ZWIBook: Difference between revisions

From Encyclosphere Project Wiki
(+ foreign word search fixed)
Line 11: Line 11:
#* Various other issues were found with very small ZWI files; many are music-only, or even software-only, files, and not books at all. Obviously these should not be counted as books, and we will be removing (and replacing) them in v1.1.
#* Various other issues were found with very small ZWI files; many are music-only, or even software-only, files, and not books at all. Obviously these should not be counted as books, and we will be removing (and replacing) them in v1.1.
#* There are 34 titles found in "<code>/book_zwis"</code> but which are not contained in the ZWIBook search database; therefore, while the earlier versions of ZWIBook do have a copy, such copies are not accessible from ZWIBook. This is a known bug. The following Project Gutenberg IDs are not included as of October 7: <code>8776, 13395, 7366, 7766, 7844, 8555, 9304, 13320, 14398, 15083, 15258, 15891, 16187, 16188, 16189, 16190, 16883, 17073, 20144, 22662, 23057, 23063, 23326, 25105, 25107, 25109, 25726, 28876, 2953, 44800, 4978, 54963, 6036, 6672</code> We apologize for the issue and will supply new builds of the ZWIBook software (with installation instructions) below soon (v1.1).
#* There are 34 titles found in "<code>/book_zwis"</code> but which are not contained in the ZWIBook search database; therefore, while the earlier versions of ZWIBook do have a copy, such copies are not accessible from ZWIBook. This is a known bug. The following Project Gutenberg IDs are not included as of October 7: <code>8776, 13395, 7366, 7766, 7844, 8555, 9304, 13320, 14398, 15083, 15258, 15891, 16187, 16188, 16189, 16190, 16883, 17073, 20144, 22662, 23057, 23063, 23326, 25105, 25107, 25109, 25726, 28876, 2953, 44800, 4978, 54963, 6036, 6672</code> We apologize for the issue and will supply new builds of the ZWIBook software (with installation instructions) below soon (v1.1).
# Some non-Latin character sets were not fully supported in the book title and author search. This has been fixed in v1.1.
# Highlighting and bookmarking does not work in the table of contents and other front matter of some books. This is a known limitation, due to the complexity of assigning IDs to complex and unpredictable HTML. This will ''not'' be fixed in version 1.1. Most of the body of the text in all books is, as far as we know, bookmarkable and highlightable. There may be a few exceptions.
# Highlighting and bookmarking does not work in the table of contents and other front matter of some books. This is a known limitation, due to the complexity of assigning IDs to complex and unpredictable HTML. This will ''not'' be fixed in version 1.1. Most of the body of the text in all books is, as far as we know, bookmarkable and highlightable. There may be a few exceptions.
# There are some remaining minor styling issues (imperfectly centered text, for example). If you want to collect those issues, we might fix them in a later version. (There will be no change on this in v1.1, because we don't know of any significant issues.)
# There are some remaining minor styling issues (imperfectly centered text, for example). If you want to collect those issues, we might fix them in a later version. (There will be no change on this in v1.1, because we don't know of any significant issues.)

Revision as of 14:57, 28 October 2024

ZWIBook is a project by Larry Sanger, with help from many (not least of which ChatGPT), to put all of Project Gutenberg books (as of December 2023), in ZWI format, on a flash drive with a bespoke reader (available for Windows, Mac, and Linux). The project is available for purchase for $50, or $100 for the numbered and signed drives. We thank Henry Sanger for help with CSS and design, and we also thank Project DARA's Louie and Dorian for creating and making available the ZWI files of Project Gutenberg books, which they prepared and which are available here.

