This topic contains 134 replies, has 2 voices, and was last updated by Josh 2 days, 8 hours ago.
October 3, 2018 at 3:46 pm #30075
It just seemed like people believe “Minecraft worlds are so big that if you keep walking, you could find ANYTHING.”October 3, 2018 at 5:09 pm #30078
You guys crack me up big time. Plagiarism cannot exist in the Library because the Library contains literally nothing! It only generates things on demand! Why can’t you people understand this??October 3, 2018 at 8:36 pm #30088
@Wade: More or less. It’s a slightly weak comparison though, given that Minecraft generates its worlds with strict rules, so that specific world features can be created. The Library has no rules like that, except for “only lower-case Latin alphabet characters, comma, period, and space”, which would basically just be the blocks. In that regard, it’s a bit more accurate to compare Minecraft to the Search feature’s “with random English words” results. If you were to search for a Spanish phrase such as “Hola, me llamo Delengroth”, the rest surrounding your results will only contain English words. It would especially stick out if you were to click the Anglishize feature, much like how in your world, a house you build will stand out against the surrounding landscape.
@Ed Gore: Nobody is denying that the pages are generated on demand. It’s just that having a website which allows you to navigate those permutations is superficially not that different than literally having them stored on the web server’s disk space. It’s sort of like a ZIP file. ZIP files don’t actually contain files, but rather, instructions on how to rebuild specific files. Whether you send someone the original file, or a ZIP, in the end they’ll have the same file either way. By that analogy, it makes no difference whether this website has every page in a TXT file, or instructions (the algorithm) to generate those same exact TXT files.October 3, 2018 at 9:14 pm #30089
Well for your philosophical ponderings, of course it makes a difference. You care about application and practicality only when it is convenient to you! But the truth is, these pages do not exist and you cannot simply mask that point! Instructions do not constitute content! Expose the lie to the people of Babel!October 3, 2018 at 9:51 pm #30093
Jesus, some of these responses are… hard to read. It’s unfortunate that so many people miss the basic concept of this website.October 3, 2018 at 11:10 pm #30096
@Ed Gore: Everyone understands that the pages don’t exist on a hard drive. Storing them all would be impossible. The point is that this website makes it possible to browse them all with math, in order to get around the physical limitations.
I made a simple example to show you what I mean. I call it the “Alphabet of Babel”:
The Alphabet of Babel is a vast library of 26 pages, each one is only one character long. This library contains all of the possible arrangements of 1-letter pages. Simply type in the page number you wish to see, and the page will be there. If you wish to find out what page a letter is on, use the Search.
The Library on this website works in basically the exact same way. The only difference is that Jonathan used random generation to scramble the pages to make it more interesting than a linear organization in alphabetical order of all the pages from “aaa…” to “zzz…”
An important thing to note though. Notice that in my code, I don’t have any alphabet listings, nor is it saving them to a hard drive or anything. It generates the page (or letter) on demand, just like this Library. Would you say then, that all of the pages in the Alphabet of Babel don’t exist? Or do they only exist once you discover them?
The point is, the pages do exist in theory and principle, and can be navigated. They don’t literally exist all at once on a hard drive at this moment, but they “exist” nonetheless. It’s silly to argue semantics at this point.October 3, 2018 at 11:20 pm #30097
The search button of your program does nothing! Case in point, the Library is a mirage! And your alphabets don’t exist at all; they exist for the brief moment they appear on a screen, and once that goes away, nothing exists but the code. Saying that something exists because the capabilities to make it exist is rubbish! Does a vehicle that goes to the center of the Earth exist just because it can exist? Same thing!October 8, 2018 at 3:50 pm #30588
I copied said gibberish from one page, inserted it in the search bar. The page I copied it from did not show up in the results. I browsed through all exact matches, it should have been exactly the same as I copy pasted it.
If this site does what it says it does and the pages are fixed like it says they are, then why aren’t they fixed?October 8, 2018 at 5:00 pm #30591
Because there are exponentially many copies of each page, so the chance of you finding the said contents in the exact same location is second to none.October 8, 2018 at 11:04 pm #30602
@ripded: If you click on the “more random char matches”, near the top you’ll see there are “20 of ~29^3215 possible matches with random characters” and a “Next Results” link to keep going. You’d have to search for quite a while.
Secondly, the search box tends to bug out when pasting text. It tends to trim off around 40 or so characters from the bottom of the page (probably because of new line characters). Look at the last few letters on a page you copied, and then the last few in the search box after you paste. You’ll notice that some are missing. Make sure you copy the rest of them. You can double check if the search box contains all of the page by copying all of it, and double checking that all 3200 are there with a character counter, such as: http://www.charactercountonline.com/October 9, 2018 at 11:13 am #30610
Look, first of all, there is such algarithm that generates random data (involving the capitalization of Hash functions), second, this isn’t a matter of actual “existance” of form, but the “existance” of the concept and what we observe.October 9, 2018 at 11:30 am #30611
Very interesting point, but still, you haven’t interpretated correctly Borges’ original idea. The idea of an infinite number of subjects is itself a “illusion” to the individual, because there are limits in human observation and the idea as a whole cannot be observed. However, if this idea of an infinte number of subjects is an “idea”, the subject of the idea exists because it can exist and because its existance is beyond human observation, also because it does not overlap human observation. For example, the supernatural does not exist if its concept overlaps human observation of its subject as a whole, mind here that this point discusses that the subject is observed “as a whole” but not as fragments or reflexions. And thus if you do believe that the concept of the library is true, and observed fragments of the library’s existance, the library exists. This is a concept of existance adapted by objective idealists, within whose category Borges is thought to belong.October 9, 2018 at 4:37 pm #30616
If I can see half of a flower, does the other half exist, even if I’ve burned it off?October 13, 2018 at 11:20 pm #30748
Correct me if I am wrong but I thought this website used random non-repeating base 10 such as 0-9 then encrypted it to base 32 with all the English alphabet and some extra to create 3,200 character encrypted text that acts as the library itself then the search engine takes the characters you entered and encrypts them to base 10 and finds a string of numbers that has those same characters and there is your answerOctober 14, 2018 at 1:10 am #30756
From what I’ve been able to piece together, the Library uses a reversible linear congruential generator as its PRNG. This allows you to find the inverse using Euclid’s Algorithm to find the greatest common divisor between the numbers 29 (character possibilities) and 3200 (page length). I believe 29 was carefully chosen since the LCG requires a prime number.
The LCG is then used to generate a 3200-byte long number and then converted to base 29. To find a page’s location (seed), you convert the page contents back out into a base 10 integer and run through the Inverse LCG. For search queries shorter than 3200 characters, just randomly fill in the rest.
Keep in mind, I haven’t been able to fully reverse engineer the functionality, so this information might not be completely accurate. For example, I still don’t know how the sequence of books are organized, and also how the wall/shelf/volume/page numbers play into the seed (could just be as simple as appending?). At any rate, you can read this StackOverflow thread that Jonathan posted up when he was originally seeking help in designing the Library of Babel: