Books online

[fladiv]

Since this wonderful site hosts so many good scanned books, I believe that their downloading size becomes an issue for some people. However, choosen format, PDF, curently is not very appropriative for such goals. As I see it, PDF is designed as a mainly vector format to store computer-generated material, as opposed to the scanned material.<br /><br />Today, there exist a great new (relatively) digital format designed especially for storing scanned material in digital form as efficient as possible, retaining quality.<br />I am talking about DjVu format. It seems that a scanned books of a decent quality, as books on this site are, take approximately 100 KB per 10 pages. Thus, FGB could have taken about 4MB instead of 12, as it is now, although I didn't check this myself because I currently lack decent internet connection. But it worth checking...<br /><br />The software for conversion into this format is not freely available, but there is a free online conversion service at http://any2djvu.djvuzone.org/ that performs such conversion - either directly from the scanned images, or from some of other formats, including PDF.<br /><br />So, perhaps, site maintainers should look into perspective of moving to DjVu, or at least offer books in this format as an alternative download.

[klewlis]

interesting... i've been in this industry for years and have never heard of this. especially interesting is that it's built on an open standard system... hmm...<br /><br />how does it compare as far as portability and printing? since those are the key selling points of pdfs and important to textkit users, that should be taken into consideration.<br /><br />unless it becomes more widespread and the demand increases, I'm not sure it would be advantageous for the administrators to spend the time required to convert documents into a second format just for a few users... although jeff will have to speak to that ;)

[Lumen_et_umbra]

Allz' Iz' cane saiz' iz' thats I preefirz' dah' reelz' buhkz'. I think I just wrote something in German or Dutch or something like that...

[fladiv]

As for accessability and printing - there are browser plugins available that render DjVu documents in browser window, with specific controls. Unfortunately, Mozilla is not supported on Windows, so I have to use IE for that. There are plugins for UNIX platforms and Macs as well, although I didn't try them. All this stuff is linked from http://djvuzone.org/, at Downloads section.<br /><br />There is also an elder product, called DjVu Solo, which is a stand-alone viewer for DjVu documents. However, I don't know where can it be downloaded currently, or perhaps I have searched their site not good enough. Anyway, the IE plugin can handle printing itself - just right-click on a document and choose Print... I believe it is similiar for the other platforms.<br /><br />It is also worth to note that the library allowing to handle DjVus to anyone intrested in writing his own viewer, is GPL'd, and can be accessed via Sourceforge (linked from the mentioned site).<br /><br />Quality of printing is as good as original scanning quality - DjVu basically is an image compression technology specially suited for things like black-and-white books, plus a format to bound it and different other existing formats into one document, similar to PDF.<br /><br />BTW, I've used to print "Teach Yourself Greek" book that I have acquired as DjVu (335 pages, 5MB... oh well, still not 12 ), but later I have came upon this site and I think I will switch to White. TYG seems to be too aligned upon english-speaking people, while my native is Russian... Besides, White did write for children, which also simplifies things <br /><br />Oh, and, Lumen_et_umbra, did this happen to be a mimic of my poor English? Any corrections are warmly appeciated...

[Keesa]

Speaking strictly for myself, I like .pdf format...and I know how to use Acrobat Reader, which is not something I can say for every technology out there...far from it! <br /><br />If you do decide to switch, Jeff, I hope you'll keep the .pdf format as an option! I am not very good with computers, and the idea of having to learn another program is, quite honestly, a few hairs short of terrifying. Please keep .pdf as an option! <br /><br />Keesa

[Jeff Tirey]

Hi Fladiv:<br /><br />This format has been suggested to us in the past and I do agree that smaller is better because it means more access for more visitors. But we'll be sticking with PDF for many reasons other than smaller files sizes which I'm sure DjVu can deliver.<br /><br />1. our download sizes are less of a problem than one might think. For example, we have five books of the iliad in one rather large PDF file. For convenience, I created a separate file for each book to make downloading easier - hardly anyone downloads the smaller files.<br /><br />2. Time is also on our side as every year more and more users enjoy high bandwidth connections.<br /><br />3. Sometime soon, we'll upgrade our files to the lastest version of PDF which has a new compression technology that should cut the file sizes in half.<br /><br />4. Nothing comes close to PDF accessibilty, almost everyone has it. It would be difficult indeed to convince visitors to actually download and install a program on their machine in order to view our files.<br /><br />thanks for the suggestion.<br /><br />jeff

[mingshey]

I'm a DjVu fan myself and have been trying to bind the Woodhouse English-Greek dictionay(images over 600MB, and the best result I got with pdf sizes 335MB) into a DjVu file. But the free trial license allows only 250 pages for one conversion process. I'm going to try their web conversion service.<br /><br />I also thought about suggesting it but changed my mind, 'cause I thought there would be many reasons the webmasters say they stick to pdf. PDF's are rather easier to handle and straightforward to use, and have many ways to make. <br /><br />DjVu files are compact, and faster to download. It uses so-called wavelet compression, and recently fractal compression, too. It separates the text of scanned materials as the "foreground" images and the pictures and backgrounds into "background" images. And it speeds up the download, to say, download the text first(a small volume of data) and you can ignore the longer download time for the background. <br /><br />But the separartion process is not perfect that sometimes tiny parts of the text are included in the background. The effect of this imperfectness is, when you want to print only the foreground images(the texts)--- for it often looks tidier if you elliminate the dusty and irrelavant background---, it becomes defective, even though the fore- and background images in whole is not defective. And this is quite annoying. :-\<br />