Known issues with version 1.0

  1. Several issues with the original collection ZWI book files in the /book_zwis directory came to light near the beginning of October 2024. Most users will not encounter these issues, as they affect fewer than 0.2% of the files. Certain files should be removed or repaired (details below) from ZWIBook flash drives from October 7, 2024 and onward; the original version of ZWIBook drives is hereby dubbed version 1.0, and the new version will be called version 1.1. When our work on this is done (as it should be by approximately October 10), then we will make a ZIP file containing all changed files, with instructions, available; this will let users update their copies of the ZWIBook files, making their local copies equivalent to version 1.1. Here are the issues:
    • At the beginning of August about 85 titles were deleted from the PG database for potential copyright reasons. This announcement was made on a small Project Gutenberg mailing list, but as soon as we became aware of the issue, we put a temporary pause in our marketing and distribution. These ~85 files should be removed from version 1.1.
    • About 871 more items will be deleted, because they were copyrighted; these items were part of the PG collection, and according to the license, the 20% royalties is given to PG as a donation, which PG waived. Nevertheless, after PG's August announcement and after discussing more with PG, we have decided to delete these items as well, out of an abundance of caution. They will be made up for with items added to PG between Dec. 2023 and sometime fall 2024.
    • The Gutenberg Webster's Unabridged Dictionary, for complicated reasons, causes the software to appear to hang, i.e., to freeze and become inaccessible. (In fact, it will eventually load, but only a blank page, and only after several minutes.) Worse, the software's internal data file recording the latest-viewed book is stays on that page. This appears to render the software unusable. The issue can be fixed as follows: (1) Go to your home directory, or wherever the software data is standardly saved in your system. (2) Look (if necessary showing "hidden files") for .zwibook. (3) Inside that directory you will find latest.txt: simply delete that file and the problem will be fixed. v1.1 drives will omit the offending title.
    • A couple ZWI files are completely missing either the HTML or the TXT file for the book. We will reconstruct the ZWI files for these items in v1.1.
    • A small number of files (15-20?) are "RTF-only", i.e., they do not have HTML or TXT files, but only RTF files; moreover, the RTF files are not part of the ZWI files. In v1.1, we will be removing these because the vast majority are copyrighted.
    • Various other issues were found with very small ZWI files; many are music-only, or even software-only, files, and not books at all. Obviously these should not be counted as books, and we will be removing (and replacing) them in v1.1.
    • There are 34 titles found in "/book_zwis" but which are not contained in the ZWIBook search database; therefore, while the earlier versions of ZWIBook do have a copy, such copies are not accessible from ZWIBook. This is a known bug. The following Project Gutenberg IDs are not included as of October 7: 8776, 13395, 7366, 7766, 7844, 8555, 9304, 13320, 14398, 15083, 15258, 15891, 16187, 16188, 16189, 16190, 16883, 17073, 20144, 22662, 23057, 23063, 23326, 25105, 25107, 25109, 25726, 28876, 2953, 44800, 4978, 54963, 6036, 6672 We apologize for the issue and will supply new builds of the ZWIBook software (with installation instructions) below soon (v1.1).
  2. Some non-Latin character sets were not fully supported in the book title and author search. This has been fixed in v1.1.
  3. Highlighting and bookmarking does not work in the table of contents and other front matter of some books. This is a known limitation, due to the complexity of assigning IDs to complex and unpredictable HTML. This will not be fixed in version 1.1. Most of the body of the text in all books is, as far as we know, bookmarkable and highlightable. There may be a few exceptions.
  4. There are some remaining minor styling issues (imperfectly centered text, for example). If you want to collect those issues, we might fix them in a later version. (There will be no change on this in v1.1, because we don't know of any significant issues.)
  5. There might still be a few books that are essentially unreadable because of styling issues. We do not currently know of any such books. We went through hundreds of books and resolved all issues we saw, but we did not test all the files in the /zwi_books folder. For such books, you can still read the HTML by exporting the ZWI (see under File), unzipping the ZWI file, and clicking on the HTML (or, if text-only, TXT) file.
  6. Unusually large books can take many seconds (up to 20 or so) to load. These are rare, however.

Downloads

This is a placeholder. Version 1.1 executables and data updates will be downloadable in this space.