too much strain on Springer's servers (and their generosity). I'll give some steps that the motivated can follow instead.

1. Extract a list of all a.btn links from the OP page (382 total).

2. Issue a HEAD request and follow redirects for each link to get the actual page URL (I used curl for this).

3. The page URL contains the book ID. Compare a sample URL's format with its PDF and ePub links. Each page URL can be easily munged into both (sed here).

4. The download pages are captcha-protected, but there's a <noscript> escape hatch in the source. Find it and pass the appropriate data in a POST request to the final file URLs (wget).

A torrent could share the load perhaps

Someone released this a bit ago: https://github.com/alexgand/springer_free_books

Looks like this script is not working anymore. I wanted to try the script when I saw this couple of months back in twitter. And in the process downloaded the whole catalogue.

Perhaps we could all seed your torrent?

I don't torrent, and it's silly to violate copyright on something they're literally offering for free...

Bear in mind that while they're letting you download it for free, Springer doesn't permit redistribution.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact